0480 coverfront - United Nations ESCAP

Vol. 18, No. 1, June 2011

IN THIS ISSUE: Investment and economic opportunities: urbanization, infrastructure and governance in the North and South of India Reaching a universal health insurance in Viet Nam: challenges and the role of government Revisiting sectoral liberalization: an alternative to the Free Trade Area of the Asia-Pacific? Implications for the Philippines Examining poverty and inequality across small areas of Islamic Republic of Iran; toolmaking for efficient welfare policies The relationships between the socio-economic profile of farmers and paddy productivity in North-West Selangor, Malaysia

The secretariat of the Economic and Social Commission for Asia and the Pacific (ESCAP) is the regional development arm of the United Nations and serves as the main economic and social development centre for the United Nations in Asia and the Pacific. Its mandate is to foster cooperation between its 53 members and 9 associate members. It provides the strategic link between global and country-level programmes and issues. It supports Governments of countries in the region in consolidating regional positions and advocates regional approaches to meeting the region’s unique socio-economic challenges in a globalizing world. The ESCAP secretariat is located in Bangkok, Thailand. Please visit the ESCAP website at for further information.

The shaded areas of the map indicate ESCAP members and associate members.

Asia-Pacific Development Journal

Vol. 17, No. 1, June 2010

i


Vol. 17, No. 1, June 2010

ASIA-PACIFIC DEVELOPMENT JOURNAL Vol. 18, No. 1, June 2011

United Nations publication Sales No. E.11.II.F.7 Copyright © United Nations 2011 All rights reserved Manufactured in Thailand ISBN: 978-92-1-120625-8 e-ISBN: 978-92-1-054871-7 ISSN: 1020-1246 ST/ESCAP/2599

This publication may be reproduced in whole or in part for educational or non-profit purposes without special permission from the copyright holder, provided that the source is acknowledged. The ESCAP Publications Office would appreciate receiving a copy of any publication that uses this publication as a source. No use may be made of this publication for resale or any other commercial purpose whatsoever without prior permission. Applications for such permission, with a statement of the purpose and extent of reproduction, should be addressed to the Secretary of the Publications Board, United Nations, New York.

ii


Vol. 17, No. 1, June 2010

Advisory Board Members Dr. YILMAZ AKYÜZ Chief Economist, South Centre (Former Director and Chief Economist, UNCTAD), Switzerland Dr. RASHID AMJAD Vice Chancellor, Pakistan Institute of Development Economics (PIDE), Pakistan Ms. MYRNA AUSTRIA Vice-Chancellor for Academics, De La Salle University, Philippines PROFESSOR RAJESH CHANDRA Vice-Chancellor and President, University of the South Pacific, Fiji PROFESSOR TAKATOSHI ITO Professor, Graduate School of Economics and Graduate School of Public Policy University of Tokyo, Japan Dr. MURAT KARIMSAKOV Chairman of the Executive body of the Eurasian Economic Club of Scientists, Kazakhstan Dr. SAMAN KELEGAMA Executive Director, Institute of Policy Studies, Sri Lanka PROFESSOR DEEPAK NAYYAR Jawaharlal Nehru University (Former Chief Economic Adviser to the Government of India), India PROFESSOR REHMAN SOBHAN Chairman, Centre for Policy Dialogue, Bangladesh Dr. CHALONGPHOB SUSSANGKARN Distinguished Fellow, Thailand Development Research Institute, Thailand PROFESSOR YU YONGDIN Chinese Academy of Social Sciences, China

Editors Chief Editor Dr. Nagesh Kumar Director, Macroeconomic Policy and Development Division Managing Editor Dr. Aynul Hasan Chief, Development Policy Section, Macroeconomic Policy and Development Division

iii


Vol. 17, No. 1, June 2010

Editorial statement The Asia-Pacific Development Journal is published twice a year by the Economic and Social Commission for Asia and the Pacific. Its primary objective is to provide a medium for the exchange of knowledge, experience, ideas, information and data on all aspects of economic and social development in the Asian and Pacific region. The emphasis of the Journal is on the publication of empirically based, policy-oriented articles in the areas of poverty alleviation, emerging social issues and managing globalization. The Journal welcomes original articles analysing issues and problems relevant to the region from the above perspective. The articles should have a strong emphasis on the policy implications flowing from the analysis. Analytical book reviews will also be considered for publication.

Manuscripts should be sent to: Chief Editor Asia-Pacific Development Journal Macroeconomic Policy and Development Division ESCAP, United Nations Building Rajadamnern Nok Avenue Bangkok 10200 Thailand Fax: (662) 288-3007 or (662) 288-1000 Email: [email protected]

iv


Vol. 17, No. 1, June 2010

ASIA-PACIFIC DEVELOPMENT JOURNAL Vol. 18, No. 1, June 2011 CONTENTS Page Kala Seetharam Sridhar, A. Venugopala Reddy

Investment and economic opportunities: urbanization, infrastructure and governance in the North and South of India

Giang Thanh Long

Reaching a universal health insurance in Viet Nam: challenges and the role of government

47

George Manzano, ¨ Myrene Bedano

Revisiting sectoral liberalization: an alternative to the Free Trade Area of the Asia-Pacific? Implications for the Philippines

73

Arman Bidarbakht Nia

Examining poverty and inequality across small areas of Islamic Republic of Iran; toolmaking for efficient welfare policies

125

Mahmudul Alam, Chamburi Siwar, Basri Talib, Mohd Ekhwan bin Toriman

The relationships between the socio-economic profile of farmers and paddy productivity in North-West Selangor, Malaysia

161

v

1


Vol. 17, No. 1, June 2010

Explanatory notes References to dollars ($) are to United States dollars, unless otherwise stated. References to “tons” are to metric tons, unless otherwise specified. A solidus (/) between dates (e.g. 1980/81) indicates a financial year, a crop year or an academic year. Use of a hyphen between dates (e.g. 1980-1985) indicates the full period involved, including the beginning and end years. The following symbols have been used in the tables throughout the journal: Two dots (..) indicate that data are not available or are not separately reported. An em-dash (—) indicates that the amount is nil or negligible. A hyphen (-) indicates that the item is not applicable. A point (.) is used to indicate decimals. A space is used to distinguish thousands and millions. Totals may not add precisely because of rounding. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the Secretariat of the United Nations concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Where the designation “country or area” appears, it covers countries, territories, cities or areas. Bibliographical and other references have, wherever possible, been verified. The United Nations bears no responsibility for the availability or functioning of URLs. The opinions, figures and estimates set forth in this publication are the responsibility of the authors, and should not necessarily be considered as reflecting the views or carrying the endorsement of the United Nations. Mention of firm names and commercial products does not imply the endorsement of the United Nations.

vi


Vol. 18, No. 1, June 2011

INVESTMENT AND ECONOMIC OPPORTUNITIES: URBANIZATION, INFRASTRUCTURE AND GOVERNANCE IN THE NORTH AND SOUTH OF INDIA Kala Seetharam Sridhar and A. Venugopala Reddy*

There are substantial disparities across the southern Indian and northern (Bihar, Madhya Pradesh, Rajasthan and Uttar Pradesh) states in terms of fundamental economic phenomena, such as per capita net state domestic product, rural and urban poverty rates, and investment flows, with the southern states taking a lead over their northern counterparts. In this paper, we make an attempt to understand what factors have caused some states to grow faster than others. We examine human capabilities, skills and awareness, resources and the efficiency of their utilization, extent of urbanization, good governance including law and order, and infrastructure across the two group of states. We conclude that the upward shift in per capita income, downward trend in poverty reduction and investment flows that occurred in the south relative to that in the northern states can be explained partly by the advantage the former had in terms of human capabilities, infrastructure, urbanization and some law and order conditions and partly by the economic liberalization of 1991.

JEL Classification: O10, O18, P47, R11. Key words: Indian states, regional disparities, southern Indian states, northern Indian states, governance.

* Kala Seetharam Sridhar is Head, Public Policy Research Group, Public Affairs Centre, Bangalore, India, [email protected]; [email protected]; [email protected]; A. Venugopala Reddy, Senior Programme Officer, Public Affairs Centre, Bangalore, India, [email protected]. The authors thank ISAS-NUS for partially funding this work and also thank Samuel Paul for his comments regarding an earlier draft. This paper draws heavily from a recently completed project at the Public Affairs Centre regarding the paradox of India’s north-south divide. The authors also wish to thank the two reviewers of this manuscript for their comments. Any errors are the responsibility of the authors.

1


Vol. 18, No. 1, June 2011

I.

INTRODUCTION

It is now well known that, in India, there is substantial variation in income across states. For instance, the per capita net state domestic product (NSDP) of the southern state of Tamil Nadu was Rs 14,000 (in 1993-94 prices) in 2004-05, whereas the NSDP of the northern state of Uttar Pradesh was only Rs 6,138 for the same year (in 1993-94 prices), less than half that of Tamil Nadu (based on data from the Economic and Political Weekly Research Foundation (EPWRF)). Similarly, according to data from the Department of Industrial Policy and Promotion, Ministry of Industry, Government of India, Maharashtra, Dadra and Nagar Haveli, Daman and Diu together accounted for nearly 32 per cent of all foreign direct investment (FDI) inflows into the country, and Delhi, parts of Uttar Pradesh and Haryana accounted for 18 per cent, whereas Uttar Pradesh and Uttaranchal accounted for less than 0.02 per cent of all FDI inflows into the country. In fact, Dreze and Sen (1997) point out that some of the southern Indian states have been growing at the rate of countries in East Asia, such as Singapore, whereas some states in the north have been crawling at the rates of those in sub-Saharan Africa causing an eventual fault line to develop between the states. While the state has policymaking power, throughout this paper we compare the four southern Indian states (Karnataka, Kerala, Tamil Nadu and Andhra Pradesh) with the northern states of Bihar, Jharkhand, Madhya Pradesh, Chhattisgarh, Rajasthan, Uttar Pradesh, and Uttarakhand. This is defensible as regions can be viewed as common markets which consist of several states and which facilitate the movement of ideas, goods and services. A comparison of this kind between the northern and southern states mentioned above is defensible because Punjab, Haryana and the western states of Gujarat and Maharashtra have been high on the economic performance scale ever since independence. For instance, as early as 1960-61, the per capita incomes of Punjab, Haryana, Gujarat and Maharashtra were respectively Rs 4,923, Rs 4,614, Rs 4,904 and Rs 5,527, compared with only Rs 3,338 for Uttar Pradesh (all in 1993-94 constant prices).1 The southern states surged sometime in the recent past, in the 1980s or 1990s (this is something we will discover in the forthcoming analysis).

1

One could argue that Rajasthan has moved out of this league of poor-performing states recently (as pointed out by Ahluwalia, 2000). Rajasthan has been in this league of slow growing states historically, and its growth is only a recent phenomenon. So we retain Rajasthan in our set of northern states, our objective being to study longer run trends. One could also make a case for including Orissa, which has also lagged economically, in the list of northern states. Orissa, being in the east, does not enter the north-south debate that is the focus of our work here. In the section on research objectives and literature survey, we clearly lay out the scope of our paper.

2


Vol. 18, No. 1, June 2011

With the exception of Rajasthan, which has been only recently surging (see Ahluwalia, 2000), the northern states have remained behind. As Paul and Sridhar (2009) and Ramachandra Guha point out, historically the north was viewed as the region which was growing and where job opportunities were being created whereas the southern states were viewed as laggards. In recent years, however, this has changed, with the southern states surging economically. Hence, a study of the four southern states with their northern counterparts offers interesting lessons in contrast. In our comparisons of the northern region, we include the three newly created states (Uttarakhand, Jharkhand and Chhattisgarh) along with their parent states because the new states were carved due to stark intra-state disparities and their backwardness. Also, when we are comparing the pre-2000 period with the post-2000 period (2000 being the year in which the three new states were created), it is necessary to account for them in the interests of comparability. Before we examine the record of investment into these states, it is instructive to examine the differences in basic economic phenomena, such as NSDP and poverty rates. As discussed, these fundamental economic phenomena show remarkable differences between the southern and northern states when observed over a period of time. Observing these phenomena over a long period of time has the advantage of demonstrating whether such disparities are a recent phenomenon or have existed for a prolonged period of time. Nunn (2009) provides a survey which gives a growing body of empirical evidence pointing towards the important long-term effects that historic events have on current economic development. This paper is organized as follows. The next section presents the research objectives followed by the literature survey, which summarizes past literature on the subject. This is followed by a section on the methodology followed by a description of trends in various explanatory factors – human capabilities, skills and awareness, resources and the efficiency of utilization, extent of urbanization, good governance, including law and order, and infrastructure, which are presented for the southern and northern states. The final section pulls all findings together, summarizes the implications of the work and concludes.

II.

RESEARCH OBJECTIVES

In this paper, we answer the question why some states have grown faster than others. An interesting issue to be explored in this context is whether a productive agricultural sector is a prerequisite for the growth of a manufacturing sector or whether comparative disadvantage in agriculture stimulates the growth of a manufacturing sector to ensure survival. One hypothesis is that with the relative

3


Vol. 18, No. 1, June 2011

growth of manufacturing, the southern states may be growing faster than the northern states, which have a comparative advantage in agriculture. To understand the questions above, we observe trends and differences in economic phenomena, such as per capita income, poverty rates and economic opportunities such as investments and FDI flows observed across the southern and northern Indian states. As indicators of explanatory factors, we examine human capabilities, skills and awareness, resources and the efficiency of the utilization of those skills, the extent of urbanization, good governance including law and order, and infrastructure across the southern and northern states. We present trends in important indicators of these differences in economic phenomena and explanatory factors, following the literature survey in the next section.

III.

LITERATURE SURVEY

There is a vast amount of literature dealing with economic growth and on convergence/divergence in Indian states. There is also a lengthy literature on intergovernmental transfers in India which shows how the fiscal disparities of the poorer states have not been adequately offset by the transfer system and how various types of subnational transfers can discourage equalization (Rao and Singh, 2005). Our paper should not be viewed as an addition to the general literature on interstate disparities. There is another strand of literature which examines the sources and timing of the shift in Indian output growth since the 1980s. This literature addresses a variety of questions such as: When did the shift in growth occur? Was the shift uniform across states? What were the factors causing the shift? Based on a review of this literature, we find that none of the studies take the distinct, north-south approach we take in this paper. We divide this literature survey into parts – one part dealing with the timing and extent of disparities among Indian states, and another part critically summarizing the literature which explains the factors behind interstate differentials, highlighting the contribution of this work. Disparities across Indian states First, we discuss the literature on disparities across the Indian states and the timing of the shift, if any. Nair’s (1982) pioneering analysis covered 14 major states. The study showed that interstate disparities in per capita NSDP, as measured by the coefficient of variation (CV), had declined over the period 1950-51 to 1964-65, but increased between 1964-65 and 1976-77. Unfortunately, this paper is quite dated and does not take into account post-1983 developments.

4


Vol. 18, No. 1, June 2011

Roy Choudhury (1993) examines interstate disparities and reported that the coefficient of variation (CV) of per capita NSDP in current prices had increased between 1967-68 and 1977-78, but declined between 1977-78 and 1985-86. However, the CV in terms of constant prices showed a persistent increase during the entire period 1967-68 to 1985-86. While this kind of analysis is useful for purposes of this work, Roy Choudhury’s (1993) study does not cover enough of the post-liberalization period for us to make an assessment. Dholakia (1994) in his analysis of interstate disparities in growth rates of 20 Indian states over the 30-year period 1960-61 to 1989-90 identified empirically the optimal years of shift in growth separately for each state through the estimation of a kinked exponential trend curve model. This analysis is interesting, but does not delve into causes of the interstate disparities in growth rates, which is much required and is attempted in this paper. Das and Barua (1996) examined several dimensions of regional economic disparities among 23 states/union territories during the period from 1970 to 1992. It was found that interstate inequality increased in almost all sectors. This paper suffers from the same limitation as the earlier papers in that it does not attempt to explain the inequality in the sectors among the states. Mathur (2001) analysed several facets of national and regional economic growth since the 1950s, with a specific focus on the 1980s and 1990s. The study reported a steep acceleration in the coefficient of variation of per capita incomes in the post-reform period of 1991-96, just as we find here (see the next section on trends in various indicators). A tendency towards convergence was noticed within the group of middle-income states, while divergence was evident within the groups of high- and low-income states. Unfortunately, the paper goes no further in explaining the convergence or divergence among the states, but some of its findings are of relevance to what we find in this paper. Kurian (2000) drew attention to interstate disparities by presenting recent data for Indian states on demographic characteristics, social characteristics, magnitude and structure of SDP, poverty ratio, developmental and non-developmental revenue expenditures, eighth plan outlay and its sectoral distribution, disbursal of financial assistance for investment, indicators of physical infrastructure development and of financial infrastructure. The paper found that a sharp dichotomy between the forward and backward groups of states had emerged.2 This paper takes a holistic 2

Kurian’s (2000) forward group consists of Andhra Pradesh, Gujarat, Haryana, Karnataka, Kerala, Maharashtra, Punjab and Tamil Nadu. The backward group comprises Assam, Bihar, Madhya Pradesh, Orissa, Rajasthan, Uttar Pradesh and West Bengal.

5


Vol. 18, No. 1, June 2011

view of development in the states, but does not explain the causes of the observed dichotomy. Kurian also groups together all states with high per capita income and others with low per capita income, without making a distinction as to when these changes occurred. Wallack (2003) finds evidence for a break in the GDP growth rate in the early 1980s. This is close to the result reported by Rodrik and Subramanian (2005). Hausmann and others (2005) analyzed transitions to higher growth in a large cross-national sample, and date the Indian growth break to 1982. However, their paper primarily deals with India in a cross-national sample and attempts to explain the Indian growth take-off in the early 1980s. They do not delve into the subnational or regional levels, as we do here. Virmani (2006) finds that the growth rate of manufacturing in Indian states accelerated after 1980-81, and this contributed to the acceleration in growth of GDP from 1981-82. Virmani finds no additional breakpoints in the 1990s once the breakpoint in 1980-81 is accounted for. It can be stated that the purpose of all these studies appears to be to examine when a break appeared in the growth rate of Indian states without worrying about why or how the break occurred. In contrast to Virmani (2006), Balakrishnan and Parameswaran (2007) find that the break in the growth rate of GDP occurred in 1978-79 – with the 1978-79 take-off in growth occurring prior to the positive break in manufacturing (1982-83). This suggests that the evidence for manufacturing having served as a primary engine of growth through appropriate market reforms is weak. In all fairness, in addition to the literature which summarizes the disparities among the states and the timing of a shift, there is a stream of literature which attempts to explain the interstate growth differentials. The next subsection summarizes this literature. Explanation of interstate growth differentials So far, a number of studies have attempted to explain the interstate differentials among Indian states. One of the earliest studies is a study of the northern and southern states in the United States of America by Olson (1984). This study explains the differences in economic performance in terms of endowments, policies and institutions. The independent variables include: the level of urbanization in 1889, lawyers per 100,000 of the population, and labour union membership and the dependent variable is the economic growth rate. The hypothesis is that distributional coalitions should be more

6


Vol. 18, No. 1, June 2011

powerful in places that have had stable freedom of organization. In the American study, the hypothesis is supported by regression results. Needless to say, this work is inspired by some variables used in the American study by Olson (1984). Ahluwalia (2000) explains interstate differences in the economic performance in terms of market development and the ability of Indian states to take advantage of economic liberalization. He finds and argues that Rajasthan and Madhya Pradesh have performed reasonably well in recent years. Sachs and others (2002) attempted a detailed qualitative assessment of the factors behind interstate growth differentials, and listed several possible hypotheses for the lack of unconditional convergence among Indian states: (a)

The geographical differences in India are larger than in the United States, Europe and Japan;

(b)

Population movements in India respond very slowly to income differentials;

(c)

Policies of the national or state governments do not facilitate convergence;

(d)

Economic convergence is slower at lower levels of economic development, as in India.

They also found that coastal access and climate are also factors in convergence, but they did not take into account the role of governance factors. In a largely agricultural country such as India, agricultural growth also may be expected to have some impact on growth. Panel data regressions by Shand and Bhide (2000) utilizing data from 15 states over three time periods (1972-82, 1982-90, and 1992-95) suggested that agricultural growth has a positive impact on industrial and service sector growth. Agricultural growth, in turn, was affected positively by land productivity in agriculture and negatively by the share of agriculture. While the regression results are useful in understanding growth, the paper does not go beyond economic factors. It is plausible that a state’s initial distribution of income and private investment impacts its current per capita income. Rao and others (1999) analyzed the determinants of growth of per capita SDP with data for the 14 major Indian states. The coefficient on the initial income variable was significantly positive in the regressions for longer periods (1965-94, 1970-94 and 1975-94). The variable indicating private investment was found to be the most important determinant of growth. Next in importance was the literacy variable. We find this paper ignores the role of non-economic factors, such as governance.

7


Vol. 18, No. 1, June 2011

Nagaraj and others (2000) gathered various factors together and used panel data for 17 states for the years 1960-94. The growth regression included, apart from lagged per capita SDP, the share of agriculture, the relative price of agricultural and manufactured goods, several infrastructure indicators and fixed effects for states as explanatory variables. Evidence for conditional convergence was found. The results of the study suggested that focusing investment efforts on physical infrastructure (electricity, irrigation and railways), and social infrastructure (human development) would raise the overall effectiveness of public investment and raise growth. However, factors such as law and order and health-related indicators such as infant mortality explain differences in growth, which are not taken into account by this paper. It is clear that urbanization and industrialization have a role to play in increasing per capita income. Ghate and Wright (2008) find that the ratio of Indian to United States per capita output over the past 45 years has displayed a distinctive V-shaped pattern.3 They show that a strikingly similar V-shaped pattern is visible not just in aggregate output figures, but also as the primary determinant of long-term movements in the cross-sectional distribution within the all-India total, at both sectoral and state output levels. They also carry out preliminary investigations of correlates of the “V-factor”, using a new panel data set for Indian states from 1960 to 2005 that extends and encompasses all previous data sets relevant to macroeconomic analysis of the Indian states. Ghate and Wright (2008) find that “V” states: •

Were on average more urbanized and more literate;4

•

Were somewhat more industrialized and somewhat less dependent on agriculture;

•

Spent somewhat less on development (revenue expenditure) than non-“V” states.

We find that Ghate and Wright (2008), like the others, focus on economic factors, such as infrastructure, but do not take into account the role of law and order or political factors in explaining growth. It is reasonable to expect that the law and order situation in a state would impinge upon private investment, economic growth 3

Their approach in using the United States as a benchmark may be debatable, but given that the United States is the head of the technological frontier and the standard neo-classical model would predict that growth rates converge to the country on the technology frontier, their choice is somewhat understandable. 4

Their “V” states are: Andhra Pradesh, Gujarat, Karnataka, Kerala, Maharashtra, Madhya Pradesh, Orissa, Rajasthan, and West Bengal. Their non-“V” states are: Assam, Bihar, Uttar Pradesh, Punjab, and Haryana. The latter two are included in the non-V states because they fit the convergence model (higher average income in the 1980s, lower growth in the 1990s).

