Africa Region Working Paper Series No. 95
South Africa’s Export Performance: Determinants of Export supply
by
Lawrence Edwards and Phil Alves
School of Economics
University of Cape Town
DECEMBER 2005
Africa Region Working Paper Series No. 95
December 2005
Abstract
This paper is a result of a wider policy research and knowledge work on growth and jobs issues in South Africa, which the Bank promotes in collaboration with leading South African researchers. The objective is to contribute to major economic and policy issues facing South Africa as it embarks on the second decade of its remarkable democratic transition. These issues include growth and jobs, export competitiveness, service delivery, small and medium-size enterprise development and investment climate, industrial concentration, infrastructure and growth, municipal and financial management, land reform, regional integration, trade and poverty, HIV/AIDS, and––especially important––service delivery. The paper provides three principal results. First, it evaluates the extent to which the composition and level of manufacturing exports have responded to reform initiatives in the 1990s and finds that the successes of these policies in generating export growth have been mixed; the inability to re-structure exports towards dynamic, high-technology products is one explanation for the relatively poor export performance of South African manufacturing during the 1990s. Second, the paper investigates the determinants of South African manufacturing export performance using estimated export supply and demand functions; it shows that South African manufacturers are on average price-takers in the international market and that exports are predominantly supply driven. And third, the paper finds that export growth is constrained by factors that affect the profitability of exports; real effective exchange rate, infrastructure costs, tariff rates and skilled labour are found to be important determinants of export supply.
The Africa Region Working Paper Series expedites dissemination of applied research and policy studies with potential for improving economic performance and social conditions in Sub-Saharan Africa. The series publishes papers at preliminary stages to stimulate timely discussions within the Region and among client countries, donors, and the policy research community. The editorial board for the series consists of representatives from professional families appointed by the Region’s Sector Directors. For additional information, please contact Momar Gueye, (82220), Email:
[email protected] or visit the Web Site: http://www.worldbank.org/afr/wps/index.htm. The findings, interpretations, and conclusions in this paper are those of the authors. They do not necessarily represent the views of the World Bank, its Executive Directors, or the countries that they represent and should not be attributed to them.
Authors’Affiliation and Sponsorship
Authors: (1) Lawrence Edwards, Professor of Economics, School of Economics, Cape Town University, South Africa, Consultant, The World Bank. (2) Phil Alves, School of Economics, Cape Town University, South Africa, Consultant, The World Bank.
Sponsor and Editor: Željko Bogetić, Lead Economist, AFTP1, The World Bank.
[email protected]
FOREWORD This paper is a result of a wider policy research and knowledge work on growth and jobs issues in South Africa, which the Bank promotes in collaboration with leading South African researchers. The objective is to contribute to major economic and policy issues facing South Africa as it embarks on the second decade of its remarkable democratic transition. These issues include growth and jobs, export competitiveness, service delivery, small and medium-size enterprise development and investment climate, industrial concentration, infrastructure and growth, municipal and financial management, land reform, regional integration, trade and poverty, HIV/AIDS, and––especially important––service delivery. It is hoped that dissemination of papers such as this will contribute to a wider exchange of ideas on policy and development experiences both within South Africa and across the African countries. Such knowledge work is key to understanding complex development issues and dilemmas confronting the policymakers. It is also a necessary ingredient in promoting sound policies and economic growth in the region. Ritva Reinikka Country Director Botswana, Lesotho, Namibia, South Africa, Swaziland The World Bank
2
South Africa’s Export Performance: Determinants of Export supply By Lawrence Edwards and Phil Alves*
1. Introduction In 1994, the new democratically elected government inherited an economic system characterised by declining economic and employment growth. In response to these pressures, the government initiated a number of policy reforms to stimulate growth, employment and redistribution. The macroeconomic reforms were encapsulated in the Growth, Employment and Redistribution macroeconomic policy (GEAR) strategy. In addition to encouraging growth and employment, this strategy aimed to transform South Africa into a “competitive, outward orientated economy” (RSA, 1996). Measures to reduce unit costs and an exchange rate policy to keep the real effective exchange rate stable at a competitive level formed key components of this strategy. To differentiate itself from the previous protectionist government, the new government also embarked upon an ambition trade liberalisation process that commenced with the government’s formal Offer in the 1995 WTO (Bell, 1997). Numerous other policy changes relating to labour markets and competition have also been implemented. This paper evaluates the extent to which the composition and level of manufacturing exports have responded to these initiatives in the 1990s. We find that the successes of these policies in generating export growth have been mixed. Exports of manufactures have increased but not by enough to generate an export-led growth boom similar to those of East Asia and a few other resource-based export economies. Moreover, South African manufactured exports remain resource-based and the country has lagged others in diversifying into new and fast growing export sectors. The inability to re-structure exports towards these dynamic high technology products is one explanation for the relatively poor export performance of South African manufacturing during the 1990s. * School of Economics, Cape Town University, South Africa. The paper was edited by Željko Bogetić, Lead Economist, AFTP1, The World Bank.