8


Vol. 18, No. 1, June 2011

and the environment. States which are in turmoil with regard to law and order would not be viewed as good places to do business. While per capita income is only one measure of economic performance, there are studies that examine reduction in poverty. Agricultural productivity determines the extent of rural poverty. Datt and Ravallion (1998) study the causes of rural poverty in a developing rural economy and question as to why some Indian states have done better than others at reducing rural poverty. They model the evolution of various poverty measures using pooled state-level data for the period 1957-91. Differences in trend rates of rural poverty reduction are attributed to differing growth rates of farm yield per acre and differing initial conditions; states starting with better infrastructure and human resources saw significantly higher long-term rates of poverty reduction. Deviations from trend are attributed to inflation (which hurt the poor in the short term) and shocks to farm and non-farm output. This paper, while being quite insightful, unfortunately does not cover institutional factors, such as the existence of the minimum support price to farmers and their impact on reducing rural poverty. In addition to agricultural growth, productivity, initial income, private investment, infrastructure, urbanization and industrialization, we would expect sweeping changes in policy also to affect economic performance. Rodrik and Subramanian (2005) argue – in similar vein to Virmani (2006) – that the improvement in India’s economic performance was driven by policy changes. In particular, Rodrik and Subramanian argue that the trigger for India’s upward break in growth – which they pin down to around 1980 – occurred because of an “attitudinal shift” on the part of the national Government in 1980 in favour of businesses. While taking a cross-national focus, this is one of the few papers that take into account the importance of non-economic factors in growth, which needs to be noted. Similarly, Basu (2004) provides empirical evidence from a study of 16 major Indian states for the period from 1980 to 2001 that, under the economic reform process, better institutional mechanisms could actually help economies to grow faster with higher level of economic well-being. This paper estimates the economic well-being index (by aggregating 15 socio-economic variables, i.e., education, infrastructure, technological progress, income, and so on) and an index of good governance (by aggregating 13 variables indicating rule of law, government functioning, public services, press freedom, and the like) by multivariate statistical measures. Panel regression showed that governance measures, and economic policy variables are crucial to explain the differential level of development performance across states in India during the last two decades. It is worthy to note that this is one of the few papers to take into account the impact of governance and institutional factors on differential economic performance of the states.

9


Vol. 18, No. 1, June 2011

An important survey article on interregional disparities by Krishna (2004) focuses on issues of growth variability and volatility in Indian states. The coefficient of variation of year-to-year growth rates for a state is used as a measure of volatility. The four most volatile states in India were Orissa, Rajasthan, Gujarat and Uttar Pradesh while the three least volatile states were Punjab, Maharashtra and Kerala. However, volatility has been declining at the national level since the 1980s. The author notes that the dispersion of growth rates of states increased considerably in the post reform period (from 15 per cent in the 1980s to 27 per cent in the 1990s). Further analysis shows that agriculture has a positive impact on industrial and service sector growth. Also, social infrastructure is an important determinant of the investment decisions. The author however stresses that there is a need to explore other approaches to explain economic growth from all perspectives. Ashraf and others (2008) assess quantitatively the effect of exogenous health improvements on output per capita in general (not with specific reference to India). They find that the effects of health improvements on per capita income are substantially lower than those that are often quoted by policymakers, and may not emerge at all for three decades or more after the initial improvements in health are implemented. These results suggest that proponents of efforts to improve health in developing countries should rely on humanitarian rather than economic arguments. Ghosh (2006) evaluates the relative performance of 15 major Indian states regarding human development, and examines the relationship between this and economic growth. The estimates of cross-sectional growth regressions provide strong evidence of regional convergence in human development despite considerable divergence in real per capita income, indicating that the poor states that have failed to catch up with the rich ones in terms of per capita income have managed to catch up in terms of human development. The results suggest that the sequencing of policy should be such that the human development-induced growth process has to be strengthened so that states can transition from a vicious to a virtuous cycle category. Although the findings from this paper make sense, it focuses only on the relationship between human development and economic growth without taking into account other factors that impinge upon economic performance. Banerjee and Iyer (2005) analyze the colonial land revenue institutions set up by the British in India, and show that differences in historical property rights institutions lead to sustained differences in economic outcomes using district-wise growth rates. They find that areas in which proprietary rights in land were historically given to landlords have significantly lowered agricultural investments and productivity in the post-independence period than in areas where these rights were given to the cultivators. This is similar to the effects Besley and Burgess (2000) find for the impact of land reform on rural poverty.

10


Vol. 18, No. 1, June 2011

While the differential rate of growth among Indian states and the issue of convergence have been probed extensively, as is clear from the literature reviewed above, few studies have searched for an explanation for the difference between the northern and the southern Indian states, taking into account the role of both economic and non-economic factors. While past research had focused mostly on economic factors, we take into account both economic and non-economic factors, such as law and order, that impinge upon growth in the northern and southern Indian states. The next two sections briefly describe the methodology used in the paper to answer these questions, and focus on analysing economic trends as well as other explanatory indicators used in this study.

IV.

METHODOLOGY

The theory is that growth in emerging economies is driven by differences in human capabilities, skills and awareness, resources and their utilization, extent of urbanization, good governance (including law and order), and infrastructure. We believe these factors explain the disparities in investments, economic opportunities and other economic phenomena, such as poverty and per capita income. Below we describe trends in the indicators we have chosen for each of these factors and then discuss them. We attempted several regressions (see Sridhar and Reddy, 2009) but did not have long enough time-series data for informative results. Therefore, we do not report them here. Most of the data we examine as it relates to economic phenomena, investment opportunities, human capabilities and skills (educational and health indicators), infrastructure, urbanization, and resource utilization exists from the 1980s onwards (although some exist only decennially for the census years). Reasonable time-series data (going back to the 1960s) does not exist for all the indicators (with the exception of per capita net state domestic product (NSDP) and installed generating capacity (of electricity)). Hence, we first examine historical trends in each of the above indicators to study if some relationship exists between economic phenomena, and urbanization, governance, infrastructure and human capabilities. As discussed above, observing these phenomena over a period of time has the advantage of demonstrating whether such disparities across the southern and northern states are a recent phenomenon or whether they have existed for a long period of time.

11


V.

Vol. 18, No. 1, June 2011

TRENDS IN SOCIO-ECONOMIC INDICATORS

In this section, we present trends in various socio-economic indicators, such as domestic product, poverty, investment and economic opportunities more generally for the northern and southern Indian states separately. It is easy to believe that per capita income is determined by private investment, which creates jobs and income. Per capita income is one of the most fundamental economic phenomena which reflect economic living conditions of the population. This is one variable on which a reasonably long time series of data was available (unlike the poverty rate, data on which was available only for a few years). Hence, per capita NSDP was chosen as a measure of aggregate economic performance of the states, as is commonly done. Figure 1 summarizes the trend of the average weighted per capita NSDP of the southern states (Karnataka, Kerala, Andhra Pradesh and Tamil Nadu) and the four northern states (Bihar, Madhya Pradesh, Rajasthan and Uttar Pradesh).5 This figure shows that, on average, the per capita NSDP of the southern states (weighted with population) is on a much higher trajectory compared with that of the northern states. Figure 1. Trends in per capita NSDP, southern and northern states, 1960-06, 1999-00 prices 25 000 20 000 15 000 10 000 5 000 0 1960- 1963- 1966- 1969- 1972- 1975- 1978- 1981- 1984- 1987- 1990- 1993- 1996- 1999- 2002- 200561 64 67 70 73 76 79 82 85 88 91 94 97 2000 2003 2006 Average per capita (weighted) NSDP, South

Source:

5

Average per capita NSDP (weighted), north

EPWRF and Authors’ computations.

For purposes of reasonable comparison, the data for Bihar, Madhya Pradesh and Uttar Pradesh include data for Jharkhand, Chhattisgarh and Uttaranchal respectively from 1993-94 onwards. Although these three new states themselves came into existence only in 2000, the EPWRF had reconstructed the population and NSDP data series for the new states from 1993-94 onwards, based on the new districts forming these states.

12


Vol. 18, No. 1, June 2011

Further, the divergence between the two groups has been increasing since the early 1990s.6 This is a source of concern. We corroborate the disparities in per capita income with data on rural and urban poverty in the southern and northern states. While aggregate per capita income portrays the general economic conditions of the state, the prevalence of poverty indicates the extent of distress. The rural and urban poverty data paint a picture similar to that of the NSDP, showing greater prevalence of economic distress in the northern states. Figure 2 summarizes the disparities in rural poverty (weighted with the population of each respective state) between the southern and northern states. The rural poverty rate summarized in figure 2 refers to the proportion of rural population in the states living below the poverty line. Figure 2 shows that the extent of rural poverty is much greater in the northern than in the southern states, where it has been declining at a more rapid rate (since 1988) than in the northern states and the disparities are widening.7 The extent Figure 2. Trends in the rural poverty rate, southern and northern states 70 60 50 40 30 20 10 0 1973

1978

1983

1988

Average rural poverty, South

Source:

1993

1998

2003

Average rural poverty, North

Planning Commission, Government of India.

6

These disparities are based on per capita NSDP data when they are expressed in 1999-00 prices. When per capita NSDP is expressed in 1993-94 constant prices as well, one finds a similar trend with the divergence beginning in the early 1990s. 7

It may be noted that the poverty series in both figures 2 and 3 end in 2004-05.

13


Vol. 18, No. 1, June 2011

of rural poverty is directly determined by agricultural yield, agricultural wages and the availability of non-farm employment (see Besley and Burgess, 2000; Fan, and others, 2000). However, historically, the southern states have not always had this edge. For instance, Datt and Ravallion (1998) report nearly 70 per cent rural poverty each for Tamil Nadu and Kerala in 1960 and 65 per cent for Andhra Pradesh,8 compared with only a 48 per cent rural poverty rate each for Uttar Pradesh and Rajasthan in 1960. Madhya Pradesh and Bihar had rural poverty rates of about 55 and 65 per cent respectively in 1960, according to Datt and Ravallion. This suggests that poverty was much more acute in the case of the southern states earlier on, but they were able to reduce it rapidly at some point. Our objective is to understand when, how and why this took place. The disparities in urban poverty rates across the southern and northern states (when weighed with population) are much lower than with rural poverty (figure 3). Figure 3. Trends in the urban poverty rate, southern and northern states 60

50

40

30

20

10

0 1973

1978

1983

1988

Average urban poverty, South Source:

8

1993

1998

2003

Average urban poverty, North

Planning Commission, Government of India.

Karnataka was the only southern state according to Datt and Ravallion (1998) to have had a lower rural poverty rate of 52 per cent even as of 1960.

14


Vol. 18, No. 1, June 2011

In 1973, the urban poverty rates in the northern states were higher than in the south, but by 1993, the urban poverty rates in both the northern and southern states were the same. They started declining more rapidly in the south (figure 3). While rural poverty is closely related to productivity in the agricultural economy, urban poverty is related to the availability of urban employment, the prevalence of rural-urban migration and the level of urban wages. Summarizing the disparities in fundamental economic characteristics, the northern states are characterized by lower per capita income with greater rural and urban poverty than the south. The rural poverty rates across the two groups of states started to diverge in 1988, whereas disparities in urban poverty rates started increasing much later, in 1993. Finally, per capita NSDP started diverging in the early 1990s. Based on their relative performance in aggregate economic phenomena, it is plausible to believe that there are significant disparities between the southern and northern states in terms of their economic environment, opportunities and potential for investment. Paul and Sridhar (2009) examined which sector(s) led the surge in per capita NSDP which is observed in Tamil Nadu, a southern Indian state as compared with that in Uttar Pradesh, a northern Indian state. Figures 4 to 6 show the trend in the composition of NSDP by sector (respectively agriculture, industry and services) in these two states. In regard to the share of agriculture, Uttar Pradesh is always above Tamil Nadu. In the share of industry, Tamil Nadu scores well over Uttar Pradesh for all but the last couple of years. It did seem that during the last few years, the share of the industrial sector in Uttar Pradesh caught up with that in Tamil Nadu and surpassed it. Figure 4. Share of agriculture in NSDP, northern and southern states 70 60 50 40 30 20 10 6 -0

3 -0

05 20

0 20

02

-0

7 99 19

-9

96 19

93 19

-9

4

1 -9

8 90 19

87

-8

5

Share of Agriculture (weighted) South

19

19

84

-8

-8

2

9 81 19

-7

78 19

75 19

-7

6

3 -7

0

72

-7 69 19

19

-6

4

66 19

1

-6

-6

63

60

19

19

7

0

Share of Agriculture (weighted) North

15


Vol. 18, No. 1, June 2011

The service sector data is interesting. Until 1980-81, the two states were more or less identical as far as the service sector share was concerned. After 1981, however, the service sector in Uttar Pradesh declined in its share in NSDP when compared with that in Tamil Nadu, where there was a constant increase. So there are grounds to believe that the service sector led the surge in per capita incomes in Tamil Nadu. This is consistent with the national growth story. Figure 5. Share of industry in NSDP, northern and southern states 30 25 20 15 10 5

6 05 20

-0

-0

3

0 02 20

-9

99 19

96 19

-0

7

4

1

-9 93 19

-8

90 19

87 19

Share of Industry (weighted) South

-9

8

5 -8

2

84 19

-7

81 19

78 19

-8

9

6 -7

3 75 19

0

-7

-7

72

-6

69 19

66 19

19

7

4 -6 63 19

19

60

-6

1

0

Share of Industry (weighted) North

Figure 6. Share of services in NSDP, northern and southern states 60 50 40 30 20 10

Share of Service (weighted) South

16

20 02 -0 3 20 05 -0 6

19 87 -8 8 19 90 -9 1 19 93 -9 4 19 96 -9 7 19 99 -0 0

19 66 -6 7 19 69 -7 0 19 72 -7 3 19 75 -7 6 19 78 -7 9 19 81 -8 2 19 84 -8 5

19 60 -6 1 19 63 -6 4

0

Share of Service (weighted) North


Vol. 18, No. 1, June 2011

Disparities in investment and economic opportunities We measure disparities in economic opportunity across the northern and southern Indian states in terms of actual private investment flows (FDI inflows and domestic investments), which are recent phenomena (at least those on which only relatively recent data is available). 9 Actual investment flows are indicators of disparities in economic opportunities because they imply the creation of jobs, income and more broadly, economic growth. They also reflect underlying conditions, such as infrastructure and public services, which influence the location choice of firms (see Sridhar and Wan). The differences between the southern states and the northern states are very pronounced in terms of the amount of actual investments. Figure 7 summarizes this trend separately for the two groups of states (weighted with population). The south is significantly ahead of the northern states in attracting investment. The northern states are in a permanent low-level equilibrium as far as investment inflows are concerned, with inflows amounting to a meagre Rs 1.7 trillion when compared to the average of Rs 4.74 trillion for the southern states from 1995 to 2003.10 There is a need to explain these disparities in economic opportunities and lack of investor interest in the northern states. This is consistent with what Kurian (2000) reports for the states. If these trends continue in investment, the northern states could stand to lose substantially in terms of investment, jobs and income. Clearly, there are some fundamental differences between these two sets of states which need to be explained. Inter- and intraregional disparities within the country can lead to civil and social unrest. They also can lead to migration resulting in undesirable consequences. Hence, it is important to understand trends in these disparities and study what is causing them. A better understanding of the factors underlying regional disparities will throw better light on economic and investment opportunities available in the Indian states. As a result, researchers, state governments and investors stand to benefit from this research. 9

We had data from the Department of Industrial Policy and Promotion, Ministry of Commerce, Government of India, regarding the number and amount of FDI approvals and domestic investments approved. However, given that there usually is a significant difference between the FDI approvals, domestic investment approvals and actual investments, and the number of approvals does not translate in terms of actual investments, we use data on actual investments made by firms in various states, from CMIE’s CAPEX database. 10

The figure presents data for 1995 to 2003. While we had (unweighted) data on actual investment flows until 2009 for all the states we study, the data on population were not available beyond 2003, with the result that we could not report weighted averages beyond that year.

17


Vol. 18, No. 1, June 2011

Figure 7. Disparities in actual investment flows, north and southern states 700 000

600 000

500 000

400 000

300 000

200 000

100 000

0 1995

1997

1999

Weighted average, South Source:

2001

2003

Weighted average, North

Centre for Monitoring Indian Economy (CMIE) CAPEX Database and Authors’ computations.

The following section focuses on indicators for the variables which explain differences in the economic phenomena observed across the northern and southern states. First, we describe what indicators have been chosen for each of the explanatory factors, we then highlight the rationale for the expected effects of each of the variables on economic performance, following which trends in the explanatory indicators are described in a subsequent section. Indicators of Human Capabilities Some measures of human capabilities may be represented by education, and health care indicators. More precisely, these education and health indicators might be respectively: literacy rate; proportion of graduates; proportion of population enrolled in technical courses; and infant mortality rate. In addition, we examine the percentage of the population in working age groups, and on the supply side, the number of institutions of higher education in the states.

18


Vol. 18, No. 1, June 2011

The literacy rate can be expected to positively affect economic growth and per capita income in the states primarily because it is a measure of the knowledge and awareness of the population. Our assumption is that a higher literacy rate prepares one for higher skills, the ability to deal with higher technology, and the discretion to make rational choices. A more literate population is able to use its skills productively to generate more output and income. The proportion of graduates and the proportion of those enrolled in technical courses also positively impact the per capita income because of their effects on creating a labour pool with certain skills. The proportion of graduates reflects the percentage of population that has attained a certain threshold level of education which equips them with certain skills needed for specific kinds of economic activity. Hence, an increase in the proportion of graduates can be expected to increase the workforce participation rate of the population and enable them to contribute to increased output and income. The proportion of technical manpower can bring about growth and has the potential to increase incomes since investors are usually attracted to a pre-existing pool of manpower with certain skills. We may expect the percentage of population in working age groups to affect per capita income positively through their impact on output because only population in the working age groups is likely to contribute to output increases.11 Turning to health indicators, the infant mortality rate (IMR) across a state is an indicator of its progress on health. While this factor does not directly affect investment in the state, it can nevertheless be viewed as an indicator which reflects the economic capability of the workforce. Good health enhances the productivity of the population. The IMR indicates prenatal care, maternal care and the existence of child-care facilities, indicating maternal mortality, fertility rate and the death rate of the population. It indicates the stage of demographic transition the state is in. Our assumption and expectation is that a lower IMR of population implies that the state’s population is healthy. Empirical studies have repeatedly brought out the finding that hospitalization is one of the most important reasons for indebtedness and abject poverty, especially in rural areas (see George, 2009). Hence, we assume that states with a lower IMR are healthier. A healthy population is capable of producing more output and income.

11

The working age group is defined as the population in the age groups 15-59 years. Only for 1971 was data on population in the upper age limit not available; hence, we used population greater than 15 years as those being in the working age group. As this was the same problem for all the states, however, we do not expect their relative positions to change.

19


Vol. 18, No. 1, June 2011

Indicators of Governance: Law and order Governance refers to the functioning of governments and public institutions that impact on economic activities and the lives of citizens (Paul and Sridhar, 2009). When the processes of public decision-making and implementation of policies are carried out with credibility, transparency and accountability, governance is considered good. Given its complex nature and scope, however, it is far more difficult to define and measure governance than the other factors discussed above. In this paper, governance has been equated with law and order. This was done because of the absence of data on other variables which reflect governance. As Paul and Sridhar (2009) point out, it is extremely difficult and challenging to come up with measures of governance that reflect the functioning of public institutions. A sound law and order system is essential for economic and social progress. Based on open-ended discussions with senior police officials, we came up with two measures of governance or law and order: (a) police firing incidents per million population and (b) percentage of pending cases under trial in courts. We selected police firings per million population as an indicator of the law and order condition in a state because it signals the intensity of agitation between groups, and the ability or inability of the administration to bring them under control or a combination of both. Because police firings are widely and regularly reported, they can add to uncertainty in the minds of investors and can adversely impact the smooth functioning of a society and its economic activity. Law and order may also be represented by events which capture the efficiency and effectiveness of the judiciary, such as the proportion of pending cases in the court. 12 One common measure that is chosen to reflect governance is corruption – the use of monetary or non-monetary bribes to have work done in government or public institutions. There are no reliable subnational data on these measures. However, the fact that other measures of governance have been used does not imply that corruption and other measures of governance have been ignored, as Paul and Sridhar (2009) point out. Good governance does manifest itself in law and order. For example, when law and order break down (as reflected in rising number of incidents of police firing), the public may be forced to resort to corruption. Similarly, when the public image of a place is that it is disorderly or when court cases take a long time to resolve (pending judicial cases), entrepreneurs will refrain from 12

In order to arrive at these indicators, we had open-ended discussions with the Director-General of Police in Karnataka on the role that law and order play and how it impinges on the economic environment and economic growth in the state. He suggested that the number of incidents of police firing and the percentage of civil to armed police are good indicators which capture the public agitation mood in the state which impinges on their economic and investment environment.

20


Vol. 18, No. 1, June 2011

investing in that state since they look for a stable environment and speedy redress of grievances in the event of disputes. Sound law and order is also essential for the retention of a skilled workforce. Hence, the measures we chose are reflective of public functioning and the governance of a state. Specifically, we expect that the greater the police firing incidents per million population, the lower the per capita income would be. Similarly, the greater the percentage of cases pending trial, the lower would be the per capita income, for reasons discussed above. We also examined SLL (cases reported under local and special laws) crimes. These refer to crimes committed under the Arms Act, Opium Act, Gambling Act, Excise Act, Prohibition Act, Explosives and Explosive Substances Act, Motor Vehicles Act, Prevention of Corruption Act, Customs Act, Indian Railways Act, and other offences. We have already demonstrated that corruption is reflected in the law and order situation. Similar is the case with crimes committed under the Explosives and Explosive Substances Act, which is likely to be reflected in police firing incidents. Most of the other SLL crimes noted above are private crimes and do not reflect the general law and order condition of the state. Furthermore, the reporting of many of these crimes is determined by the filing of a First Information Report (FIR). If no FIR is filed, then these crimes are not reflected in the data. However, since data on police firings and pending court cases are reported widely, we chose them to reflect governance and public functioning. Thus the measures of law and order we choose reflect to a substantial degree the governance of a state. We consider our work pioneering in that we find that no other earlier studies (with a few exceptions as we noted) have examined non-economic factors, such as law and order and their impacts on the economic environment. Measures of infrastructure Why is infrastructure important for economic growth and investment? Infrastructure is an important enabler of economic growth. Electricity is required for manufacturing; telecommunications are necessary for reducing firms’ transaction costs; good roads are required for transportation of inputs and connectivity to markets. Mani and others (1996) find that power availability rather than its price, reliable infrastructure and factors of production played significant roles in decisions regarding the location of firms across major Indian states. In line with this literature, our chosen measures of infrastructure or public services are (a) installed capacity for generating electricity and (b) penetration of telecommunications.

21


Vol. 18, No. 1, June 2011

First, we take the instance of electricity – installed capacity in the states. While electricity consumption is concomitant with growth and may be expected to increase monotonically with growth, installed capacity is a precondition for growth. Installed capacity is critical for manufacturing processes and is necessary to increase output and raising per capita incomes. Another measure of physical infrastructure we examine is the percentage of households with electricity. This indicates the extent to which electricity is extensively available in the state. However, given that the percentage of households with electricity could be correlated with the installed generating capacity in a state, we include only installed capacity in the estimation, although we present trends and disparities in the percentage of households with electricity connections across the northern and southern states. Telecommunications are crucial for firms in reducing their transaction costs (see Norton, 1992; Roller and Waverman, 2001) and for increasing organizational efficiencies, output and per capita incomes. The literature conclusively shows that tele-density has positive impacts on growth. A number of researchers have hypothesized that telecommunication infrastructure lowers both the fixed costs of acquiring information and the variable costs of participating in markets (Norton, 1992). They point out that, as such infrastructure improves, transaction costs decline, and output increases for firms in various sectors of the economy. Sridhar (2007) found positive impacts of mobile and landline phones on national output when controlled for the effects of capital and labour. Hence, we expect both installed capacity and tele-density to have positive effects on the economic environment in the states, especially for manufacturing, and positively impact per capita NSDP. Indicators of resources We choose per capita public (both capital and revenue) expenditure as an indicator of resources available to a state which could endogenously determine its per capita NSDP. This is because it is assumed that all public expenditure translates into output of goods and services, increasing per capita NSDP. This could be endogenous since rising public expenditure could be partly financed out of rising NSDP. However, we circumvent endogeneity by using the lagged form of this variable. While current year expenditure can be endogenous, per capita income in a current year cannot impact a previous year’s expenditure.

22


Vol. 18, No. 1, June 2011

Measures of urbanization We use the percentage of urban population in a state as the measure of urbanization which we expect will impact per capita income. Urbanization is a causal factor underlying high per capita incomes because scale economies and agglomeration economies make it possible to accumulate output rapidly. How is urbanization defined in India’s context? The Census of India defines settlements having the following characteristics as urban areas: (a)

A population of five thousand or more;

(b)

A minimum density of 1,000 people per square mile;

(c)

At least 75 per cent of workforce outside agriculture.

It should be mentioned that India’s definition of urbanization is quite conservative when compared with that of China where all areas with a minimum of 10 per cent non-agricultural employment are classified as urban. As Cohen (2004) argues, if India were to reclassify its urban areas using a more liberal definition, a majority of India would be urban today. In fact, higher levels of urbanization also attract firms to locate, invest and create jobs due to urbanization economies and localization economies. Trends in explanatory factors: Human capabilities In this section, we review the trends and disparities in the explanatory factors. Figure 8 compares the average weighted (with population) literacy rate across the southern and northern states over time. Figure 8 shows that the southern states’ literacy rate has always been at a higher level when compared with that of their northern counterparts. Further, the rate of growth of literacy also has been occurring at the same rate in the two regions, with the result that the northern states’ literacy rate has remained well below that of the south as of 2001. Despite its remarkable stability when compared with per capita NSDP (which is quite volatile, see figure 1), we surmise that the literacy rate may have been one of the preconditions necessary for economic growth to have taken off in the southern states. In terms of examining trends in educational outcomes, we do not stop at the literacy rate. We compare the trends in the average proportion of graduates (or above) from 1971 to 2001 between the southern and northern states as yet another educational measure.

23


Vol. 18, No. 1, June 2011

Figure 8. Trends in literacy rate, southern and northern Indian states 80

70

60

50

40

30

20

10

0 1961

1971

1981

Average weighted literacy rate, South

Source:

1991

2001

Average weighted literacy rate, North

Census of India and authors’ computations.

Figure 9 compares the trend in the average weighted proportion of graduates during 1971 to 2001 for the southern and northern states separately. The interesting finding here is that the northern states had on average the same proportion of graduates as the southern states, from 1971 to 1981. However, they gradually lost out to the southern states, from 1991 to 2001 (figure 9). Thus, we have more evidence here that the surge in the south is a more recent phenomenon, not historical. Over and above general graduates, we made an attempt to examine the proportion of technical manpower in the two groups of states. Enrolment in and graduation with degrees in technical courses, such as B.E. (Bachelor’s of Engineering), B.Sc. (Bachelor’s in Science), B.Arch. (Bachelor’s in Architecture), Medicine, Dentistry, Nursing, Pharmacy, Ayurvedic and Unani, B.Ed. (Bachelor’s in Education), and B.T. (Bachelor’s in Technology) have a role in the building of a skilled labour force. So we compared enrolment by year in (all the above) technical degree courses as a proportion of the population in the relevant age group (above 15 years)

24


Vol. 18, No. 1, June 2011

Figure 9. Proportion of graduates, south and northern states, 1971-2001 Per cent 6

5

4

3

2

1

0 1971

1976

1981

1986

Average weighted % graduates, South

Source:

1991

1996

2001

Average weighted % graduates, North


to examine if the southern states have anything more of an edge compared with their northern counterparts (see figure 10).13 Certainly, the southern states have a larger skilled and technical labour force when compared with the northern states for all the years of study. All these measures of human capabilities and skills can be expected to impact not only the per capita income through their effect on skilled jobs, but also impact investment due to the existence of a pool of skills which impact firm location

13

The methodology we used to arrive at the enrolment in these technical courses as a proportion of population above 15 years of age, from 2001 to 2005 is as follows: We took the age-wise distribution of population in 2001 from the 2001 Census for all the states. Then we assumed that the same age-wise distribution of population holds good for the period from 2002 to 2005. Since we had the state-wise populations during the period from 2002 to 2005, we applied the age distribution (of population above 15 years of age) of 2001 to obtain population above 15 for the non-Census year. Then we took the enrolment in the technical courses as a proportion of the population above 15 years of age. Next, we averaged this proportion for the southern states and the northern states (including Jharkhand, Uttaranchal and Chhattisgarh) separately. Although we tried, we were unable to obtain data for earlier years on this important indicator.

25


Vol. 18, No. 1, June 2011

Figure 10. Enrolment in technical courses, south and northern states Per cent 0.35

0.30

0.25

0.20

0.15

0.10

0.05

0.00 2001

2002 Average proportion of technical enrolment (weighted), South

Source:

2003

2004

Average proportion of technical enrolment (weighted), North

Ministry of Human Resource Development, Government of India, and authors’ computations.

decisions. In fact, it is plausible to believe that technology giants such as Infosys and WIPRO have located in Bangalore only because of the pre-existence of a large pool of skilled and technical workforce there. Paul and Sridhar (2009) report that the southern Indian state of Tamil Nadu had a total of over 540 engineering colleges in 2008 compared to only 11 colleges in the 1970s. They report that Uttar Pradesh, on the other hand, had less than half this number of engineering colleges despite having a head start in this arena in the nineteenth century. On the labour market aspects, we examined the percentage of the population in the working age groups. Figure 11 presents the disparity across the northern and southern states in terms of their working age group population. Figure 11 shows that, although the northern and southern states were the same as far as the percentage of population in working age group is concerned in 1961, there was a divergence after 1971, when there was a steady increase in the working age population in the southern states when compared with that of the north.

26


Vol. 18, No. 1, June 2011

Figure 11. Population in working age groups, south and northern states Per cent 60

50

40

30

20

10

0 1961

1971 Weighted average, South

Source:

1981

1991

2001


Census of India, various years, and authors’ computations.

This lends credence to the belief that this may have been a contributing factor to the rising incomes we observe in the south.14 Finally, on the supply side of human resources, we examined the number of higher educational institutions in the northern and southern states. To obtain this information, we aggregated data on the number of colleges from the Census town directories. 15 The assumption is that only towns and cities contain institutions of higher education, which is reasonable. There is also no source which would contain this data for rural areas as well. Figure 12 presents these data for the northern and southern states. Figure 12 shows that as with the other measures of human resources we observe, even with respect to higher educational institutions, the southern states have stolen a march over their northern counterparts. This must have created the

14

We also obtained data on the man-days lost by state, but there was no reliable data on the number of man-days of employees or workers against which we could compare the man-days lost. Hence, this variable could not be used. 15

Information on arts, science, commerce, engineering, medical and law colleges, vocational training institutes and polytechnics is included here.

27


Vol. 18, No. 1, June 2011

Figure 12. Higher educational institutions, south and northern states 3 000

2 500

2 000

1 500

1 000

500

0 1991

1993

1995

Weighted average, South Source:

1997

1999

2001


Census of India town directories and authors’ computations.

necessary institutional capacity to turn out a large pool of skilled labour responsible for increasing levels of output and income. Thus, we find that in terms of educational outcomes measured in the literacy rate, proportion of graduates, enrolment in technical courses, proportion of population in the working age group and supply-side factors, such as the number of higher educational institutions, the southern states have an edge over the northern states during the entire period of our study.16 Figure 13 summarizes the selected health indicator of human capabilities – the infant mortality rate (IMR), weighted by population of the respective states. Figure 13, which summarizes the historical trend in the infant mortality rate across the southern and northern states, shows that the southern states (with their lower IMR) have always been better than their northern counterparts, consistent with our expectation. This implies better prenatal medical care and related facilities in the south, which implies a population with a much better health and productivity than in the north, although the IMR in the northern states has been declining post-1991. 16

A caveat to note is that the mere existence of a large number of educational institutions in the southern region does not mean that enrolments consists only of people from the southern states or that the graduates of these institutions necessarily constitute the labour force for the industries in the south.

28


Vol. 18, No. 1, June 2011

Figure 13. Infant mortality rate (IMR), southern and northern states 200 180 160 140 120 100 80 60 40 20

IMR, South

Source:

97 19

93

95 19

19

19

91

89 19

87 19

85 19

83 19

81 19

79

77

19

19

75 19

73 19

19

71

0

IMR, North

Compendium of India’s Fertility and Mortality Indicators, 1971-97, Registrar-General of India and Authors’ Computations.

These trends in education and health, which are indicators of human capabilities and skills, strongly suggest that at least some of the southern states have had an advantage historically over their northern counterparts. Trends and disparities in governance: northern and southern states We have already discussed the indicators of governance – police firing incidents (per million population) and proportion of pending court cases under trial. An examination of trends in police firing incidents and the proportion of pending cases in court (filed under the Indian Penal Code) is very revealing when we look at these separately for the southern and northern states. Figure 14 summarizes the trend in the average number of police firing incidents (per million population) separately for southern and northern states, weighted by their population.17 17

For the northern states, since we are comparing the undivided states (Madhya Pradesh, Bihar and Uttar Pradesh) prior to 2000 with years beyond 2000 after the three new states were created, we have added the data for Chhattisgarh, Jharkhand and Uttaranchal post-2000 to ensure that we are comparing the same set of states.

29


Vol. 18, No. 1, June 2011

Figure 14. Trends in the average number of police firing incidents per million population, south and northern states 4.00 3.50 3.00 2.50 2.00 1.50 1.00 0.50

Average police firing/million population (weighted), South Source:

02 20

00 20

98 19

96 19

94 19

92 19

19

88

0.00

Average police firing/million population (weighted), North

National Crime Record Bureau and authors’ computations.

During the 1990s, the south Indian states had relatively more incidences of police firing per million population when compared with that in the northern states. This is a surprising finding. After probing this, we found that such incidents in the south are dominated by Andhra Pradesh, which was characterized by frequent Naxalite disturbances (1987-2002) during which there was a sharp increase in the number of police firings. It should be noted that Andhra Pradesh, which is high on this score (law and order problems), is lowest on the per capita income front among the southern states (implied in chart 1, see Paul and Sridhar, 2009). By and large, if Andhra Pradesh were excluded, the number of police firing incidents in the southern states would always be lower than in the north. Paul and Sridhar (2009) also find evidence of this. Next, we take the proportion of pending cases in courts in the two sets of states and examine their trends during 1991 to 2004. This is calculated as cases filed under the Indian Penal Code pending trial in the courts as a proportion of the total number of cases for trial including pending cases from previous years. Court cases should be viewed as a measure of public faith in the judiciary, and pending cases demonstrate its efficiency/inefficiency. Figure 15, which compares the weighted (weighted by the population of the respective states) proportion of cases pending trial in courts, shows that the judiciary

30


Vol. 18, No. 1, June 2011

Figure 15. Trends in proportion of cases pending trial in courts of northern and south Indian states Per cent 100 90 80 70 60 50 40 30 20 10 0 1991

1993

1995

1997

Average percentage of court cases pending (weighted), South

Source:

1999

2001

2003

Average percentag of court cases pending (weighted), North

National Crime Record Bureau and authors’ computations.

in the northern Indian states is quite inefficient when compared with that in the southern Indian states, where on average the proportion of cases pending trial stood at only 67 per cent as of 2004, compared with 76 per cent in the northern states (including the three new states – Jharkhand, Chhattisgarh and Uttaranchal). It may be argued that cases may be pending because the number of cases registered is higher or the number of judges lower. If the number of judges is lower, it means the state is unable to recruit judges to increase its efficiency. This is yet another indicator which impacts the economic environment in these states and is very representative, since investors also look for speedy redress of grievances in the event of disputes. We investigated the possibility of using other law and order measures such as the proportion of civil to total police force (consisting of civil and armed police). But since that is correlated with the number of police firing incidents, we decided to use the police firing incidents per 100,000 population. For instance, only when the number of police firing incidents is on the increase that we may expect armed police strength in a state to increase. Hence, we expect that the number police firing incidents coupled with the number of court cases – both ongoing and pending provide reasonably good measures of law and order.

31


Vol. 18, No. 1, June 2011

Summarizing, with the average weighted number of police firings (leaving out Andhra Pradesh which is the outlier) being lower in the south and with their high judicial efficiency (low proportion of cases pending trial), there is reason for us to believe that the southern states would offer a peaceful and stable environment, resulting in better economic and investment opportunities compared with their northern counterparts. Trends in indicators of infrastructure Our findings with respect to the infrastructure indicators – installed generating capacity, and tele-density are interesting. We find that the southern states are ahead of the northern states in these respects. Figure 16 summarizes weighted (weighted by the respective states’ population) installed capacity (thousands of kilowatts) per million population in the southern and northern states separately. These data series cover a reasonably long period of time. Not only was the installed capacity per million population always Figure 16. Trends in installed capacity of electricity per million population: southern and northern states 100 90 80 70 60 50 40 30 20 10 0 1960

1970

1980

Average IC/mill pop (weighted), South Source:

32

1990

2000

Average IC/mill pop (weighted), North

Central Electricity Authority, Ministry of Power, Government of India, and authors’ computations.


Vol. 18, No. 1, June 2011

lower in the northern than in the southern states, but also there was a widening of these disparities between them from the mid-1980s. 18 The southern states experienced a continuous increase in their installed capacity after the 1990 liberalization whereas the northern states likely faced several constraints with regard to installed capacity expansion. Because of this, their average weighted installed capacity stagnated beginning from the mid-1980s. Thus, it is possible that a number of preconditions necessary for the existence of industry and services were getting ready in the southern states, which prepared them to take the plunge when the reforms of 1991 took place. Another measure of physical infrastructure we examine is the percentage of households with electricity. This indicates the extent to which electricity is extensively available in the state. Figure 17 summarizes the weighted proportion of households in Figure 17. Percentage of households with electricity, northern and southern states Per cent 80 70 60 50 40 30 20 10 0 1981

1986

1991

Weighted average, South

Source:

1996

2001


Census of India, various years, and authors’ computations.

18

Here, as with other indicators, post-2000, we added the installed capacity for Chhattisgarh, Jharkhand and Uttaranchal to that for Madhya Pradesh, Bihar and Uttar Pradesh respectively, in all fairness to the northern states, since Chhattisgarh especially received many power plants after its separation from Madhya Pradesh. We have ensured, based on our discussions with the Central Electricity Authority, that it is possible for installed capacity to decline when old plants are retired or when plants are degraded.

33


Vol. 18, No. 1, June 2011

the two groups of states and presents the trends over time. This figure shows that the southern states clearly have a lead in the percentage of households with electricity for all the years of our study. This implies that the southern states’ physical infrastructure was much better when compared with that of the northern states. This prepared the southern states to grow rapidly when the liberalization of 1991 took place. The northern, states with their low level equilibrium with regard to the electricity infrastructure, were not prepared, and therefore lagged behind even when the reforms of 1991 took place. Another measure of infrastructure we looked at relates to tele-density – the number of fixed land lines and mobile phones per 100 population for the southern and northern states. Figure 18 presents weighted tele-density (weighted with the states’ population) for the two sets of states. As with electricity, tele-density for the southern states on average is not only much higher than that for the northern states for the limited period (1999-2004) over which we observe this, but it also increased at a much higher rate in the south than in the northern states over this period. Figure 18. Trends in tele-density: southern and northern states 14

12

10

8

6

4

2

0 1999

2000

2001

2002

2003

Average (weighted) teledensity, South Average (weighted) teledensity, North

Source:

34

Department of Telecommunications, Government of India.

2004


Vol. 18, No. 1, June 2011

Recall that we have defined tele-density to consist both of land lines and mobile telephones. Given land lines are mostly offered by government operators (such as Bharat Sanchar Nigam Limited, Mahanagar Telephone Nigam Limited and so on), there is not much of a difference in penetration between the southern and northern states there.19 The differences in total telephone penetration across the northern and southern states could be attributed to the extent of mobile telephone penetration, which is much higher in the southern states. This is primarily due to the competition prevalent in the mobile telephony sector (see Sridhar, 2007) which is much greater in the southern than in the northern states. We made an attempt to examine the road length in the northern and southern states. We did obtain data on this from the CMIE, but found that the road length declined during some years in most of the states. This suggests that there were changes in road classification which were not captured by this database and hence are not reported here. In summary, all infrastructure indicators including installed capacity (electricity), percentage of households with electricity and tele-density, show clear advantage for the southern states and steep disparities between the northern and the southern states over a reasonably long period of time. This disparity continues to widen even today. This strongly suggests that the southern states had all the preconditions necessary for growth and were ready to take the plunge when the reforms of 1991 took place. However, the northern states, with their poor infrastructure and pre-conditions, were simply not ready to take advantage of the opportunities when economic liberalization started to take place in the country. Trends in resource utilization The efficiency with which resources are utilized has impact on economic growth. If resources are used in a manner that maximizes the useful goods and services derived from those resources, then we may expect greater economic growth to occur. The “doing more with less” slogan indicates the focus on more outputs with fewer impacts (fewer resources). While we focus on outputs with fewer resources, we are unable to examine other resource utilization impacts, such as that on the condition of the poor (relating to equity), due to data limitations.

19

The three new states created in 2000 (namely Chhattisgarh, Jharkhand and Uttaranchal) are still treated as part of the circles of their parent states, with the result that pre-2000 and post-2000, we are comparing the same states.

35


Vol. 18, No. 1, June 2011

Some measures of resources and their utilization would be public expenditures. We examined the trend in total expenditures20 (consisting of both developmental and revenue expenditure) on various social sectors (such as education and public health) which are inputs. Figure 19 summarizes the average total (developmental and nondevelopmental) per capita expenditure of the southern states (weighted with population), and that of the northern counterparts. The record of the southern and northern states in terms of spending (summarized in figure 19) indicates that the southern states took a leap forward in their developmental expenditures post-1991 compared with their northern counterparts. However, the fact that the southern states did not always have this advantage may be seen by the fact that, in the 1980s, the northern states’ per capita expenditure was more or less the same as that of the southern states. Ghate and Wright (2008) found that revenue expenditure by the V states was lower than by their non-V states. Figure 19. Total per capita expenditure, south and northern states 5 000 4 000 3 000 2 000 1 000 0 1980-81

1983-84

1986-87

1989-90

1992-93

1995-96

1998-99

2001-02

Weighted per capita expenditure for Southern states Weighted per capita expenditure for northern states

Source:

Economic and Political Weekly Research Foundation (EPWRF) and authors’ computations.

Next, we review the sectoral expenditure for the southern and northern states. When we examine such sectors as education and health, it is important to take into account total expenditures rather than merely capital expenditures. Much of the education and health outcomes depend upon the number of teachers and health workers and in respect of these items current/revenue expenditures constitute more than 80 per cent of the total expenditure. Thus, we compared total expenditures on 20

36

This includes developmental expenditure incurred on the capital and the revenue accounts.


Vol. 18, No. 1, June 2011

sectors such as education and health with respective outcomes, such as the literacy rate, proportion of graduates, enrolment in technical courses, and infant mortality rate. Figure 20 summarizes over time the average per capita (total) expenditure on education, sports and culture by the southern and northern states.21 While we did not have this data disaggregated separately for education, sports and culture, we surmise that the expenditure on education must account for a major part of this expenditure. Having noted this, figure 20 shows that the southern states spent less on education than the northern states during the 1980s. It was only after 1990 that the southern states’ spending on education, sports and culture started diverging from that of the northern states.22 It is not clear if the increased spending is a sign of inefficiency or indicates better outcomes. To assess this, we compared this expenditure on education to relevant outcomes in the southern and northern states. The foremost of educational outcomes is the literacy rate, which we have compared for the southern and northern Figure 20. Trends in total per capita expenditure on education, sports and culture, south and northern states

19

80 19 -81 81 19 -82 82 19 -83 83 19 -84 84 19 -85 85 19 -86 86 19 -87 87 19 -88 88 19 -89 89 19 -90 90 19 -91 91 19 -92 92 19 -93 93 19 -94 94 19 -95 95 19 -96 96 19 -97 97 19 -98 98 19 -99 99 20 -00 00 20 -01 01 20 -02 02 20 -03 03 -0 4

800 700 600 500 400 300 200 100 0

PC total expenditure on education, sports and culture (weighted) Southern states PC total expenditure on education, sports and culture (weighted) Northern states

Source:


21

In the case of education and health, the total expenditure was developmental expenditure on the capital and revenue account. There was no non-developmental expenditure reported. 22

It must be mentioned that, as with the other indicators, for the northern set of states, we have included post-2000 data for the three newly created states – Jharkhand, Chhattisgarh and Uttaranchal, – so that the pre-2000 and post-2000 data are comparable.

37


Vol. 18, No. 1, June 2011

states in figure 5. That figure clearly shows that the south has been well ahead of the north historically in terms of the level and progress of the literacy rate. This means that the per capita expenditures on education are either not completely reflected in the literacy rate, or the southern states are more efficient (recall from figure 20 that their spending on this sector during the 1980s had been lower than that of the northern states) when compared with their northern counterparts as far as the outcomes are concerned. We have also compared the trends in the average proportion of graduates (or above) during 1971 to 2001 between the southern and northern states as yet another educational outcome (see figure 9). We found that the surge of the south is a more recent phenomenon, rather than historical (recall that initially the southern and northern states had the same proportion of graduates until 1981). Even when we compared enrolment in technical courses, we found that the southern states have a higher proportion of technical labour when compared with the northern states for all the years of study (figure 10). Thus, educational outcomes measured in terms of the literacy rate, proportion of graduates (post-1981), and the enrolment in technical courses, show that the southern states have maintained an edge over the northern states. This is so despite the fact that their spending on education has not been always higher than that of the northern states (figure 20, see the decade of the 1980s). Thus it must be the case that the southern states’ expenditures on education are efficiently spent compared with those of the northern states. Next, we examined per capita expenditures by the states on public health and medical facilities.23 Figure 21 summarizes the trends in per capita spending on public health and medical facilities in the south and northern states. Figure 21 shows considerable variability in the per capita expenditures on public health and medical facilities across the two set of states, with the spending of the southern states diverging from that of the northern states beginning from the late 1980s. We have already reviewed the outcomes of health spending – manifested in the infant mortality rate, which is lower for the southern states (see figure 13). Given that the southern states’ spending on public health was clearly lower than that of the northern states during the decade of the 1980s, but its health outcomes, such as IMR, were clearly better always, it must be the case that the quality of spending in the south is much better than in the northern states even as it relates to public health. 23

As with the other indicators, for the northern set of states, we have included post-2000 data for the three newly created states – Jharkhand, Chhattisgarh and Uttaranchal, – so that the pre-2000 and post-2000 data are comparable.

38


Vol. 18, No. 1, June 2011

Figure 21. Per capita expenditure on public health, south and northern states 200 150 100 50

19

80 19 -81 81 19 -82 82 19 -83 83 19 -8 84 4 19 -85 85 19 -86 86 19 -87 87 19 -8 88 8 19 -89 89 19 -90 90 19 -91 91 19 -9 92 2 19 -93 93 19 -94 94 19 -95 95 19 -9 96 6 19 -97 97 19 -98 98 19 -99 99 20 -0 00 0 20 -01 01 20 -02 02 20 -03 03 -0 4

0

PC total expenditure on medical and public health (weighted) Southern states PC total expenditure on medical and public health (weighted) Northern states

Source:


Summarizing, when we compare spending on education and health with their outcomes across the two groups of states, we find the south is relatively more efficient as it is able to ensure better outcomes than the northern states with its lower record of spending on these sectors during the 1980s. In order to compare the expenditure on roads and bridges by the southern and northern states, we did not have reliable data on significant outcomes (road length) to enable us to make an assessment of this component of public spending.24 We also had data related to spending on energy, which could have been easily compared with the outcome on installed capacity generated, but the data on energy were not complete.25

24

We found in the case of some states that road length actually declined in some years, which is not plausible except in the event of a reclassification of roads. Further, we found that the data on road length from the Centre for Monitoring Indian Economy (CMIE) was disaggregated by various types of roads, such as surfaced national highways, surfaced state highways, district roads, panchayat roads, urban roads, project roads and so forth. But the length of different types of roads did not add up to the total road length reported. We attempted to contact CMIE regarding this, but did not obtain a satisfactory response. 25

We found negative values in the developmental expenditure on energy in the case of both the southern and northern states. Based on our discussions with the EPWRF, the actual developmental expenditure on energy is worked as follows: if the actual expenditure is Rs 100,000,000 during any given year and the receipts were Rs 116,000,000, then the -16,000,000 appears as deficit. The problem with using these data is that they do not indicate what was spent, but only the deficit.

39


Vol. 18, No. 1, June 2011

While in this section, we have compared the public expenditure on social sectors with their outcomes to assess the efficiency of spending, in the regression, we use total public expenditure per capita (lagged) as an explanatory variable. Disparities in Urbanization Given the importance of urbanization in increasing aggregate productivity and incomes, what do we observe with respect to the urbanization pattern of the southern states versus the northern states? Figure 22 shows that the southern states are way above the northern states in terms of the percentage of urbanization (weighted with population) since 1971. This indicates that the southern states experienced agglomeration and scale economies in production throughout the period, which must have contributed to increased aggregate output and their higher per capita incomes. Figure 22. Trends in urbanization, south and northern states, 1971-2001 40 35 30 25 20 15 10 5 0 1971

1976

1981

1986

Average urbanization (weighted), South

Source:

40

1991

1996

2001

Average urbanization (weighted), North



VI.

Vol. 18, No. 1, June 2011

SUMMARY OF FINDINGS AND IMPLICATIONS

The important question examined by this paper is: Why have some Indian states grown faster than others? While our hypothesis was that, with the relative growth of manufacturing, the southern states may be growing faster than the northern Indian states, which have more comparative advantage in agriculture, we find that the service sector has led the surge in growth in the southern states (taking the example of Tamil Nadu). The additional purpose of the exposition in the previous sections of this paper has been to examine whether factors, such as human capabilities, skills and awareness, infrastructure, governance, urbanization and resource utilization shed light on the divergent paths of per capita income growth, divergent trends in poverty reduction, and disparities in FDI inflows and domestic investment observed across the southern and northern states. What do we gather from the analysis of disparities in these factors between the southern and northern states? It is possible that differences in the underlying and relatively more stable conditions such as literacy rate, and infant mortality rate in the two set of states could at least in part account for the divergence in per capita income and poverty reduction although there could be some simultaneity there. Our premise is that the marked upward shift in per capita income and the subsequent reduction in poverty that the southern states experienced since the early 1990s can be attributed to the flow of substantial investments into these states. We find that, from 1995 to 2003, the southern states attracted private investment worth Rs 4.74 trillion when compared with an average of only Rs 1.7 trillion during the same period for the northern states. Based on our research and analysis, we surmise that disparities in governance, educational outcomes, urbanization, infrastructure and resource utilization could account for disparities in investment flows across the southern and northern states. A limitation is that this paper provides more descriptive and qualitative analysis than quantitative analysis. Therefore, policy conclusions need to be read with caution. Being aware that we may not have taken into account some factors that could have contributed to the outcomes studied here (for instance, Paul and Sridhar (2009) discuss the impact of social movements and caste on education in the case of Tamil Nadu), the following specific findings from our analysis are worth noting: (a)

With respect to most of the factors representing human capabilities – literacy, infant mortality, stock of graduates, enrolment in technical courses, and proportion of population in working age group, supplyside factors, such as the number of institutions of higher education, and infrastructure, such as installed capacity, percentage of households with electricity and telephone penetration, the southern

41


Vol. 18, No. 1, June 2011

states had certainly an advantage over the northern states. We have to note that technical manpower (indicated by enrolment in technical courses), in which the southern states appear to have a lead, signals a critical resource that modern industries and the service sector need. Unfortunately, data on this was not available for a reasonably long time period for us to include it in the regression. Given this limitation, we surmise that the supply of this factor must have played a key role in the transformation that the south experienced from the mid-1990s; (b)

In terms of factors indicating law and order, such as the proportion of cases pending trial in courts, the southern states have an edge over their northern counterparts. With the exception of Andhra Pradesh, the number of police firing incidents was lower in the southern than in the northern states. Based on this, we surmise that the potential (measured in terms of the initial conditions) for economic growth existed more in the south than in the north;

(c)

With respect to total per capita spending and per capita spending on education and public health, while the southern states spent less than the northern states during the 1980s, they spent more than the northern states post-1991, presumably implying a greater level and quality of public services.

We thus find that while the southern states had an edge with regard to the initial conditions of several factors that we have taken into account, it did not have an initial advantage in all of them (police firing (when AP is included), the stock of graduates and the proportion of population in the working age group in which the south and the north started off at the same point in 1961). A surprising finding is also that, while the southern states’ average weighted per capita NSDP was nearly twice that of the northern states in 2004, the growth rate of weighted per capita income was on average higher in the northern states (2.2 per cent) than that in the south (which was 1.78 per cent) during the period from 1960 to 1991. However, during 1992 to 2004, the average growth rate of weighted per capita NSDP in the southern states was 4.6 per cent compared with only 1.62 per cent in the northern states. This shows that the surge in the south is indeed a recent, post-1991 phenomenon. For an explanation of the intriguing phenomenon of the sudden growth of southern states in the 1990s, we turn to major policy changes that were occurring in the Indian economy since the mid-1980s, when the first steps towards decontrol and liberalization occurred in India (see Joshi and Little, 1997; Rodrik and Subramanian,

42


Vol. 18, No. 1, June 2011

2005). This is also consistent with what the earlier literature (for example, Mathur, 2001; Basu, 2004) finds. De-licensing of industries and more liberal policies towards foreign investment were adopted during this period. In 1991, the full-fledged economic liberalization further enabled the opening up of the Indian economy, which created favourable conditions for private sector investment, both domestic and foreign. These policy changes were exogenous and national, with all the states being free to take advantage of the opportunities it offered. So states that were more prepared (in terms of governance and infrastructure) to take the plunge forward succeeded, whereas the states that were less prepared in these terms could not do so. Ahluwalia (2000) also highlights how the economic liberalization reduced the degree of control exercised by the centre in many areas, leaving much greater scope for state-level initiatives, which is particularly true as far as attracting investment, both domestic and foreign, is concerned. Ahluwalia concludes that state-level performance and policies therefore deserve much closer attention than they receive. It is particularly important to study the differences in performance among states in order to extract lessons about what works and what does not. A better understanding of the reasons for the superior performance of some states would help to spread success from one part of the country to another. Overall, the upward shift in per capita income, downward trend in poverty reduction and much greater investment flows that occurred in the southern states relative to that in the northern states can be explained partly by the advantage the former had in terms of human capabilities, infrastructure, urbanization and some (not all) law and order indicators and partly by the economic liberalization of 1991.

43


Vol. 18, No. 1, June 2011

DATA APPENDIX Data on investments are from the Centre for Monitoring Indian Economy (CMIE) data set CAPEX. Data sources for education/health and urbanization indicators are the Census of India. Historical data on infant mortality rate are obtained from the publication, Sample Registration System: Statistical Report 2006, published by Census of India. SDP data are from the Central Statistical Organization (or the Economic and Political Weekly Research Foundation (EPWRF). Poverty data are from the Planning Commission. Law and order indicators, such as the number of police firing incidents, proportion of pending court cases, are from the National Crime Record Bureau. Infrastructure measures, such as installed electrical capacity, are from the Central Electricity Authority, Ministry of Power, Government of India. Data on percentage of households with electricity by state are from the Census of India. Data on telephone penetration are from the Department of Telecommunications (DoT), Ministry of Information Technology and Communications, Government of India. Data on total and developmental expenditures by sector (education, sports and culture (and that on energy, roads and bridges not reported for various reasons discussed in the paper), public health and medical facilities are from the EPW Research Foundation. Literacy rates, proportion graduate and percentage of working age group for all states by year are from the Census of India. Data on the proportion of technical degree holders are from the Ministry of Human Resources Development’s publication, Selected Educational Statistics. Annual time series data on the population in various states are from the EPWRF. Data on urbanization are from the Census of India. Data on higher educational institutions are aggregated at the state level from the Census of India town directories for towns and cities in the respective states.

44


Vol. 18, No. 1, June 2011

REFERENCES Ashraf, Quamrul H., Ashley Lester and David N. Weil (2008). “When does improving health raise GDP?”, NBER Working Paper No. 14449, October. Balakrishnan, Pulapre and M. Parameswaran (2007). “Understanding economic growth in India: A prerequisite”, Economic and Political Weekly, 14 July, pp. 2915-2922. Banerjee, Abhijit and L. Iyer (2005). “History, institutions, and economic performance: The legacy of colonial land tenure systems in India”, American Economic Review, vol. 95, No. 4, September, pp. 1190-1213. Basu, Sudip Ranjan (2004). “Economic growth, well-being and governance under economic reforms: Evidence from Indian States”, HEI Working Paper No: 05/2004, The Graduate Institute of International Studies (Geneva). Besley, Timothy J. and Robin Burgess (2000). “Land reform, poverty reduction and growth: Evidence from India”, Quarterly Journal of Economics, vol. 115, No. 2, pp. 389-430. Chaudhuri, S. (2000). “Economic growth in the State – Four Decades – I”, Money and Finance, October-December, pp. 45-69. Cohen, Barney (2004). “Urban growth in developing countries: A review of current trends and a caution regarding existing forecasts”, World Development vol. 32, No.1, pp. 23-51. Datt, Gaurav and Martin Ravallion (1998). “Why have some Indian states done better than others at reducing rural poverty? Economica vol. 65, Issue 257, February, pp. 17-38. Das, S.K. and A. Barua (1995). “Regional inequalities, economic growth and liberalisation: A study of the Indian economy”, Journal of Development Studies, vol. 32, No. 3, pp. 364-390. Dholakia, R.H. (1994). “Spatial dimension of acceleration of economic growth in India”, Economic and Political Weekly, 27 August , pp. 2303-2309. Dreze, Jean and Amartya Sen (1997). Indian Development: Selected Regional Perspectives. (Oxford: Clarendon Press). Fan, Shenggen, Peter Hazell, and Sukhadeo Thorat (2000). “Government spending’, growth and poverty in rural India”, American Journal of Agricultural Economics vol. 82, No. 4, November, pp. 1038-1051. George, Benny (2009). “Sanitation programmes: A glass half-full”, Economic and Political Weekly, 21 February, pp. 65-67. Ghate, Chetan and Steven Wright (2008). “The ‘V-Factor’: Distribution, timing and correlates of the great Indian growth turnaround, Unpublished paper, accessed from http://www.isid.ac.in/ ~cghate/vfactor_final.pdf. Ghosh, Madhusudan (2006). “Economic growth and human development in Indian States”, Economic and Political Weekly, 29 July, pp. 3321-3329. Guha, R., Southern Exceptionalism, Outlook. Hausmann, R., Lant Pritchett and Dani Rodrik (2005). “Growth accelerations”, Journal of Economic Growth, vol. 10, pp. 303-329. Joshi, Vijay and I.M.D. Little. (1997). India’s Economic Reforms: 1991-2001 (Delhi: Oxford University Press).

45


Vol. 18, No. 1, June 2011

Krishna, K.L. (2004). “Patterns and determinants of economic growth in Indian states”, ICRIER Working Paper No. 144, September. Kurian, N.J. (2000). “Widening regional disparities in India: Some indicators, Economic and Political Weekly, 12 February, pp. 538-550. Mani, M., S. Pargal and M. Huq (1996). “Does environmental regulation matter? Determinants of the location of new manufacturing plants in India in 1994”, World Bank Working Paper 1718, (Washington, D.C., World Bank). Mathur, A. (2001). “National and regional growth performance in the Indian economy: A sectoral analysis”, paper presented at the National Seminar on Economic Reforms and Employment in Indian Economy, IAMR. Nagaraj, R., A. Varoudakis and M.A. Veganzones (2000). “Long-run growth trends and convergence across Indian states”, Journal of International Development, vol. 12. Nair, K.R.G. (1982), Regional Experience in a Developing Economy, (New Delhi: Wiley Eastern). Norton, S.W. (1992). “Transaction costs, telecommunications, and the microeconomics of macroeconomic growth. Economic Development and Cultural Change, vol. 41, No. 1, pp. 175-196. Nunn, Nathan (2009). “The importance of history for economic development”, NBER Working Paper No. 14899, April, available at www.nber.org/papers/w14899, accessed on 29 June 2009. Olson, Mancur (1984), Rise and Decline of Nations (New Haven, Yale University). Paul, Samuel and Kala S. Sridhar (2009). “The paradox of India’s North-South divide”, Unpublished mimeo, May, Public Affairs Centre, Bangalore. Rao, M. Govinda, R.T. Shand and K.P. Kalirajan (1999). “Convergence of incomes across Indian states: A divergent view”, Economic and Political Weekly, 27 March 27, pp. 769-778. Rao, Govinda M. and N. Singh (2005). Political Economy of Federalism in India (New Delhi, Oxford University Press). Rodrik, Dani and Arvind Subramanian (2005). “From Hindu growth to productivity surge: The mystery of the Indian growth transition”, IMF Staff Papers, vol. 52, No. 2, (Washington, D.C., IMF). Roller, L.H. and Leonard Waverman (2001). “Telecommunications infrastructure and economic development: A simultaneous approach” American Economic Review vol. 91, No. 4, pp. 909-923. Roy Choudhury, U.D. (1993). “Interstate variations in economic development and standard of living”, (New Delhi, National Institute of Public Finance and Policy). Sachs, J.D., N. Bajpai and A. Ramiah (2002). “Geography and regional growth”, The Hindu, 25-26 February. Shand, R. and S. Bhide (2000). “Sources of economic growth: Regional dimensions of reforms”, Economic and Political Weekly, 24 October, pp. 3747-3757. Sridhar, V. (2007). “Growth of mobile services across regions of India,” Journal of Scientific and Industrial Research, vol. 66, April, pp. 281-289. Virmani, Arvind (2006). “India’s economic growth history: fluctuations, trends, breakpoints, and phases”, Indian Economic Review, vol. 41, No. 1, pp. 81-103. Wallack, J. (2003). “Structural breaks in Indian macroeconomic data”, Economic and Political Weekly, 11 October.

46


Vol. 18, No. 1, June 2011

REACHING A UNIVERSAL HEALTH INSURANCE IN VIET NAM: CHALLENGES AND THE ROLE OF GOVERNMENT Giang Thanh Long*

Given the level of economic development in Viet Nam, the country’s health sector is quite impressive. The country has now set its sights establishing universal health insurance coverage by 2015. This article aims to, first, describe the current health-care system in Viet Nam accompanied by a number of challenges the country faces in delivering and financing health-related services, and then provides some policy discussions on how to achieve this ambitious plan. The article also stresses the important role in attaining universal health insurance coverage by providing quality services and guaranteeing financial protection for both health-care suppliers and consumers.

JEL classification code: I14, I18. Key words: Health-care system, universal health insurance, Viet Nam.

I.

INTRODUCTION

The implementation of the Doi Moi (renovation) programmes about 20 year ago has transformed Viet Nam from being one of the poorest countries in the world in the late 1980s to a low middle-income country since 2008. The average gross domestic product (GDP) growth rate was about 7.4 per cent during 1991 to 2010, and this, in turn, has helped boost GDP per capita from $98 in 1990 to $1,170 in 2010. Also, thanks to this remarkable economic growth, the national poverty rate decreased significantly from 58.1 per cent in 1993 to about 9.5 per cent in 2010. Along with

* Giang Thanh Long, Ph.D., National Economics University (NEU) and Indochina Research and Consulting (IRC), Hanoi, Viet Nam, [email protected]. The author would like to thank the participants in the UNDP-VASS Seminar on Human Development in Viet Nam for their useful comments on an earlier draft of this article. The author is also thankful to two anonymous referees of the Journal for their insightful comments and suggestions, which in turn helped to improve the article.

47


Vol. 18, No. 1, June 2011

these remarkable social and economic achievements, great strides have been made in the country’s health sector, in which the health outcomes are similar to those of high-middle income and advanced countries in many aspects. The Vietnamese now enjoy a healthier and longer life, which, in turn, has lifted the country into the ranks of medium human development countries, placing 113 out of 169 countries in the world in 2010 (UNDP, 2010). The Government recently announced its plan to achieve universal health insurance coverage by 2015. This plan entails: strengthening the health-care system, extending delivery of health-care services to all people and improving the quality and contents of health-care packages offered by both public and private health providers to insured beneficiaries. By 2010, health insurance coverage had reached 62 per cent of the total population or about 53 million persons. Under current regulations, the compulsory health insurance scheme covers people working in the formal sector, as well as specific groups of the population (including the poor, children under six years of age and veterans). As indicated in the Health Insurance Law enacted on 1 January 2007, coverage will be gradually expanded to the rest of the population until universal coverage is reached: first to all students, then to the “near-poor” (who are not poor, but his/her household per capita expenditure is less than 125% poverty line), and finally to all workers in the informal sector. This is expected to be completed by 2015. Several challenging issues must be dealt with in the transition towards a universal health insurance system, including the delivery and financing of healthcare services. With regard to delivery, achieving such an ambitious policy goal will largely depend on the ability to extend the existing voluntary health insurance scheme to people working in the informal sector. In other words, making health insurance compulsory for all Vietnamese. Notably, in the past years, few of the people working in the informal sector have voluntarily joined the health insurance system, and based on this alone, reaching the goal of universal health insurance coverage by 2015 will require a change in mindset among informal workers and non-working family members towards health insurance. In addition, achieving universal health insurance coverage will also require dealing with financing issue. By 2008, health insurance revenue represented less than 1 per cent of GDP, and health insurance covered only 13 per cent of total health expenditure (Viet Nam, 2007). Other financing sources for health expenditure, particularly from public ones, remain stable and low. As such, out-of-pocket (OOP) spending consequently remains high. Increasing public spending on health to encourage participation of informal workers and their dependents will expand the coverage of health insurance as well as help relieve financial constrains in the longer term.

48


Vol. 18, No. 1, June 2011

The main objective of this paper is to describe the current state and challenges of the health-care system in Viet Nam with a special focus on financing and delivery of services through health insurance, and to provide some policy suggestions to deal with issues related to the goal of achieving universal health insurance by 2015. The paper is organized as follows. In the next section, we will provide an overview on the development of the Vietnamese health-care system, with a discussion on its current state and the challenges in financing and the delivery of services. The third section will provide an analysis of the delivery patterns of health-care services through health insurance. The paper will also discuss financial requirements for reaching a universal health insurance with simulation results from Lieberman and Wagstaff (2008). The final part will provide concluding remarks.

II.

HEALTH-CARE PERFORMANCE, SERVICES DELIVERY AND FINANCING IN VIET NAM

The health sector has received special attention from the Government, in particular the Ministry of Health (MOH), which has instituted a number of policies, guidelines, and regulations to strengthen the health system and improve the health of the Vietnamese people. This section briefly presents several main health policies which aim to deliver efficient and financially viable health care services. Performance of health sector and challenges Among the eight Millennium Development Goals, three are related to basic health indicators, including reduction of child mortality; improvement of maternal health, and tackling of HIV/AIDS, malaria, and other diseases. As assessed by a number of studies, such as Adams (2005) and WHO (2009), Viet Nam has achieved certain successes along the pathway towards the Goals. Viet Nam has achieved most of the Goals during the period 2006 to 2010. For instance, in 2009, the maternal mortality rate (MMR) reached 75 per 100,000 live births in comparison with 233 per 100,000 live births in 1990; under-five child malnutrition rate was 20.6 per cent, which is close to the target of 20 per cent by 2010; and under-five mortality (U5MR) and infant mortality rate (IMR) per 1,000 live births reached 25 and 15, respectively, attaining the targets of 25 and 16, respectively, by 2010 (figure 1).

49


Vol. 18, No. 1, June 2011

Figure 1. U5MR and IMF in Viet Nam, 1990-2009 IMR

U5MR 50

70 58

44.4

40

25 19.3

20

24.8

20

21 17.8

16

16

15

15

2009

26 25.9 25.5 25

2008

30

30 27.5

2007

42

40 30

2006

50

16 14.8

10

10 0

MDG 2015

M. tiêu 2010

2005

2003

2002

2001

05 20 06 20 07 20 08 20 M .t 09 iê u 20 M 10 D G 20 15

01

Source:

20

20

19 9

0

0

1990

60

Ministry of Health (various years)

Figure 2. Trends in age-adjusted mortality: Viet Nam, Malaysia and Thailand

Source:

WHO Life Tables for Member States, as quoted by Lieberman and Wagstaff (2008).

Viet Nam has also done well in terms of child mortality, in which there has been a rapid reduction in comparison with Malaysia and Thailand – two Asian countries with significantly higher per capita incomes than Viet Nam. Unlike Thailand, which saw rising age-specific mortality rates at some ages between 2000 and 2005, and Malaysia, where the decline in age-specific mortality appeared to have stagnated,

50


Vol. 18, No. 1, June 2011

Viet Nam saw reductions in age-specific mortality at all ages between 2000 and 2005. Also, Viet Nam’s age-specific death rates compared favourably with those of Malaysia across the full range of ages (figure 2). In accordance with a number of other studies, Lieberman and Wagstaff (2008) show that such impressive achievements in the health sector are results of a widely-covered grass-roots health system in Viet Nam focusing on various preventive health-care services. However, a number of critical challenges need to be responded to if Viet Nam is to continue its above-mentioned achievements. First, there are wide gaps and notable variations in IMR by region. For instance, in the three regions with the highest IMRs, an infant has a probability of dying before reaching his or her first birthday about two and a half times higher than an infant in the three regions with the lowest IMRs (figure 3). Particularly high IMR has been observed in certain provinces. In the period 2005 to 2008, the highest IMR incidence was in Kontum (62.6 per cent in 2005; 48 per cent in 2008); Ha Giang (55.8 per cent and 40.0 per cent, respectively); Lai Chau (44.0 per cent and 33.0 per cent, respectively). These provinces experienced an IMR of about 5 to 6 times as much as developed provinces, such as Hanoi (7.9 per cent and 7.0 per cent, respectively) and about 2 to 3 times as much as the national average (17.8 per cent and 15.0 per cent, respectively). Such a disparity could be explained in large part by differences in the level of development among the regions in such factors as income, education, infrastructure and availability of services. Figure 3. Regional differences in IMR 40 Northwest Central Highlands

30

Northeast North Central 20 Central Coast Mekong River Delta 10 Red River Delta Southeast 0 2002 Source:

2003

2004

2005

Hayes and others (2009).

51


Vol. 18, No. 1, June 2011

Significant differences in IMR have also been observed between income groups and rural and urban areas. The gaps in IMR between the poorest and richest groups, and between rural and urban areas have widened (figure 4). This situation indicates that persons from rural areas and with low incomes will have difficulty in overcoming the high incidence of IMR, which in turn, will have a negative impact on their human development. Figure 4. Differences in IMR between income groups and rural and urban areas 100

60

71 59

56

12%

50 cc to 250 cc

Static converters, nes

Transformers electric power handling capacity not exceeding 1 KVA, nes

Digital data processing systems, nes

Electrical relays for a voltage not exceeding 60 volts

Boxes, cases, crates andsimilar articles of plastic

Aircraft under-carriages and parts thereof

Amino-alcohol-phenol, amino-acid-phenol & oth amino-compds with oxygen func

Horizontal lathes numerically controlled for removing metal

Description

1 958

2 423

11 632

3 353

14 305

2 436

6 260

1 818

3 732

1 886

2 311

2 302

APEC

3 179

3 216

21 785

5 051

21 978

3 787

9 435

3 011

6 889

2 559

3 465

3 704

World

2 428

2 997

14 371

4 132

17 626

3 001

7 706

2 235

4 580

2 314

2 832

2 819

World less intraEuropean Union trade

Exports (USD billion)

ANNEX A (continued)

61.61

75.35

53.39

66.39

65.09

64.33

66.35

60.36

54.17

73.69

66.69

62.15

To world

80.65

80.86

80.94

81.14

81.16

81.17

81.24

81.31

81.49

81.50

81.60

81.66

19.04

5.51

27.55

14.75

16.07

16.84

14.89

20.95

27.32

7.81

14.91

19.51

To world less intraDifference European Union rade

Share (percentage)

Asia-Pacific Development Journal Vol. 18, No. 1, June 2011

107

108

854150

180

854390

176

611120

390190

175

179

540233

174

390760

442190

173

740311

820730

177

Radio navigational aid apparatus

852691

171

172

178

Knittd/crochetd tex fab, width >30 cm, >/= 5 per cent of elastomeric/rubber, nes

600230

170

Semiconductor devices, nes

Babies garments and clothing accessories of cotton, knitted

Copper cathodes and sections of cathodes unwrought

Polyethylene terephthalate

Parts of electrical machines and apparatus havg individual functions, nes

Polymers of ethylene nes, in primary forms

Textured yarn nes, of polyester filaments, not put up for retail sale

Wood articles nes

Tools for pressing, stamping or punching

Plywood, outer ply of non-conifer wood nes, ply 600 mm x 600 mm x 0.5-1 mm

Tableware and kitchenware of plastics

Lumber, coniferous (softwood) 6 mm and thicker

Combined refrigerator-freezers, fitted with separate external doors

Electric conductors, for a voltage not exceeding 80 V, nes

Tools, nes, hand-held, with selfcontained electric motor

Organo-sulphur compounds, nes

Gold (incl. gold plated with/platinum) unwrought/semi-manufactured or powdered

Description

2 047

1 573

7 178

4 316

2 188

12 029

3 347

2 138

2 324

2 797

22 584

APEC

4 210

2 787

13 744

7 747

4 231

22 614

6 005

4 025

4 518

5 023

31 732

World

2 791

2 143

9 723

5 841

2 956

16 244

4 516

2 876

3 123

3 754

30 223



ANNEX A (continued)

48.63

56.45

52.22

55.71

51.72

53.19

55.74

53.13

51.45

55.68

71.17

To world

73.36

73.42

73.82

73.90

74.02

74.06

74.11

74.35

74.42

74.50

74.73

24.73

16.97

21.60

18.19

22.30

20.87

18.37

21.22

22.97

18.82

3.56


Share (percentage)


HS code

611030

392690

8525

901819

441219

640399

852910

284410

890190

902780

Rank

242

243

244

245

246

247

248

249

250

251

Instruments and apparatus for physical or chemical analysis, nes

Cargo vessels nes & oth vessels for the transport of both persons & goods

Natural uranium & its compounds; mixtures cntg natural uranium/its compds

Aerials & aerial reflectors of all kinds; parts suitable f use therewith

Footwear, outer soles of rubber/ plastics uppers of leather, nes

Plywood nes, at least 1 outer ply of coniferous wood (ply’s /= 600 mm wide

Lumber, non-coniferous nes

Wire of refind copper of which the max cross sectional dimension >6 mm

Maize (corn) nes

721070

440799

740811

100590

847989

8704

870324

721913

853650

851790

253

254

255

256

257

258

259

260

261

262

Parts of electrical apparatus for line telephone or line telegraphy

Electrical switches for a voltage not exceeding 1 000 volts, nes

Flat rolld prod, stainless steel, hr in coil, w>/= 600 mm, 3

Automobiles with reciprocating piston engine displacing >3 000 cc

Motor vehicles for the transport of goods

Machines andmechanical appliances nes having individual functions

Lubricatg oil additives cntg pet oils/ oils obtaind from bitu minerals

381121

252

Description

HS code

Rank

15 839

5 859

2 374

78 339

40 479

20 841

6 035

4 298

2 363

1 843

2 005

APEC

26 351

12 433

4 035

118 245

82 000

33 939

10 124

8 977

3 699

4 594

3 927

World

22 015

8 141

3 299

108 610

56 105

28 879

8 359

5 943

3 266

2 547

2 768



ANNEX A (continued)

60.11

47.12

58.85

66.25

49.37

61.41

59.62

47.87

63.87

40.13

51.06

To world

71.94

71.97

71.98

72.13

72.15

72.17

72.20

72.32

72.35

72.37

72.44

11.83

24.85

13.13

5.88

22.78

10.76

12.58

24.45

8.48

32.24

21.38


Share (percentage)


Isocyanates

292910

391910

853710

264

265

266

Table, kitchen or other household art & parts thereof, stainless steel, nes

Dog or cat food put up for retail sale

732393

230910

720421

854212

100190

711719

269

270

271

272

273

274

Household and toilet articles nes, of plastics

Imitation jewellery nes of base metal whether o not platd with prec metal

Wheat nes and meslin

Cards incorporating electronic integrated circuits

Waste and scrap, stainless steel

Motor vehicle parts nes

392490

870899

267

268

Boards, panels, includg numerical control panels, for a voltage < V >

Self-adhesive plates, sheets, film etc., of plastic in rolls / = 600 mm, 0.5 mm
1), ρ1 = 1 in which y– is the average per capita expenditure in the region. In fact, the Gini coefficient is the area above the Lorenz curve and below the diagonal 45-degree line divided by the area under the diagonal line.

III.

NUMERICAL RESULTS

An extensive number of estimates has been produced by using the abovementioned techniques, making comparisons and interpretations difficult. To prevent confusion, this report does not focus on detailed estimations over small areas but instead concentrates on some extreme cases. Maps are very instrumental tools to visually compare different areas and diagnosing the patterns and disparities of

8

The reader may refer to Deaton (1997) for more details.

136


Vol. 18, No. 1, June 2011

indicators over the country.9 Annex figure A.1 shows the location and names of provinces. National and regional poverty and inequality The results indicate that the national headcount poverty rate (P0) in rural areas (15 per cent) is significantly higher than in urban areas (11 per cent). This gap is even wider when comparing the depth of poverty (P1). The national level in rural areas is 0.04 compared to 0.029 in urban areas. As for the severity of poverty, the (P1) indicator, in which the distribution of the expenditure is rated in addition to the poverty gap, the figure is 50 per cent higher in rural areas at 0.018 versus 0.012 in urban areas. For the study, the Gini index was used to measure inequality in terms of per capita consumption expenditure. Even though poverty incidence is higher in rural areas, the results show that income (expenditure) is distributed more equally in rural areas than in urban areas. This is due to homogeneity of living standards among rural households. At the regional level, the rankings of urban and rural areas in terms of P0 by province tend to vary. This is not the case, however, for province 11 (Sistãn va Baluchestãn), ranked as the poorest province in both its urban and rural areas and province 02 (Mãzandarãn), ranked as the second richest province in both areas. The largest gap between urban and rural areas in terms of poverty incidence has been detected in provinces 23 (Tehrãn) and 15 (Lorestãn). In these two provinces, the people in the rural areas are much poorer than those in urban areas (more than twice). Annex table A.5 shows national and regional estimates of poverty and inequality. Poverty and inequality in small areas The Islamic Republic of Iran consists of 336 counties. To account for the urban and rural areas in each county, welfare indicators have been estimated in 672 small areas which range in population from 781 to 7,872,280. This level of disaggregation is very informative for pro-poor policies since it provides key information for studying patterns in poverty and inequality and their relationships with other characteristics of small areas.

9

Detailed statistics on the small area estimations are presented in Bidarbakht-Nia (2009) and are also available from the author upon request.

137


Vol. 18, No. 1, June 2011

Poverty and inequality in rural areas A comparison of the estimates of the regional level with those of the small areas indicates that using only regionally disaggregated information in policymaking is unreliable. Although province 01 (Gilãn), for example, has a relatively low rate of poverty and inequality, there is considerable diversity in all of the indicators among the counties in the province. In province 05, (Kermãnshãh), the discrepancy is wide with regards to expenditure inequality and in provinces 06 (Khuzestãn), 08 (Kermãn), 10 (Esfahãn) and 27 (Golestãn), only the poverty indicators vary within each province. In province 00 (Markazi), most of the counties fall in the first quintile (poorest 20 per cent) and in 03 (East Ajarbãijãn), all counties fall in the last quintile of inequality (20 per cent highest inequality). In province 06 (Khuzestãn), most of the counties have high poverty and low inequality. In provinces 18 (Bushehr), 26 (Ghazvin), 24 (Ardebil) and 02 (Mãzandarãn), all of their counties are relatively better off when compared with provinces 11 (Sistãn va Baluchestãn) and 21 (Yazd) in which almost all of their counties suffer from high poverty and inequality. Kohrang and Ardal are the only counties in 14 (Chãharmahãl va Bakhtiãri) that suffer from a high poverty rate. As extreme cases, Savãd-Kuh, Jam and Tonekabon have the lowest poverty rate (less than 3 per cent) and Irãnshahr, Sarbãz and Shãdegãn have the highest poverty rate (more than 47 per cent). Of note, all the counties in the 11 (Sistãn va Baluchestãn) fall in the last quintile in terms of poverty but vary among quintiles in terms of inequality. Poverty and inequality in urban areas As for the rural areas, diversity among counties inside the provinces is obvious in some cases. Poverty indicators in provinces 27 (Golestãn), 14 (Chãharmahãl va Bakhtiãri), 17 (Kohgiluyeh va Boyrahmad) and 23 (Tehrãn) vary widely among their small areas. In province 08 (Kermãn), the county Bam, is an exception as its poverty incidence and inequality (poverty rate 6 per cent and Gini index 0.37) are both relatively low. This is due to the reconstruction activities and aid tied to the 2003 earthquake in this area. In province 23 (Tehrãn), the characteristics of the capital city, also called Tehrãn, is a totally different than what is found in the counties. Tehrãn enjoys a low poverty rate compared with other counties in the province but its Gini coefficients are almost the same, with the exception of some small counties in which inequality is not surprisingly very low due to similarities between households. Province 21 (Yazd), 22 (Hormozgãn) and 29 (South Khorasãn) have high levels of poverty and inequality in almost all their counties. In (13) Hamedãn, most of the areas have high inequality but low poverty rates. In 11 (Sistãn va Baluchestãn), a majority of its counties have poverty rates between 30 per cent

138


Vol. 18, No. 1, June 2011

and 62 per cent. Two counties in the province that are an exception to this are Zahak and Konãrak, which have very low poverty and inequality levels. In province 06 (Khuzestãn), half of its counties are in the last quintile (poorest 20 per cent) with respect to the poverty rate despite the fact that the province has a lot of industry, including oil production, contributes a share of 15.4 per cent to the country’s GDP (the second highest share after Tehrãn) and has the second highest per capita GDP after province 17 (Kohgiluyeh va Boyrahmad). This shows that there is a gap between production value added (or GDP) and disposable income. Although per capita GDP mainly represents the productivity of the people in one region, it is not a good proxy for income in regional studies. This is because the central government tends to control the reallocation of the budget and one of the factors for determining how the funds are distributed is based on the political power of local governments, which often takes precedence over regional GDP. Surprisingly, inequality is relatively low in this province (mostly around 0.35). The least poor small urban areas (with a poverty rate of less than 3 per cent) are Jam, Bushehr, Gilãngharb, Bijãr, Nowshahr, Ghasr-e-Shirin, Tonekãbon and Marivãn. In contrast, Sarbãz, Sarãvãn, Kalãt, Bahme’ei and Khavãf have the highest level of poverty incidence (more than 47 per cent). In summary, the estimated poverty rate is less than 0.10 in 31 per cent of the rural counties and more than 0.20 in almost 29 per cent of counties. These statistics are correspondingly 36 per cent and 26 per cent for urban counties which shows that more rural areas are in poverty (in terms of proportion) than urban areas. Mapping poverty and inequality Poverty mapping methods have been utilized to visually investigate patterns in welfare indicators in the country. The incidence of poverty in urban areas is shown in annex figures A.2 and A.3. A comparison between two figures shows that the poverty rate varies widely within one province and that the scale of discrepancy is very different among provinces. For example, in provinces 09 (Khorãsan Razavi) and 11 (Sistãn va Baluchestãn), poverty rates among counties range between less than 6 per cent and more than 60 per cent, while in provinces 04 (West Azerbãijãn), 18 (Bushehr), 12 (Kordestãn), 05 (Kermãnshãh) and 02 (Mãzandarãn), the rates differ between 1 per cent and 15 per cent. Generally speaking, in the north-central and north-western parts of the country, there is homogeneity among the interior areas, while in the eastern and south-western regions, the provinces are heterogeneous and generally have higher rates of poverty. In terms of climate circumstances, with the exception of 18 (Bushehr) and 29 (South Khorãsãn), the provinces under study with favourable weather conditions, have relatively low poverty levels in their urban areas.

139


Vol. 18, No. 1, June 2011

Some may assume that this is tied to productive agricultural activities but it could easily be tied to other income-generating factors, such as tourism and access to the foreign market (either legal or illegal). In the case of province 18 (Bushehr), the low level of poverty may be attributed to income generated from a special economic zone set up there that enables access to foreign markets through the Persian Gulf. Nevertheless, the information available is not definitive and can only serve as guide in determining the patterns and the magnitude of the effects of some explanatory variables. More information is needed to ascertain the spatial differences and their determinants. Annex figures A.4 and A.5 illustrate rural poverty rates in provinces and small areas, respectively. Similar to the urban case, the disparity in some provinces is considerable in rural areas. The pattern of poverty in rural areas in the north-western part of the country is quite different from its urban counterparts. The urban areas are better off than the rural areas, with the exception of the borderline areas in the northern part of this region where the economy benefits from access to foreign markets and productive agricultural activities. Another way to determine patterns of poverty is to estimate the number of people whose income is under the poverty line in each small area (poverty density). Poverty density for each area is estimated by multiplying the corresponding poverty rate times the population. Annex figures A.6 and A.7 depict poverty densities for urban and rural areas, respectively. In these maps, each dot represents 400 individuals living under the provincial poverty line. Notably, in both urban and rural areas, regions with high levels of poverty tend to have low poverty density. An exception to this is the south-east region. As poor areas are mainly located in the eastern part of the country, which, except for the extreme north-east, has a desert climate and is less populated than other regions. Generally speaking, many areas with low level of poverty incidence, which may not be considered as a priority in pro-poor policies, have high density of poverty. Therefore, it must be noted that relying only on the poverty rate may cause exclusion of the majority of the poor people in the country. To avoid this, it is very helpful to carefully examine all poverty and inequality indicators simultaneously for areas. Since none of the existing poverty indicators contributes poverty density in the formula it is more likely to be ignored. A comparison of the maps on poverty rate and poverty density shows that in some provinces, such as 06 (Khuzestãn), 25 (Ghom), and in some areas in 11 (Sistãn va Baluchestãn), there is a high level of poverty not only in terms of percentage but also regarding population in poverty. Consequently, multivariable clustering is required to put similar counties in the same category (discussed later in this section). A comparison between rural and urban areas regarding their inequality Gini index (annex figures A.8 – A.11) for the country as a whole illustrates that the general level of inequality in rural areas is lower than in urban areas. This can attributed to the

140


Vol. 18, No. 1, June 2011

narrower differences in living standards among rural households. Despite the diversity in poverty, inequality is homogeneous across the country. As discussed in detail later in this study, location is the main explanatory variable for poverty incidence. On the other hand, household characteristics have more of an effect on the level of inequality than location effects. Finally, to assist in setting priorities for policymaking, small areas have been clustered based on the poverty and inequality indicators (three FGT indicators, the Gini coefficient and poverty density) utilizing a fuzzy clustering technique. Small areas in each cluster are similar regarding to all of their poverty and inequality indicators.10 Table 1 shows the cluster means of poverty and inequality indicators. In this table, small areas in cluster 1 can be considered non-poor on average, based on all indicators, and sequentially clusters 2 and 3 are worse off except in the case of Gini coefficient in rural areas. Table 1. Cluster means of welfare indicators Cluster* Urban

Rural

P0

P1

P2

GINI

Poverty Density

1

0.08

0.02

0.005

0.37

5 808

2

0.14

0.03

0.012

0.39

13 342

3

0.24

0.07

0.028

0.41

27 978

1

0.09

0.02

0.007

0.34

4 682

2

0.16

0.04

0.015

0.39

8 348

3

0.26

0.07

0.029

0.37

19 493

* Number of clusters (3) is arbitrarily determined.

This classification does not mean that, for example, every small area in cluster 1 has a lower poverty incidence than small areas in other clusters. Detailed information shows that there might be small areas with one or more relatively high welfare indicators in cluster 1. This is because the membership function for each small area is a multivariate function in which all the indicators are simultaneously considered. Annex figures A.12 and A.13 map the clustered small areas. Poverty and inequality correlates The evaluations outlined in previous sections indicate that the patterns of poverty and inequality vary greatly across the country and between urban and rural areas. Policies aimed at mitigating household deprivations need to be informed with 10

For more detail explanations about fuzzy clustering, please refer to Bezdek (1980).

141


Vol. 18, No. 1, June 2011

disaggregated statistics that visually illustrate areas and subpopulations with high priorities regarding poverty and inequality. Nevertheless, at this point, two important questions remain unanswered; “what characteristics explain the inequality?” and “who is more likely to be poor?” This study does not assume that there is causality relationship between poverty and inequality and household characteristics. Instead, it determines what factors explain inequality and are more closely related to the likelihood of being poor. To show the general effect of explanatory variables on poverty and inequality, the country level models (for rural and urban separately) have been fitted. Moreover, a linear logistic model has been estimated separately for each province in order to monitor different effects of explanatory variables on poverty across regions. Regarding to the inequality decomposition, the contribution of each explanatory variable (Sk) and, to be more precise, the share of each variable in the R-square (Pk) have been calculated for the national level.11 Bar charts in figure 1 show the percentage share of each category in the predictability of the country-level model. In other words, it indicates the overall effect of each group of explanatory variables on the expenditure distribution in country level. The asset category is the most influential factor in both urban and rural with almost half of the contribution in the per capita expenditure variation. In both rural and urban areas, education level and location are two other effective categories with the difference being that location has much greater effect in rural areas than in urban areas. This may be attributed to the many spatial characteristics which could affect the income level and income inequality in rural areas, such as transportation facilities, natural endowments and distances to the big cities. As location is an exogenous variable, this result suggests that investing in infrastructure in rural areas could mitigate inequality. However, to determine this, more Geographical Information Systems (GIS) information is required to analyse the spatial determinants of the poverty and inequality. Inside categories, the most effective variables are car ownership and the floor area per person in the asset group for both urban and rural areas. In the case of demographic the size of household has the most influence on the per capita expenditure disparity. The interpretation of coefficients in the logit model is not straightforward and usually requires some extra work after parameter estimations have been made. In this paper, the explanatory variables (annex table A.4) are divided into three groups according to their measurement scale: dummy variables, ratios and other types (continuous or counts). The coefficients are interpreted as the amount of change in log odds due to a one unit increase in explanatory variables. Thus, absolute values of the coefficients are comparable in each category; the corresponding explanatory

11

Since semi-logarithmic functional form applied to inequality decomposition is not mean-independent, only a country model has been fitted for inequality decomposition.

142


Vol. 18, No. 1, June 2011

Figure 1. Percentage contribution of each category of variables in the R-square of expenditure model Per cent 60

47

Urban

Rural

41 40

21 20

17 16

17 12 11 6

8

3

2

n tio ca Lo

ic bl Pu

pl

oy

U

m

til

en

ity

t

n io at uc Ed

Em

D

em

og

ra

As

ph

se

ic

t

0

variable becomes more effective as the absolute value of the coefficient increases. The logit model has been applied at the country and the provincial levels in both rural and urban areas. Table 2 illustrates the most effective variables in country model. Note that location dummy variables (336-1 dummies for counties’ effects in the model) have been excluded from the comparison. However, when the location parameter estimation is included, it can be observed that in the dummy group, about 90 per cent of location effects (either positive or negative effects depending on the degree of poverty) in urban areas and 60 per cent in rural areas are significant and more influential than other dummy variables. Location effects are more informative when using provincial models for geographically targeted pro-poor policies. The regional results (not presented here) indicated that in almost every province location is highly correlated with poverty likelihood. These results only show the degree of importance of location when analysing poverty. However, information available cannot explain why location is such a determinant factor. Lack of GIS information in the Islamic Republic of Iran, such as distance to the markets (domestic and foreign) and facilities, soil fertility and climate

143


Vol. 18, No. 1, June 2011

Table 2. The most effective explanatory variables in each group for the country logit model Urban Variable

Type

Rural Parameter estimates

Variable

Type

Parameter estimates

Internet

Dummy

-1.11

Car

Dummy

-0.75

Car

Dummy

-0.85

Computer

Dummy

-0.53

Heating fuel

Dummy

-0.8

Structure

Dummy

-0.49

Employed

Ratio

-1.36

Employed

Ratio

-2.54

Literate

Ratio

-1.34

Under 6

Ratio

-2.29

Area per capita

Ratio

-1.26

Rooms per capita

Ratio

-1.46

College

Other

-.079

Size

Other

0.73

Size

Other

0.34

College

Other

-0.71

does not allow spatial analysis in our case. Although the most effective variables from the country level were presented in the above table, locally targeted policies need to consider provincial models in which every province has its own characteristics and poverty attendant correlates. For instance, in country as a whole, car ownership significantly increases the likelihood that a person will be classified as non-poor, while in provincial models, this variable has very different effects in different provinces.

IV.

POLICY IMPLICATIONS AND CONCLUSION

The results of the study clearly show that the Islamic Republic of Iran is in need of a transparent and comprehensive anti-poverty policy in order to overcome its high level of poverty and inequality. Maps and statistics not only delineate these high levels but also show the disparity of the extent to which people are in poverty among regions. Numerical results and maps also illustrate patterns of poverty and inequality in small geographic regions and provide useful information about their attendant correlates. Maps depict the priority areas regarding each aspect of poverty and inequality. The results also suggest that location has a great deal to do with understanding the economic welfare of households. Small area estimations uncovered that even within prosperous regions, there may be a few areas with extreme poverty levels. Some regions are uniformly in poverty while in others, the poverty rate is deceptively low and the poverty density is high. Clustering methods are efficient instruments to classify small areas in similar groups regarding all aspects of poverty and inequality and should be utilized to avoid confusion.

144


Vol. 18, No. 1, June 2011

Small area estimations of poverty gap are very helpful for estimating the minimum budget required for eliminating poverty in particular region. This enables governments to avoid extreme payments during budget allocation. Table 3 shows the budget required to raise all members of the population above the corresponding poverty line using estimates of poverty gap in different levels of disaggregation. The difference between two budget requirements is a measure of efficiency that government can gain from small area estimations. Applying an informed welfare policy with small area estimations and saving 27 per cent of budget, government can take the opportunity to spend a significant share of the subsidy revenue on development programmes according to the regional requirement. Table 3. Required budget for eliminating poverty Level

Required budget* (USD billion)

County

1.00

Province

1.37

* Required budget to raise every one above poverty line = poverty gap x poverty line x population

The discussion in this paper suggests that small area estimations are beneficial for setting a geographically targeted pro-poor policy. Although the results do not indicate which geographical factors affect economic welfare, a spatial analysis using detailed GIS information can be applied to obtain additional input as complementary information for policy implications. Reallocating a budget based on regional requirements not only alleviates poverty and inequality but also creates jobs and boosts production and income, which consequently raises living standards. For instance, geographical factors in rural areas, which cannot be addressed by current policy, may be insufficient educational facilities, an inefficient health care system and limited infrastructure. One reason behind this is that government has no control on intrahousehold allocations and usually in poor households, the allowance received by head of household is not spent on education, health and other needs of children. In contrast, in a growth-oriented policy, the government takes into account regional factors and household characteristics, and invests more efficiently. As a result, governmental bodies do not waste time and energy on redistributional activities, but instead focus on reallocating budgets and setting regional policies. In short, policymakers may switch their focus from addressing “poor people” to “poor areas” (considering available information and policy instruments) in order to attain more accurate outcomes.

145


Vol. 18, No. 1, June 2011

Generally speaking, the effective use of SAE can lead to fundamental changes in government interventions and development plans. This method is particularly useful when a government faces a budget deficit and must rationalize in detail its expenditures. Visualizing welfare indicators in small areas can facilitate evidence-based policymaking and efficient use of scarce resources. The policy implications of this paper are applicable to other countries that are similar to the Islamic Republic of Iran. Indonesia’s policy on energy subsidies is an extreme case in East Asia in which different households and different regions are disproportionately benefiting from the subsidies. The government has been struggling for a long time on whether to free fuel prices or balance the subsidiary benefits among households and regions. Nevertheless, according to Bedi, and others (2007), many government agencies had never utilized SAE and poverty maps for setting polices despite the substantial variation in poverty and inequality among local areas. In fact, the study done for the World Bank indicates that poverty maps have influenced policymaking in several countries. In Sri Lanka, poverty incidence is highly correlated with access to the market and the SAE technique has served as an essential tool for Samurdhi, the country’s main poverty alleviation programme, which is under the authority of the Samurdhi (or prosperity) Ministry. Through this programme, 113 of the poorest areas were identified as being target points as the ministry reformed its transfer programme. As another example, a simulation study by Elbers and others (2007) using data from Cambodia shows that by utilizing SAE for efforts aimed at alleviating poverty, specific areas can be pinpointed and targeted for a specific programme, resulting in a budget that is less than one-third of one for a comparable untargeted programme. Poverty maps from other countries in Asia, such as Bangladesh, Cambodia, Nepal and Viet Nam, show that there are large-scale discrepancies in depth and incidence of poverty among small geographic areas, but this information has been underutilized. The additional evidence from the Islamic Republic of Iran to the substantive literature on poverty and inequality studies in developing countries reinforces the findings that SAE is an efficient tool for evidence-based policymaking.

146


Vol. 18, No. 1, June 2011

REFERENCES Akhavi, A. (1996). “Has number of poor increased? (Aya Faghiran Afzayesh Yafteh-and?)”. Economic Analysis of Poverty, Institute for Trade Studies and Research. Assadzadeh, A. and S. Paul (2004). “Poverty, growth, and redistribution: a study of Iran”. Review of Development Economics, November, vol. 8, Issue 4, pp. 640-653. Bezdek, J.C. (1980). “A convergence theorem for the fuzzy ISODATA clustering algorithms”. IEEE Transactions on Pattern Analysis and Machine Intelligence, January, vol. 2, Issue 1, pp.1-8. Bedi, T., A. Coudouel, K. Simler (eds) (2007). More Than a Pretty Picture: Using Poverty Maps to Design Better Policies and Interventions (Washington, D.C., World Bank). Bidarbakht-Nia, A. (2009). “Empirical Analysis of Poverty and Inequality in Small Areas; Case Study of Iran”. MA Thesis, Tokyo International University (TIU). Deaton, A. (1997), The analysis of household surveys: a microeconometric approach to development policy (Washington, D.C., World Bank). Elbers, C., Lanjouw, J.O., and P. Lanjouw (2002). “Micro-Level estimation of welfare”, Policy Research Working Paper No. 2911, October, World Bank. Elbers, C., T. Fujii, P. Lanjouw, B. Özler, W. Yin (2007). “Poverty alleviation through geographic targeting: How much does disaggregation help?” Journal of Development Economics, Elsevier, May, vol. 83, Issue 1, pp. 198-213. Fields, G.S. (1998). “Accounting for income inequality and its change”, Department of Economics, Cornell University, mimeo. Fields, Gary S. (2001). Distribution and Development, A New Look at the Developing World. (Russell Sage Foundation and the MIT Press). Fields, G.S. (2003). “Accounting for income inequality and its change: a new method, with application to the distribution of earnings in the United States”. in Polachek, S.W. (Ed.), Worker Well Being and Public Policy, Research in Labor Economics, vol. 22. Elsevier, pp. 1-38, New Jersey. Foster, J., J. Greer and E. Thorbecke (1984). “A class of decomposable poverty measures”. Econometrica, May, vol. 52, Issue 3, pp. 761-766. Fujii, T. (2007). “To use or not to use?: Poverty mapping in Cambodia.” In More Than a Pretty Picture: Using Poverty Maps to Design Better Policies and Interventions, edited by T. Bedi, A. Coudouel, and K. Simler. pp. 125-142. (Washington, D.C., World Bank). Greene, W.H. (2007). Econometric Analysis (Sixth edition, New York, Prentice Hall). Haslett, S., and G. Jones (2005). “Local estimation of poverty in the Philippines”. The National Statistics Coordination Board of the Philippines and the World Bank, pp. 58. Longford, N.T. (2005). Missing Data and Small-area Estimation: Modern Analytical Equipment for the Survey Statistician (New York, Springer). Maddala, G.S. (2001). Introduction to Econometrics (New York, John Wiley & Sons, Ltd.). Management and Planning Organization (2000). Barnameh mobarezeh ba faghr (program for fighting poverty). Technical report, Management and Planning Organization, Tehran, Iran. Report (in Persian).

147


Vol. 18, No. 1, June 2011

Minot N., B. Baulch, and M. Epprecht (2006). “Poverty and inequality in Vietnam: Spatial Patterns and Geographic Determinants”. International Food Policy Research Institute, Research Report 148, Washington, D.C. Morduch, J., T. Sicular (2002). “Rethinking inequality decomposition, with evidence from rural China”. Economic Journal, vol. 112, pp. 93-106. Pajouyan, J. (1994). “Establishing the poverty line In Iran” Iran Economic Review, vol. 1(1). Tehran Univerity. Rao, J.N.K. (2003). Small Area Estimation (New York, John Wiley & Sons, Inc). Ravallion, M. (1994). Poverty Comparisons. (Chur, Harwood Academic Publishers). Ravallion, Martin (1998). ”Poverty lines in theory and practice”. Living Standards Measurement Study Working Paper 133, World Bank, Washington, D.C. Salehi-Isfahani, Djavad (2009), “Poverty, inequality, and populist politics in Iran”. Economic Inequality, Springer, March, vol. 7(1), pp. 5-28.

Journal of

Särndal, C.E. (2007). “The calibration approach in survey theory and practice”. Survey Methodology, December, vol. 33, No. 2, pp. 99-119. Suryahadi A.W., Widyanto, R.P. Artha, D. Perwira and S. Sumarto (2005). “Developing a poverty map for Indonesia: A tool for better targeting in poverty reduction and social protection programs.” SMERU Research Institute, pp. 42. Statistical Centre of Iran (2004). “Poverty line estimation in Iran 1991-2001”, paper presented at the 2004 International conference on official poverty statistics; methodology and comparability, Manila, Philippines, 4-6 October. Wan, Guanghua (2004). “Accounting for income inequality in rural China: a regression-based approach,” Journal of Comparative Economics, Elsevier, June, vol. 32, Issue 2, pp. 348-363.

148


Vol. 18, No. 1, June 2011

APPENDIX MAPS AND TABLES* Annex figure A.1 – Name and location of Provinces

Caspian Sea

24

03 04

19

02

26

12 13 05

00

23

09

20

25

15

16

28

27

01

10

29

21 06

14 17 08 18

07

Pe rs

N

n

S

f

ul

G

E

11

22

ia

W

Provinces 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 *

Markazi Gilãn Mãzandarãn East Azerbãijãn West Azerbãijãn Kermãnshãh Khuzestãn Fãrs Kermãn Khorãsãn Razavi Esfahãn Sistãn va Baluchestãn Kordestãn Hamedãn Chãharmahãl va Bakhtiãri Lorestãn Ilãm Kohgiluyeh va Boyrahmad

18 19 20 21 22 23 24 25 26 27 28 29

Bushehr Zanjãn Semnãn Yazd Hormozgãn Tehrãn Ardebil Ghom Ghazvin Golestãn North Khorãsãn South Khorãsãn

All maps are produced by author, using Arcview.

149


Annex figure A.2. Provincial poverty rate in urban

0.04 - 0.06 0.06 - 0.10 0.10 - 0.13 0.13 - 0.20 0.20 - 0.26 Lake & Sea

Annex figure A.4. Provincial poverty rate in rural areas

0.05 - 0.07 0.07 - 0.15 0.15 - 0.18 0.18 - 0.25 0.25 - 0.35 Lake & Sea

150

Vol. 18, No. 1, June 2011

Annex figure A.3. Poverty rate in urban counties

0.00 - 0.06 0.06 - 0.10 0.10 - 0.15 0.15 - 0.20 0.20 - 0.28 0.28 - 0.42 0.42 - 0.62 Lake & Sea

Annex figure A.5. Poverty rate in rural counties

0.01 - 0.07 0.07 - 0.11 0.11 - 0.15 0.15 - 0.20 0.20 - 0.26 0.26 - 0.34 0.34 - 0.53

Lake & Sea


Vol. 18, No. 1, June 2011

Annex figure A.6. Poverty density in urban areas

Annex figure A.7. Poverty density in rural areas

Annex figure A.8. Provincial Gini index in urban

Annex figure A.9. Gini index in urban counties

0.31 - 0.36 0.36 - 0.39 0.39 - 0.42 0.42 - 0.46 Lake & Sea

0.26 - 0.34 0.34 - 0.38 0.38 - 0.43 0.43 - 0.53 Lake & Sea

151


Annex figure A.10 Provincial Gini index in rural

Vol. 18, No. 1, June 2011

Annex figure A.11. Gini index in rural counties

0.26 - 0.28 0.28 - 0.36 0.36 - 0.40 0.40 - 0.47

0.22 - 0.30 0.30 - 0.36 0.36 - 0.41 0.41 - 0.50

Lake & Sea

Lake & Sea

Annex figure A.12. Clustering of urban counties

Annex figure A.13. Clustering of rural counties

1 2 3

1 2 3

Lake & Sea

Lake & Sea

152


Vol. 18, No. 1, June 2011

Annex table A.1. Common variables between census and HIES Survey Variable Name

Description

Variable Name

Description

AREA

The area of the house (only indor)

MALE_R

Proportion of males

AREA_R

Area per person

MALE15_65

Number of males aged 15 to 65

BATH

Bath room (yes, No)

MOTOR

Motor cycle (Yes, No)

CAR

Car (yes, No)

N_EMP_R

Proportion of employed members

CENTERAL_CO

Central cooling and heating system

N_LITE_R

Number of literal members

COLLEGE

Number of members attend(ed) college

N_SCH_R

Proportion of school-going members

COMPUTER

Computer (Yes, No)

N_UNEMP_R

Proportion of unemployed

COOKING_FU

Type of fuel is used for cooking

NUMBER_EMP

Number of employed members

ELECT

Electricity (Yes, No)

NUMBER_LIT


FEMALE

Number of females

NUMBER_SCH

Number of school-going members

FEMALE_R

Proportion of females

NUMBER_UNE

Number of unemployed members

FEMALE15_65

Number of females aged 15 to 65

OCCUPANCY

Ownership (Yes, No)

FM15_65_R

proportion of females aged 15 to 66

OVER64

Number of members aged over 64

GAS

Gas (yes, No)

OVER64_R

Proportion of members aged over 64

HEAD_AGE

Age of head

PIPING

Piping system (Yes, No)

HEAD_CERT

Educational certificate of Head

ROOMS

Number of rooms occupied

HEAD_EMP

Employment statues for head

ROOMS_R

Number of rooms per person

HEAD_LITE

Literacy of head

SANITATION

Sanitation system (Yes, No)

HEAD_MARI

Marriage statues of head

SIZE

Household size

HEAD_SEX

Sex of head

SPOUSE_EMP

Spouse employment (employed, else)

HEAD_STUDY

Does head currently study?

STRUCTURE

Building structure (skeleton)

HEATING_FU

Type of fuel is used for heating

TEL

Telephone (Yes, No)

INTERNET

Internet (Yes, No)

UNDER10

Number of members aged under 10

JOB_OCCUP

Job occupation of the head

UNDER10_R

Proportion of members aged under 10

KITCHEN

Kitchen (Yes, No)

UNDER6

Number of members aged under 6

M15_65_R

Proportion of males aged 15 to 65

UNDER6_R


MALE

Number of males

WATER_FUEL

Type of fuel is used for water heating

153


Vol. 18, No. 1, June 2011

Annex table A.2. List of area (rural/urban areas of each county) variables using only census information Cluster Variables

Description

CAR_CLUSTER

Proportion of HHs owning car

CERT_CLUSTER

Proportion of individuals attend(ed) college

DIS_CLUSTER

Proportion of individuals who are disable

GAS_CLUSTER

Proportion of HHs have access to gas system

ILIT_CLUTER

Proportion of individuals who are illiterate

INTER_CLUSTER

Proportion of HHs have access to internet

MIGJOB_CLUSTER

Proportion of migrants with job purposes

MIG_CLUSTER

Proportion of migrants (for any purpose)

MORT_CLUSTER

Proportion of HHs owning motor cycle

OCCUP_CLUSTER

Proportion of HHs who are owner of their house

PIP_CLUSTER

Proportion of HHs who have water piping

SEWE_CLUSTER

Proportion of HHs who have sanitation system

SIZE_CLUSTER

The average of HH size in the county

STRUC_CLUSTER

Proportion of HHs who live in the buildings with metallic or concrete skeleton

TEL_CLUSTER

Proportion of HHs who own Telephone

UNEMP_CLUSTER

Proportion of individuals who are unemployed

154


Vol. 18, No. 1, June 2011

Annex table A.3. National and regional poverty lines (Rial per person per year) Urban Code

Rural

Province Lower

Upper

Average

CV (%)

Lower

Upper

Average

CV (%)

00

Markazi

3 493 723

9 973 592

6 733 658

3.3

2 947 225

6 717 072

4 832 149

2.1

01

Gilan

4 191 186

10 851 256

7 521 221

5.7

3 404 107

8 018 060

5 711 084

3.3

02

Mazandaran

4 394 477

9 668 731

7 031 604

5.3

3 500 858

7 636 493

5 568 676

2.9

03

Eastern A.

3 128 657

8 012 694

5 570 675

2.5

2 894 098

6 383 956

4 639 027

2.9

04

Western A.

4 402 270

7 187 635

5 794 952

2.6

3 153 017

6 195 625

4 674 321

4.2

05

Kermanshah

3 319 208

6 831 652

5 075 430

3.5

2 453 928

5 271 614

3 862 771

2.9

06

Khouzestan

3 943 185

9 595 854

6 769 520

3

2 740 309

5 712 448

4 226 379

1.4

07

Fars

4 271 749

9 941 939

7 106 844

2.7

3 363 919

6 515 073

4 939 496

1.1

08

Kerman

3 418 476

8 238 429

5 828 453

2.8

2 550 666

4 814 827

3 682 747

4.3

09

Khorasan Razavi

3 355 913

8 145 405

5 750 659

3.5

2 666 419

4 922 376

3 794 397

4.2

10

Isfahan

4 417 375

12 936 896

8 677 136

6

3 161 773

8 309 821

5 735 797

6

11

Sistan & Baloochestan

2 829 017

6 135 342

4 482 180

3.7

2 310 094

3 956 047

3 133 071

1.9

12

Kordestan

3 606 900

5 931 093

4 768 996

3.8

3 366 731

5 233 150

4 299 940

2.5

13

Hamedan Chãharmahãl va

3 099 264

8 878 638

5 988 951

3

2 725 695

5 052 391

3 889 043

1.7

14

Bakhtiãri

4 091 547

8 225 512

6 158 530

3.9

3 427 235

4 586 149

4 006 692

2.8

15

Lorestan

4 353 950

7 822 248

6 088 099

4.8

3 339 866

5 071 294

4 205 580

3.2

16

Ilam

4 522 102

8 385 821

6 453 962

2.8

2 941 303

5 006 827

3 974 065

2.8

17

Kohgilooye va Boyrahmad

3 829 266

8 646 433

6 237 850

3.5

3 412 399

5 164 559

4 288 479

4.2

18

Bushehr

4 242 628

6331275

5 286 951

2.9

3 280 654

6 520 904

4 900 779

2.2

19

Zanjan

3 416 410

9 609 467

6 512 939

4.1

2 704 781

5 449 856

4 077 318

5.9

20

Semnan

4 025 705

11 217 948

7 621 826

2.8

2 817 681

5 710 407

4 264 044

3.2

21

Yazd

4 022 875

11 247 065

7 634 970

6

2 930 643

7 358 807

5 144 725

4.2

22

Hormozgan

3 845 677

11 702 243

7 773 960

2.3

2 739 670

6 161 872

4 450 771

2.2

23

Tehran

4 579 854

15 015 342

9 797 598

3.2

3 509 615

11,000,000

7 088 207

0.9

24

Ardebil

3 921 123

9 288 709

6 604 916

3.2

2 977 964

5 418 978

4 198 471

3.6

25

Ghom

3 657 864

8 828 154

6 243 009

5.5

3 036 813

6 066 223

4 551 518

3.3

26

Ghazvin

4 513 458

10 929 176

7 721 317

3.6

3 314 179

6 511 832

4 913 005

4.3

27

Golestan

3 069 159

9 194 987

6 132 073

6

2 406 588

4 612 899

3 509 744

1.2

28

North Khorasan

3 213 248

8 947 058

6 080 153

5.9

2 345 577

4 426 599

3 386 088

2.3

29

South Khorasan

3 176 720

6 409 299

4 793 009

3.2

2 591 731

4 042 683

3 317 207

1.7

Country

3 834 377

10 165 924

7 000 151

2 947 777

6 290 588

4 619 183

155


Vol. 18, No. 1, June 2011

Annex table A.4. Explanatory variables for models (7) and (10) Name AREA_R

Type Ratio

Category Asset

Expression Area per person

BATH

Dummy

Asset

Bath (yes, No)

CAR

Dummy

Asset

Car (yes, No)

CENTRAL_CO

Dummy

Asset

Central cooling and heating system (yes, No)

COMPUTER

Dummy

Asset

Computer (Yes, No)

INTERNET

Dummy

Asset

Internet (Yes, No)

KITCHEN

Dummy

Asset

Kitchen (Yes, No)

MOTOR

Dummy

Asset

Motor cycle (Yes, No)

OCCUPANCY

Dummy

Asset

House ownership (Yes, No)

ROOMS_R

Ratio

Asset

Number of rooms per person

STRUCTURE

Dummy

Asset

Building structure (skeleton) in 8 levels

FEMALE_R

Ratio

Demographic

Proportion of females

FM15_65_R

Ratio

Demographic

Proportion of females aged 15 to 65

HEAD_AGE

Other

Demographic

Age of head

HEAD_AGE2

Other

Demographic

Square of Age of head

HEAD_MARI

Dummy

Demographic

Marriage statues of head

SEX

Dummy

Demographic

Sex of head

OVER64_R

Ratio

Demographic

Proportion of members aged over 64

SIZE

Other

Demographic

Household size

SIZE2

Other

Demographic

Square of Household size square

UNDER10_R

Ratio

Demographic


UNDER6_R

Ratio

Demographic


COLLEGE

Other

Education

Number of members attend(ed) college

HEAD_CERT

Dummy

Education

Educational certificate of Head

HEAD_LITE

Dummy

Education

Literacy of head

HEAD_STUDY

Dummy

Education

Does head currently study?

N_LITE

Other

Education


N_SCH_R

Ratio

Education

Proportion school-going members

HEAD_EMP

Dummy

Employment

Employment statues for head

JOB_OCCUP

Dummy

Employment

Job occupation of head

N_EMP_R

Ratio

Employment

Proportion of employed members

N_UNEMP_R

Ratio

Employment

Proportion of unemployed members

SPOUSE_EMP

Dummy

Employment

Spouse employment (employed, else)

156


Vol. 18, No. 1, June 2011

Annex table A.4. (continued) Name

Type

Category

Expression

COOKING_FU

Dummy

Public utility

Type of fuel is used for cooking

ELECT

Dummy

Public utility

Electricity (Yes, No)

GAS

Dummy

Public utility

Gas (yes , No)

HEATING_FU

Dummy

Public utility

Type of fuel is used for heating

PIPING

Dummy

Public utility

Piping system (Yes, No)

SANITATION

Dummy

Public utility

Sanitation system (Yes , No)

TEL

Dummy

Public utility

Telephone (Yes, No)

WATER_FU

Dummy

Public utility

Type of fuel is used for water heating

COUNTY

Dummy

Location

Dummy for counties

157

158

0.18

0.15

0.24

0.15

0.22

19

20

21

22

23

0.21

0.05

0.07

17

0.2

15

16

18

0.22

0.14

12

0.12

0.35

13

0.12

10

11

14

0.18

0.15

7

0.14

0.23

6

8

0.11

5

9

0.15

2

0.14

0.06

1

3

0.18

0

4

P0

0.25

Province

0.026

0.008

0.018

0.012

0.01

0.009

0.006

0.006

0.007

0.005

0.013

0.008

0.024

0.007

0.006

0.006

0.019

0.02

0.01

0.011

0.007

0.005

0.008

0.008

se_P0

P1

0.06

0.04

0.07

0.04

0.05

0.01

0.05

0.01

0.04

0.02

0.07

0.02

0.11

0.03

0.04

0.05

0.03

0.06

0.03

0.04

0.04

0.01

0.05

0.08

P2 se_P2

0.0449 0.0051

0.0011

0.009 0.0227 0.0038

0.003 0.0183 0.0016

0.005 0.0303 0.0028

0.004 0.0174 0.0022

0.004 0.0189 0.0023

0.002 0.0025 0.0006

0.002 0.017

0.001 0.0032 0.0005

0.002 0.0108 0.0006

0.001 0.0034 0.0002

0.005 0.0278 0.0027

0.002 0.0063 0.0009

0.01

0.003 0.0142 0.0014

0.002 0.0142 0.0012

0.002 0.0171 0.0009

0.005 0.0108 0.0016

0.008 0.0259 0.0039

0.004 0.0133 0.0016

0.004 0.0176 0.0017

0.003 0.0137 0.0012

0.001 0.0032 0.0004

0.003 0.0194 0.0013

0.004 0.0365 0.0024

se_P1

Rural

0.38

0.42

0.47

0.38

0.34

0.32

0.39

0.32

0.26

0.28

0.39

0.34

0.38

0.4

0.4

0.42

0.38

0.33

0.43

0.46

0.35

0.36

0.38

0.39

0.011

0.008

0.031

0.01

0.009

0.007

0.007

0.007

0.004

0.005

0.013

0.014

0.009

0.008

0.007

0.011

0.013

0.018

0.02

0.014

0.007

0.008

0.007

0.008

Gini se_Gini

0.09

0.15

0.18

0.2

0.09

0.06

0.18

0.09

0.08

0.08

0.16

0.05

0.26

0.12

0.12

0.12

0.11

0.18

0.04

0.07

0.1

0.04

0.09

0.13

P0

0.003

0.006

0.013

0.018

0.006

0.003

0.004

0.006

0.006

0.011

0.008

0.009

0.046

0.018

0.023

0.01

0.007

0.01

0.003

0.009

0.006

0.003

0.005

0.007

se_P0

0.02

0.04

0.05

0.05

0.02

0.01

0.06

0.02

0.01

0.02

0.05

0.01

0.09

0.03

0.03

0.03

0.02

0.04

0.01

0.01

0.03

0.01

0.02

0.04

P1

0.001

0.002

0.005

0.009

0.002

0.001

0.002

0.002

0.001

0.003

0.003

0.001

0.018

0.005

0.007

0.003

0.002

0.003

0.001

0.002

0.002

0.001

0.002

0.003

se_P2 0.0014

0.0004

0.0007

0.0011

0.0017

0.0068 0.0005

0.0177 0.0011

0.0209 0.0025

0.0175 0.0047

0.0063 0.0008

0.0023 0.0002

0.0228 0.0008

0.0051 0.0006

0.0042 0.0005

0.0046 0.0013

0.022

0.0017 0.0003

0.0417 0.0094

0.0124 0.0022

0.0127 0.0028

0.0096 0.001

0.008

0.015

0.0018 0.0002

0.003

0.0113 0.0011

0.0028 0.0003

0.0093 0.0007

0.015

P2

Urban se_P1

Annex table A.5. Regional poverty and inequality indicators

0.39

0.44

0.46

0.38

0.39

0.36

0.42

0.41

0.36

0.31

0.43

0.35

0.44

0.39

0.44

0.40

0.38

0.39

0.35

0.43

0.41

0.41

0.42

0.42

Gini

0.006

0.008

0.017

0.015

0.006

0.006

0.004

0.007

0.005

0.007

0.007

0.012

0.011

0.009

0.01

0.007

0.01

0.007

0.005

0.008

0.01

0.009

0.009

0.009

se_Gini


0.15

0.21

0.17

28

29

Country

0.07

0.17

26

27

0.06

0.23

24

P0

25

Province

0.007

0.011

0.007

0.005

0.012

0.005

se_P0

0.04

0.03

0.06

0.04

0.01

0.06

0.02

P1 P2 se_P2

0.018

0.002 0.0093 0.0006

0.005 0.0231 0.0027

0.003 0.0158 0.0013

0.001 0.0045 0.0005

0.005 0.0256 0.0026

0.002 0.0053 0.0007

se_P1

Rural

0.4

0.35

0.39

0.4

0.37

0.39

0.37

0.006

0.011

0.007

0.008

0.011

0.009

Gini se_Gini

0.11

0.07

0.22

0.18

0.07

0.24

0.07

P0

Annex table A.5. (continued)

0.006

0.014

0.005

0.011

0.013

0.005

se_P0

0.03

0.01

0.06

0.05

0.01

0.05

0.01

P1

0.001

0.005

0.002

0.003

0.007

0.001

se_P1 se_P2

0.012

0.0038 0.0004

0.0244 0.0024

0.0215 0.0009

0.0033 0.0008

0.0173 0.0036

0.0041 0.0004

P2

Urban

0.42

0.39

0.44

0.41

0.41

0.38

0.36

Gini

0.008

0.011

0.007

0.025

0.013

0.006

se_Gini


159


Vol. 18, No. 1, June 2011

THE RELATIONSHIPS BETWEEN THE SOCIO-ECONOMIC PROFILE OF FARMERS AND PADDY PRODUCTIVITY IN NORTH-WEST SELANGOR, MALAYSIA Md. Mahmudul Alam, Chamhuri Siwar, Basri Talib, Mohd Ekhwan bin Toriman*

Paddy is the main food crop in Malaysia, but due to a low rate of productivity, the land area for paddy production is gradually decreasing. As a consequence, it is critically important to know the socio-economic characteristics of the paddy farmers and their linkage with agricultural productivity to ensure the sustainability of paddy farms as well as farmers’ livelihood. This study analyses the relationships between the paddy yields and the socio-economic characteristics of the farmers in the Integrated Agricultural Development Area (IADA), North-West Selangor, Malaysia. The data was collected through a survey. To analyse the data, this study conducted cross-sectional multiple OLS regressions. The findings of the study revealed that several socio-economic and physical characteristics had significant effects on paddy productivity in Malaysia. Based on the findings, some policy recommendations and action plans have been proposed focusing on paddy productivity in relation to the socioeconomic sustainability of the livelihood. The findings of the study are important for the policymakers and relevant agencies.

JEL Classification: C21, Q12, Q15, Q18. Key words: Paddy productivity, socio-economic characteristics, physical characteristics, agricultural policy, Malaysia.

* Md. Mahmudul Alam, Ph.D. Student, Institute for Environment and Development (LESTARI), Universiti Kebangsaan Malaysia, and Research Fellow at the Integrated Education and Research Foundation, Dhaka, Bangladesh, [email protected]; Chamhuri Siwar, Emeritus Professor, Institute for Environment and Development, Universiti Kebangsaan Malaysia, [email protected]; Basri Talib, Associate Professor, Faculty of Economics and Business, Universiti Kebangsaan Malaysia, [email protected]; and Mohd Ekhwan bin Toriman, Associate Professor, School of Social, Development and Environmental Studies (FSSK), Universiti Kebangsaan Malaysia, [email protected].

161


Vol. 18, No. 1, June 2011

I.

INTRODUCTION

The growing industrialization of Malaysia has meant that the agricultural sector’s share of the country’s gross domestic product (GDP) has been declining since 1975. In 1970, the contribution of agriculture to GDP was 30.8 per cent, which was the highest among all sectors. It fell to 22.7 per cent in 1975 and stood at 22.9 per cent and 20.8 per cent in 1980 and 1985, respectively, but still remained the top sector contributing to GDP. In 1990, agriculture became the second largest sector, contributing 18.7 per cent to the national GDP. In 1995, the contribution of agriculture to the national GDP further declined to 13.6 per cent, but it remained as the second largest sector in the economy. The contribution of the sector continued to decline to 8.9 per cent and 8.2 per cent in 2000 and 2005, respectively. As the agriculture sector lost its importance to the national economy, services and manufacturing sectors became the first and second highest contributing sectors, respectively, placing agriculture as the third engine of economic growth in the country. Land use for agricultural activities in Malaysia is declining due to the country’s rapid economic development. As a consequence, many land areas once used to grow crops are now used for housing, business, and industrial purposes. From 1960 until 2005, the land use for industrial crops increased while land use for food crops decreased. This implies that more and more agricultural land is being used for growing industrial crops and that the importance of growing food crops is on the decline. In terms of numbers, land use for food crops accounted for 31.5 per cent of the total agricultural land in Malaysia, compared to 16.3 per cent in 2005. Among the industrial crops, palm oil accounted for the largest share of the total land utilization in the country. Agricultural land used by the palm oil sector has significantly increased over the last five decades from only 2.1 per cent of the total amount in 1960 to 63.4 per cent in 2005. This is an indication that palm oil production has been gaining in importance and contributing significantly to the national economy. Yamada (2003) mentioned that the agriculture sector in Malaysia is characterized by a dualistic structure, consisting of large plantation companies, which are professionally managed and focus on perennial crops, such as oil palm, rubber and cocoa, and small farmers, which are generally not well managed and concentrate their activities on cultivating food crops. Rice is the main staple food in Malaysia. There are 300,000 paddy farmers in the country, of which only 40 per cent of them are full-time farmers (Man and Sadiya, 2009). According to a World Bank study cited by Pio Lopez (2007), 65 per cent of farms take up less than one hectare, and Malaysia is an inefficient producer of rice. The study noted that the price of locally produced rice was double that of imported rice. It was estimated that 74 per cent of paddy producers’ monthly income came

162


Vol. 18, No. 1, June 2011

from income support measures, suggesting that the Malaysian paddy subsector is non-viable and non-sustainable. Government support for research and development, production and marketing in this subsector have taken many forms. Credit facilities, fertilizer subsidies, irrigation investment, guaranteed minimum price, income support programmes, subsidized retail price as well as research and extension support (training and advisory services), amounting to billions of dollars over the past 50 years, have been a fiscal drain on the nation. Despite the massive fiscal outlays, rice production remains chronically inefficient with regards to meeting market demand. Given the continued decline in cultivated area, negligible gains in productivity, continued increases in the cost of production and decreasing profitability, rice production in Malaysia can be considered a sunset industry (Pio Lopez, 2007). The current record also shows a negative trend of land usage for paddy production. There are a total of 426,260 hectares of paddy planted area, and average yield is 3.5 tons per hectare (Malaysia, 2008). Singh and others (1996) pointed out that the actual farm yields of rice in Malaysia varied from 3 to 5 tons per hectare, where potential yield is 7.2 tons. Pio Lopez (2007) noted that rice production in Malaysia was going to end due to the continued decline in the cultivated area, negligible gains in productivity, continued increases in the cost of production and decreasing profitability. Jayawardane (1996) indicated that 90 per cent of total paddy productivity is contributed by labour, farm power, fertilizer and agro-chemicals, out of which 45 per cent is contributed by labour. The Third Malaysia Plan (Malaysia, 1976) reported that the incidence of poverty was 88 per cent among the rice farmers. The National Response Strategies (Malaysia, 2001) also reported that the poor are the most vulnerable population group to climate change and that the hardcore poor tended to have a relatively large number of household members and are involved in agricultural activities. While working in the Muda Agricultural Development Authority (MADA) area, Corner (1981) observed that off-farm employment should be expanded as part of an anti-poverty strategy. Shand and Chew (1983) conducted research in Kelantan, Malaysia and found that a large majority of farmers had relied heavily on off-farm employment in order to attain a modest standard of living. An assessment of the socio-economic profile of farmers in Selangor, Malaysia, indicated that 81 per cent of the farmers were between 20 and 60 years old, and that 84 per cent of them had at a least primary education (TaniNet 2nd Report, 2000). To remedy the above, elements of the Second National Agricultural Policy 1992-1997 (Malaysia, 1992) were revised for the Third National Agricultural Policy 1998-2010 (Malaysia, 1998), which envisioned the sustainable development of

163


Vol. 18, No. 1, June 2011

a dynamic agricultural sector focusing on the market-led, commercialized, efficient and competitive growth of agriculture. The principal aim of the Third National Agricultural Policy was to maximize the income of the stakeholders through optimal utilization of resources. In addition to actual production, several socio-economic characteristics as well as the physical assets must be accounted for when measuring productivity pertaining to agricultural activities. Moreover, variables that may affect productivity, such as climate change, must also be considered by officials when setting agricultural policies. The purpose of this paper is to analyse the effects of the socio-economic characteristics of farmers on the productivity of the paddy and to discuss the relevant policy implications.

II.

BACKGROUND OF THE STUDY AREA

The agricultural history of the study area started before 1930. An irrigation controlled area was built at Pancang Bedena in 1932 and 15,000 acres of land was converted to a rice field, with planting starting in this area in 1936. The yield was very low at that time due to a water shortage, and, so, a barrage at the Tengi River was built the same year. Since 1948, a number of concrete waterways have been built in the area. In 1962, double planting of paddy per year was introduced in the area and a pump house was built in Began Terap to provide sufficient water in remote land areas in the south. By 1966, all of the paddy areas in the study area were being planted twice yearly. In 1978, Projek Barat Laut Selangor (PBLS) was established to help boost productivity, upgrade and improve the infrastructure of the agricultural plots, and look after the welfare of the farmers. In 1982, planting by direct seeding was introduced in this area. Before it was changed to “North West Selangor Integrated Agricultural Development Project” (IADP), it was called “Tanjung Karang Drainage Scheme”. In 1985 IADP was completed at a cost of $87 million, financed through a World Bank loan. This project focused on improving irrigation infrastructure for rice intensification. Recently, the name of IADP has been changed to “Integrated Agricultural Development Area” (IADA). Now the IADA in North-West Selangor consists of eight areas – Sawah Sempadan, Sg. Burong, Sekinchan, Sg. Leman, Pasir Panjang, Sg. Nipah, Panchang Bedena, and Bagan Terap. The total agricultural land area of the IADA is 100,000 hectare (ha), with 55,000 ha used for palm oil, 20,000 ha for coconut, 5,000 ha for fruits and vegetables and 20,000 ha for paddy. This 20,000 ha for the paddy area contains a river, a drainage system, and rural roads. The total paddy producible area is 18,638 ha, and currently paddy is planted on 18,355 ha. Total paddy irrigated area is 18,980 ha, with the additional 625 ha being used for drainage. There are 10,300

164


Vol. 18, No. 1, June 2011

paddy farmers and 30,000 producers of other crops. The total size of the agricultural community is 50,000.

III.

DATA AND METHODOLOGY

As mentioned above, the productivity of crops is determined by many socioeconomic and physical factors. To ascertain the effects of these factors, this study mostly relied on primary data from a research project entitled “The economics of climate change: Economic dimensions of climate change, impacts and adaptation practices in agriculture sector: Case of paddy sector in Malaysia”, conducted by the Institute for Environment and Development (LESTARI) of the National University of Malaysia (UKM) (Alam and others, 2010a). Data for this project were collected through a sample survey of paddy producing farmers in the eight sections of the Integrated Agricultural Development Area (IADA) of North-West Selangor, Malaysia. The target group of the survey was paddy producing farmers. The survey was a structured questionnaire taken by regular enumerators of the IADA authority under the direct supervision of IADA officials. The population size of the area was 10,300 while the sample for the study consisted of 198 respondents proportionately distributed among the eight areas based on the size of the irrigated land area. The 198 households covered 577.53 ha of paddy areas. The sample within the area is selected randomly. Details of socio-economic profiles of the respondents are available at Alam and others (2010a, 2010b). While analysing the relationship between socio-economic characteristics of farmers and agricultural yield, this study conducted cross sectional-multiple OLS regressions based on the survey data. To avoid the problem of heteroskedasticity, further regression was conducted by the “White Heteroskedasticity – Consistent Standard Errors and Covariance” method by using EViews econometric software. The econometric model used in this study avoided the direct inputs variables that determine the crop yield, such as labour hours, use of pesticides, use of fertilizers and use of irrigation. Instead, it assumes that all of the direct input factors are exogenous and constant based on the view that the government provides these facilities in fixed quantities, which very few farmers exceed. In this model, factors relevant to socio-economic characteristics of farmers (W1- W11), physical characteristics of the farm (W12- W22), and farmers’ knowledge about relevant current issues, such as climate change (W23), are considered to determine its relationship with the yield of paddy production.

165


Vol. 18, No. 1, June 2011

Pi = α + β1W1i + β2W2i + β3W3i + β4W4i + β5W5i + β6W6i + β7W7i + β8W8i + β9W9i + β10W10i + β11W11i + β12W12i + β13W13i + β14W14i + β15W15i + β16W16i + β17W17i + β18W18i + β19W19i + β20W20i + β21W21i + β22W22i + β23W23i+ε Here, Pi = Yield of paddy production W1i = Race of the farmer (dummy variable where Malay = 1, Other = 0) W2i = Gender of farmer (dummy variable where Male = 1, Other = 0) W3i = Age of the farmer W4i = Secondary occupation (dummy variable where Having secondary occupation = 1, No secondary occupation = 0) W5i = Tertiary education (dummy variable where Having tertiary education = 1, Other = 0) W6i = Primary education (dummy variable where Having primary education = 1, Other = 0) W7i = Secondary education (dummy variable where Having secondary education = 1, Other = 0) W8i = Family size (number of family members) W9i = Non-agriculture to agricultural income ratio W10i = Number of available vehicle W11i = Distance between home and field W12i = Farm size (size of paddy cultivated area) W13i= Mixed ownership of land (dummy variable where Both own and rent land Use = 1, Other = 0) W14i = Own land (dummy variable where Own land use = 1, Other = 0) W15i = Number of available machineries W16i = Locality Sawah Simpadan (dummy variable where Sawah Simpadan = 1, Other = 0) W17i = Locality Sekinchan (dummy variable where Sekinchan = 1, Other = 0)

166


Vol. 18, No. 1, June 2011

W18i = Locality Sg. Nipah (dummy variable where Sg. Nipah = 1, Other = 0) W19i = Locality Bagan Terap (dummy variable where Bagan Terap = 1, Other = 0) W20i= Locality Sg. Burong (dummy variable where Sg. Burong = 1, Other = 0) W21i = Locality Pasir Panjang (dummy variable where Pasir Panjang = 1, Other = 0) W22i = Locality Sg. Leman (dummy variable where Sg. Leman = 1, Other = 0) W23i = Know about current issues (such as climate change) α = Constant β = Coefficient of respective explanatory variable ε = Residual

IV.

RESULTS AND DISCUSSIONS

The multiple regression based on the socio-economic characteristics of farmers on yield of paddy shows mixed results. The results differ based on normal output and heteroskedasticity consistent standard errors and covariance (the White Test for heteroskedasticity) output. For reporting purposes both data are reported, but for analysis and drawing inferences only the heteroskedasticity adjusted results are considered here. In both cases, the R2 is 0.44, which means that the independent variables explain 44 per cent of the dependent variable. As explained earlier, this model assumes the main input variables that determine yields of paddy are exogenous due to an almost fixed value for all observations. As these exogenous variables highly determine the dependent variable, excluding them in the model would result in a moderate value of R2. The demographic data, including, among other factors, ethnicity, gender and age, are considered in terms of labour quality for the yield of production. The result shows that race has significant impacts on productivity. In the dummy variable, Malay farmers are classified as 1 and Chinese farmers are classified as 0. The coefficient of the regression result shows a significant negative relationship, a 1 per cent significance level, between the race of the farmers and productivity, which means that the productivity of Malay farmers is lower than that of Chinese farmers. The elasticity shows -0.4 for Malay farmers with regard to paddy yield. In terms of

167


Vol. 18, No. 1, June 2011

Table 1. Descriptive statistics Variable Pi

Mean

Standard deviation

N

6.95

2.169

198

W1i

.90

.302

198

W2i

.90

.295

198

W3i

52.93

11.629

198

W4i

.16

.364

198

W5i

.04

.197

198

W6i

.47

.501

198

W7i

.43

.496

198

W8i

4.74

1.972

198

W9i

2.92

2.681

198

W10i

.30

.459

198

W11i

.53

.500

198

W12i

2.97

2.071

198

W13i

.13

.333

198

W14i

.10

.295

198

W15i

.10

.302

198

W16i

.13

.333

198

W17i

.16

.369

198

W18i

.18

.382

198

W19i

.12

.327

198

W20i

.56

3.082

198

W21i

.59

.493

198

W22i

1.42

4.232

198

W23i

3.62

1.921

198

age and gender, the quality of labour does not have a significant impact on paddy yield. The availability of secondary occupations for the farmers has a significant negative impact, a 7 per cent significance level, on paddy yield. For a 1 per cent increase in secondary occupations among farmers, the yield of paddy production decreases by 0.02 per cent. This variable indicates that the quality of labour input, in terms of concentration or effort and managing enough time for the agricultural purposes, is required for increasing the yield.

168


Vol. 18, No. 1, June 2011

Table 2. Regression output for paddy yield on farmers’ socio-economic characteristics Normal value

Variable Coefficient α

t-stat

P-value

Heteroskedasticity consistent value Elasticity Coefficient

t-stat

P-value

Elasticity

8.93*

6.12

0.00

8.93*

6.12

0.00

W1i

-3.12*

-5.00

0.00

-0.40

-3.12*

-5.00

0.00

-0.40

W2i

-0.36

-0.69

0.49

-0.05

-0.36

-0.69

0.49

-0.05

W3i

-0.01

-0.59

0.56

-0.06

-0.01

-0.59

0.56

-0.06

W4i

-0.68***

-1.84

0.07

-0.02

-0.68***

-1.84

0.07

-0.02

1.56

0.12

0.01

1.38

1.56

0.12

0.01

W5i

1.38

W6i

1.08***

1.75

0.08

0.07

1.08***

1.75

0.08

0.07

W7i

1.29**

2.11

0.04

0.08

1.29**

2.11

0.04

0.08

W8i

0.04

0.54

0.59

0.03

0.04

0.54

0.59

0.03

W9i

0.06

1.09

0.28

0.00

-0.09**

-2.16

0.03

-0.01

W10i

0.42

1.15

0.25

0.22

-0.11

-1.25

0.21

-0.06

W11i

-0.20

-0.54

0.59

-0.04

-0.03

-1.48

0.14

-0.01

2.07

0.04

0.06

0.06

1.09

0.28

0.02

W12i

0.14**

W13i

-0.20

-0.40

0.69

-0.01

0.42

1.15

0.25

0.02

W14i

0.44

0.53

0.59

0.03

-0.20

-0.54

0.59

-0.02

W15i

0.63

1.60

0.11

0.27

2.07

0.04

0.06

W16i

-0.21

-0.34

0.73

0.00

-0.20

0.14**

-0.40

0.69

0.00

W17i

0.88***

1.82

0.07

0.01

0.44

0.53

0.59

0.01

W18i

0.58

0.96

0.34

0.01

0.63

1.60

0.11

0.01

-0.21

W19i

0.01

0.01

0.99

0.00

-0.34

0.73

0.00

W20i

-0.09**

-2.16

0.03

0.00

0.88***

1.82

0.07

0.02

W21i

0.04

0.13

0.90

0.00

0.58

0.96

0.34

0.01

W22i

-0.03

-1.48

0.14

0.00

0.01

0.01

0.99

0.00

W23i

-0.11

-1.25

0.21

-0.01

0.04

0.13

0.90

0.00

Note:

´ ´ Wji ´ = mean of respective Wi; Pji ´ = mean of dependent variable Pi; Elasticity = βj* (Wji/Pji); j = respective independent variable. *, **, and *** are significant at the 1%, 5% and 10% significance level, respectively.

The education level of the farmer plays a major role in the yield of agricultural output. If a farmer has a tertiary education, there is no significant impact on the paddy yield, but for those with primary and secondary educations, the impacts on paddy yield are significant, with significance levels of 8 per cent and 4 per cent, respectively. Farmers need a basic level of education to understand and read relevant

169


Vol. 18, No. 1, June 2011

news, rules and notices, which can affect productivity significantly. The elasticity of primary and secondary educated farmers on paddy yield is 0.07 and 0.08, respectively. This means that a 1 per cent increase in primary level-educated farmers leads to a 0.07 per cent increase in paddy yield, and a 1 per cent increase in secondary level-educated farmers leads to 0.08 per cent increase in paddy yield. The size of a family or number of family members in terms of quantity of labour has no significant impact on paddy yields. This finding is consistent with the initial assumption that labour is an exogenous factor in the model, meaning that the same amount of labour is provided by all farmers. Consequently, increased labour inputs provided by a large family have no significant effects on yield. Moreover, a small family tends to cover required labour needs by hiring workers and thus the impact of labour is on profitability rather than on productivity. The ratio of non-agriculture to agricultural income has significant negative impacts, at 3 per cent significance level, on paddy yields. If farmers have more nonagricultural income compared to agricultural income, the productivity of paddy declines due to the less attention farmers provide to agricultural work. If the ratio increases by 1 per cent, the yield of paddy productivity declines by 0.01 per cent. The distance between the paddy field and home and the number of available vehicles has no significant impact on paddy yield. Farmers usually live very close to their fields, and have enough vehicles on hand to reach the fields when necessary. As a consequence, these factors have no significant impacts on paddy yield. The farm level production or individual farmer level production has no significant impacts on productivity. Due to the use of almost constant rates of inputs by farmers, provided by government which is considered as exogenous variables in the model, the yields of farms or individual farmers have no significant difference. The ownership of the land also has no significant impacts on yield of paddy production. However, both the farm size and ownership criteria have impacts on the profitability rate as opposed to the productivity rate. The available number of machines for carrying out agricultural activities has significant impacts on yield productivity of paddy. This is an indicator of using technology and timing of technology usage. At a 4 per cent significant level, it shows that the total number of machines available to the farmer or farm has significant impacts on paddy yield. A 1 per cent increase in the number of machines leads to a 0.06 per cent increase in the productivity of the paddy. The productivity of paddy among the eight areas of IADA varies. Some areas of the IADA have statistically significant impacts on the yield. One of them is

170


Vol. 18, No. 1, June 2011

Sg. Burong which is significant at 7 per cent. The area of Sekinchan is not statistically significant, but the productivity of this area is very high. The variation in crops from area to area results from the variation in land fertility and climate-related factors, such as the land moisture level, the rate of rainfall and temperature rates. The knowledge of farmers about recent issues, such as awareness about climate change, has no significant impacts on yield of paddy, because this information is readily available to everybody. Moreover, the IADA authority provides the information, controls and monitors all relevant issues of paddy production in this locality. Several socio-economic variables of the farming community that have statistically significant relationships with paddy yield change due to declining paddy productivity. Farmers with secondary occupations have negative relationships with yield. When yields decline, the secondary occupations of farmers increase further. Moreover, education has a positive relationship with yield, as yield decreases, the number of educated farmers declines. In addition, the ratio of non-agricultural income to agricultural income will increase more in the future as yields decline due to climate change. The socio-economic characteristics of the farming community are changing and will change even more depending on the intensity of the adverse effects stemming from climate change. Physical characteristics of farms that have significant impacts on paddy yield carry important implications to commercial farms, for instance, will likely benefit more from the positive relationship between machinery and productivity than individual farmers because they are in a better position to purchase new machines. Furthermore, few localities have positive relationships with yields that lead to an increase in the demand for land in these areas. Due to differentiations in paddy productivity among geographical areas, income inequality among the farmers in different areas is increasing.

V.

CONCLUSIONS AND RECOMMENDATIONS

As shown in the study, several socio-economic and physical characteristics of farmers have significant impacts on paddy productivity. One characteristic in particular is race. The productivity of Malay farmers compared with Chinese farmers is significantly low, with the elasticity at -0.4 in terms of paddy yield between the two groups. The option of pursuing a secondary occupation for farmers has significant negative impacts on paddy yield. If the number of farmers pursuing secondary occupations increases 1 per cent, the yield of paddy production decreases by 0.02 per cent. The level of education also has significant impacts on paddy yield.

171


Vol. 18, No. 1, June 2011

A 1 per cent increase in farmers having a primary level education results in a 0.07 per cent increase in paddy yield, and 1 per cent increase in farmers having a secondary level education raises paddy yield 0.08 per cent. The ratio of non-agriculture to agricultural income has significant negative impacts on paddy yield. If the ratio increases by 1 per cent, the yield of paddy productivity declines by 0.01 per cent. Moreover, examples of physical farm characteristics that have significant impacts on paddy yield are technology and geographical position. A 1 per cent increase in the number of farm machines leads to a 0.06 per cent increase in the productivity of the paddy. Among the areas, Sg. Burong shows positive impacts on yield where the elasticity value is 0.02. Based on the findings, there are several policy implications. Currently, the Government allocates a certain amount of land for Malay farmers even though their productivity is lower than that of other ethnic groups. It is also a threat for achieving self-sufficiency level of paddy production at the national level. At the same time, due to the high productivity of Chinese farmers, allotting more land to them will create a social imbalance. Therefore, the Government should initiate policies that aim to help Malay farmers boost their productivity through specific training or education programmes, awareness creation programmes or extra incentive programmes. Production practices are an important element in government policy. Currently, the Government specifies certain areas for paddy production. This approach inhibits farmers who tend to select the crops they produce of their own choosing and results in higher land degradation due to mono crop production. In this area, based on soil suitability, crop rotation and crop variety are needed to maintain land fertility and reduce the risk of climate change. The Government should also pay more attention to other relevant factors, such as agricultural wages, land leasing systems and rates and maximum farm size. These factors are very important for the sustainability of small farmers, poverty reduction and reducing income inequality and a positive effort directed at them would increase overall productivity to achieve self-sufficiency or close to the self-sufficiency level in rice production to ensure national food security.

172


Vol. 18, No. 1, June 2011

REFERENCES Alam, M.M., Chamhuri Siwar and M.E. Toriman (2010a). Socioeconomic Study of Climate Change: An Assessment of Agriculture and Livelihood Sustainability on Paddy Farming in Malaysia. LAP Lambert Academic Publishing: Saarbrucken. Alam, M.M., Chamhuri Siwar, Wahid Murad, R.I. Molla and M.E. Toriman (2010b). “Socioeconomic profile of farmer in Malaysia: study on Integrated Agricultural Development Area in NorthWest Selangor”, Agricultural Economics and Rural Development, vol. 7, No. 2, pp. 249-26, 2010. available at http://ideas.repec.org/s/iag/reviea.html. Corner, L. (1981). “The impact of rural outmigration: labor supply and cultivation techniques in a double cropped padi area, West Malaysia”, Ph.D. thesis, Macquarie University, Sydney. Jayawardane, S.N. (1996). “Socio-economic constraints and future prospects for crop diversification in minor irrigation schemes”, workshop on crop diversification, Colombo. Malaysia (2008). Agriculture Statistical Handbook, “Paddy”. Ministry of Agriculture. Malaysia (2001). National Response Strategies to Climate Change, Ministry of Science, Technology and the Environment. Malaysia (1998). Third National Agricultural Policy (1998-2010), Ministry of Agriculture. Malaysia (1992). Second National Agricultural Policy (1992-1997), Ministry of Agriculture. Malaysia (1976). Third Malaysia Plan (1976-1980). Department.

Economic Planning Unit, Prime Minister’s

Man, N. and Sadiya, S.I. (2009). “Off-farm employment participation among paddy farmers in the MUDA Agricultural Development Authority and Kemasin Semerak Granary areas of Malaysia”. Asia-Pacific Development Journal, vol. 16, No. 2, pp. 141-153. Pio Lopez, G. (2007). “Economic reforms for paddy sub-sector”, The Star Online, 25 June, available at http://biz.thestar.com.my/news/story.asp?file=/2007/6/25/business/18087959&sec= business. Shand, R.T. and T.A. Chew (1983). “Off farm employment in the Kemubu Project in Kelantan, Malaysia”. Presented at a conference in Chiang-Mai, Thailand, 23-26 August. Available at http://www.cababstractsplus.org/abstracts/Abstract.aspx?AcNo=19876704459. Singh, S., R. Amartalingam, W.S. Wan Harun and M.T. Islam (1996). “Simulated impact of climate change on rice production in Peninsular Malaysia”. Proceeding of National Conference on Climate Change. pp. 41-49, UPM, Malaysia. TaniNet 2nd Report (2000). Second Quarter, DAGS Report, TaniNet Project, UKM-MTDC Bangi, Selangor, Malaysia. Yamada, Saburo (2003). “Regional Survey Report On Agricultural Productivity Index”. Asian Productivity Organization, Japan. Available at http://www.apo-tokyo.org/projreps_acd/ 15_03-AG-GE-SYP-01-B.pdf.

173


Vol. 18, No. 1, June 2011

SUBSCRIPTION FORM (Please type or print) NAME:

_______________________________________________________________________________

POSITION:

_______________________________________________________________________________

ORGANIZATION:

_______________________________________________________________________________

ADDRESS:

_______________________________________________________________________________ _______________________________________________________________________________ _______________________________________________________________________________ _______________________________________________________________________________ COUNTRY: _________________________________ POSTCODE: ______________________

TELEPHONE: ______________________________ FACSIMILE: ___________________ E-MAIL: ___________________ ---------------------------------------------------------------------------------------------------------------------------------SUBSCRIPTION RATES FOR ASIA-PACIFIC DEVELOPMENT JOURNAL (2 ISSUES PER YEAR)

❏ ❏

1 year

US$ 66.00

3 years

US$ 198.00

Please mail this form together with your subscription fee in US dollars in the form of a bank draft/cheque drawn in a bank in the United States, issued in the name of UN-ESCAP, crossed and marked “Account Payee Only”. The draft/cheque may be mailed to the following address: Chief, Conference Management Unit Administrative Services Division Economic and Social Commission for Asia and the Pacific (ESCAP) United Nations Building, Rajadamnern Nok Avenue Bangkok 10200, Thailand

✁ ----------------------------------------------------------------------------------------------------------------------------This publication may be obtained from bookstores and distributors throughout the world. Please consult your bookstore or write to any of the following: Sales Section Room DC2-0853 United Nations Secretariat New York, NY 10017 United States of America Sales Section United Nations Office at Geneva Palais des Nations CH-1211 Geneva 10 Switzerland Chief, Conference Management Unit Administrative Services Division Economic and Social Commission for Asia and the Pacific (ESCAP) United Nations Building, Rajadamnern Nok Avenue Bangkok 10200, Thailand

Tel.: (212) 963-8302 Fax: (212) 963-4116 Telex: 422311 UN UI

Tel.: (41) (22) 917-1234 Fax: (41) (22) 917-0123 Telex: 23711 ONU CH

Tel.: (662) 288-1234 Fax: (662) 288-3022 Telex: 82392 ESCAP TH


Vol. 18, No. 1, June 2011

READERSHIP SURVEY The Macroeconomic Policy and Development Division of ESCAP is undertaking an evaluation of the publication Asia-Pacific Development Journal, with a view to improving the usefulness of future publications to our readers. We would appreciate it if you could complete this questionnaire and return it, at your earliest convenience, to: Director Macroeconomic Policy and Development Division ESCAP, United Nations Building Rajadamnern Nok Avenue Bangkok 10200, THAILAND E-mail: [email protected]

QUESTIONNAIRE Rating for quality and usefulness (please circle) 1.

● ● ● ● ●

Average

Poor

presentation/format readability timeliness of information coverage of subject matter analytical rigour overall quality

4 4 4 4 4 4

3 3 3 3 3 3

2 2 2 2 2 2

1 1 1 1 1 1

4 4 4 4 4

3 3 3 3 3

2 2 2 2 2

1 1 1 1 1

How useful is the publication to your work? ● ● ● ● ●

3.

Very good

Please indicate your assessment of the quality of the publication in terms of: ●

2.

Excellent

provision of information clarification of issues its findings policy suggestions overall usefulness

Please give examples of how this publication has contributed to your work: ....................................................................................................................................................... ....................................................................................................................................................... ....................................................................................................................................................... .......................................................................................................................................................

......................................................................................................................................


4.

Vol. 18, No. 1, June 2011

Suggestions for improvement of similar publications: ....................................................................................................................................... ....................................................................................................................................... ....................................................................................................................................... .......................................................................................................................................

5.

Your background information, please: Name: ............................................................................................................................ Title/position: ................................................................................................................ Institution: ...................................................................................................................... Office address: .............................................................................................................. .......................................................................................................................................

Please use additional sheets of paper, if required, to answer the questions. Thank you for your kind cooperation in completing this questionnaire.


Vol. 18, No. 1, June 2011

ASIA-PACIFIC DEVELOPMENT JOURNAL INSTRUCTIONS TO CONTRIBUTORS Published by the Macroeconomic Policy and Development Division of the United Nations Economic and Social Commission for Asia and the Pacific, the Asia-Pacific Development Journal provides a platform for the exchange of ideas and experiences on development issues and concerns facing the region, and aims to stimulate policy debate and assist policy formulation. Published twice a year, the Journal welcomes policy-oriented articles and original pieces of work, focusing on development issues and challenges relevant to the Asian and Pacific region. 1.

MANUSCRIPTS Authors are requested to provide copies of their manuscripts in English. Contributors should indicate in their covering letter to the Editorial Board that the material has not been previously published or submitted for publication elsewhere. The manuscripts should be typed, double-spaced, on one side of white A4 paper and the length should not exceed 30 pages. Manuscripts are accepted subject to editorial revision. Since all manuscripts will be refereed by professionals in the field, the name(s) of the author(s), institutional affiliation(s) and other identifying information should be placed on the title page only, in order to preserve anonymity. The title page should contain the following: (a) title; (b) name(s) of the author(s); (c) institutional affiliation(s); and (d) complete mailing address, telephone number, facsimile number and e-mail address of the author, or of the primary author in the case of joint authors, and (e) JEL classification and keywords relevant to the article. The second page should contain the title, the name(s) of the author(s) and an abstract of approximately 150 words. Acknowledgement (if any) should appear after the abstract. It is preferred that manuscripts be submitted by e-mail to the address below (if hard copies are submitted, kindly provide two copies of the manuscript to the address below). The preferred word-processing software is MSWord. Once a manuscript is accepted for publication, the author(s) may be asked to submit electronic files of their manuscript, figures, tables and charts, as appropriate. 2.

FOOTNOTES AND QUOTATIONS Footnotes, if any, should be numbered consecutively with superscript Arabic numerals. They should be typed single-spaced and placed at the bottom of each page. Footnotes should not be used solely for citing references. Quotations should be double-spaced. A copy of the page(s) of the original source of the quotation, as well as a copy of the cover page of that source, should be provided. 3.

TABLES AND FIGURES All tables and figures should be numbered consecutively with Arabic numerals. Each table should be typed double-spaced. Tables and figures should be planned to fit the proportions of the printed page. Full information on the source(s) should appear below the table/figure, followed by notes, if any, in lowercase letters. 4.

REFERENCES

Authors should ensure that there is a complete reference for every citation in the text. References in the text should follow the author-date format, followed, if necessary, by page numbers, for example, Becker (1964: 13-24). List only those references that are actually cited in the text or footnotes. References, listed alphabetically, should be typed double-spaced on a separate page in the following style: Ahmed, E. and N. Stern (1983). University of Warwick.

“Effective taxes and tax reform in India”, Discussion Paper 25,


Vol. 18, No. 1, June 2011

Desai, Padma, ed. (1883). Marxism, Central Planning, and the Soviet Economy (Cambridge, MA, MIT Press). Krueger, Alan B. and Lawrence H. Summers (1987). “Reflections on the inter-industry wage structure”, in Kevin Lang and Jonathan S. Leonard, eds., Unemployment and the Structure of Labour Markets (London, Basis Blackwell). Sadorsky, P. (1994). “The behaviour of U.S. tariff rates: comment”, American Economic Review, vol. 84, No. 4, September, pp. 1097-1103. Terrones, M. (1987). “Macroeconomic policy cycle under alternative electoral structures: a signaling approach”, unpublished. For further details on referencing, please refer to the editorial guidelines at: www.unescap.org/ pdd/publications/questionnaire/apdj_editorial_guidelines.pdf. The Editorial Board of the Asia-Pacific Development Journal wish to emphasize that papers need to be thoroughly edited in terms of the English language and authors are kindly requested to submit manuscripts that strictly conform to the attached editorial guidelines. Manuscripts should be sent to: Chief Editor, Asia-Pacific Development Journal Macroeconomic Policy and Development Division United Nations Economic and Social Commission for Asia and the Pacific United Nations Building, Rajadamnern Nok Avenue, Bangkok 10200, Thailand. Tel: (662) 288 1902; Fax: (662) 288 1000; (662) 288-3007; E-mail: [email protected] Cable: ESCAP Bangkok

United Nations Economic and Social Commission for Asia and the Pacific Macroeconomic Policy and Development Division United Nations Building, Rajadamnern Nok Avenue Bangkok 10200, Thailand Fax: (662) 288-3007 E-mail: [email protected] [email protected] Website: www.unescap.org

United Nations publication Sales No. E.11.II.F.7 Copyright © United Nations 2011 ISBN: 978-92-1-120625-8 ISSN: 1020-1246 ST/ESCAP/2599

The Asia-Pacific Development Journal (APDJ) is published twice a year by the Macroeconomic Policy and Development Division of the United Nations Economic and Social Commission for Asia and the Pacific. The primary objective of the APDJ is to provide a platform for the exchange of knowledge, experience, ideas, information and data on all aspects of economic and social development issues and concerns facing the region and aims to stimulate policy debate and assist policy formulation. The Asian development experience has stood out as an extraordinary example of what can be achieved when policymakers, experts, scholars and people at large harness their creativity, knowledge and foresight. The APDJ has been a proud partner in this process, providing a scholarly means for bringing together research work by eminent social scientists and development practitioners from the region and beyond for use by a variety of stakeholders. Over the years, the Journal has emerged as a key United Nations publication in telling the Asian development story in a concise, coherent and impartial manner to stimulate policy debate and assist policy formulation in the region.

USD $33 ISBN 978-92-1-120625-8

United Nations Publication Printed in Bangkok October 2011 – 1,000