3
The paper also investigates the determinants of South African manufacturing export performance using estimated export supply and demand functions. The analysis finds that South African manufacturers are on average pricetakers in the international market and that exports are predominantly supply driven. Export growth is therefore not predominantly dependent on the economic prosperity of South Africa’s trading partners or on their ability to compete in the export market on the basis of price. Furthermore, the paper finds that many of the constraints to export growth can be found in factors that negatively affect the profitability of export supply. The real effective exchange rate, infrastructure costs, tariff rates and skilled labour are found to be important determinants of export supply. Section 2 of the paper presents a very brief review of South Africa’s trade regime, the increased openness in the 1990s, the changing composition of South Africa’s exports, and its dynamic export performance in comparative perspective. Section 3 develops the export model used to identify the determinants of export performance and then discusses the results. Section 4 contains concluding remarks and some policy implications.
2. South African Trade regime and Trade Patterns During the 1990s This section consists of three components. Progress made in liberalising the economy is first discussed. This is then followed by an analysis of the changing patterns of South Africa’s exports from the 1970s. Finally, South Africa’s dynamic export performance during the 1990s is assessed relative to a range of developing and natural resource-based economies.
Trade liberalisation in the 1990s The democratically elected government in 1994 inherited a protectionist trade regime characterised by high levels of protection, a wide dispersion of tariffs, and a complicated array of tariff types (Belli et al., 1993). The protective trade regime arose from a policy of import substitution industrialisation that began in the 1920s with the substitution of imports of consumer goods by domestic manufactures, but then shifted in the 1970s and 1980s towards import replacement in downstream industries, particularly the chemical and basic metals sub-sectors. Although some initiatives had been made in opening the economy from the 1970s (Export Development Assistance scheme in 1970s, General Export Incentive Scheme in 1990 and the relaxation of quantitative restrictions), reform of the trade regime accelerated with South Africa’s formal Offer in 1995 to the WTO. In this Offer South Africa agreed to bind 98% of all tariff lines, reduce the
4
number of tariff rates to six, to rationalise the over 12000 tariff lines and to replace quantitative restrictions on agricultural products with tariffs. 1 Substantial progress has been made in simplifying the tariff structure of the early 1990s, but further progress can be made. The total number of HS8-digit tariff lines fell from over 11200 in 1994 to 6707 in 2004. The tariff structure has also been simplified with the number of HS8-digit lines bearing formula, mixed or specific duties declining from 3524 in 1994 (30% of total) to 205 in 2004 (3% of total), although almost half of this reduction took place between 2003 and 2004. The number of ad valorem tariff rates also remains high (37 in 2004 vs. 31 in 1994) and exceeds the 6 tariff rates proposed in South Africa’s GATT/WTO Uruguay Round offer. If non-ad valorem tariff rates are included, the number of different rates in 2004 rises to 99. 2 Therefore, there is further scope to simplify the tariff structure as per the Offer to the WTO. Average nominal and effective protection rates have also fallen. The simple average Most Favoured Nation (MFN) tariff rate, inclusive of surcharges, fell from 22% in 1994 to 11.3% in 2003 (Figure 1), although most of this decline occurred prior to 2000. 3 4 Since 2000, tariff rates facing EU and SADC countries have also fallen in accordance with the SA-EU Free Trade Agreement (2000) and the SADC Free Trade Protocol (1996), reaching 9.7% and 5.1% in 2003, respectively. Average effective rates of protection (ERP) have also fallen, but remain high, particularly in manufacturing where they averaged 25% in 2003 (Table 1). 5 Further simplifications of the tariff structure initiated in early 2004 have led to addition reductions in average tariff rates facing MFN (8.3%), EU (7.1%) and SADC (2.4%) countries. 6
1
This is a very cursory overview of the liberalisation process, which has been heavily debated in the South African context. For more detailed discussions see Holden (1992), Belli et al. (1993), Bell (1997), Jenkins et al. (1997), Fedderke and Vase (2001, 2004) and Rangasamy and Harmse (2004).
2
The number of rates in 1990 and 1994 were 733 and 717, respectively.
3
These rates include ad valorem equivalents of formula, specific, compound and mixed duties and are based on HS8-digit tariff lines. Duty collection rates are used to calculate the ad valorem equivalents. See Edwards and van de Winkel (2005) for further details.
4
The decline in average nominal protection in South Africa has been marginally higher than reductions in other developing economies, but this has not significantly affected its ranking (43-44 percentile) in terms of tariff levels.
5
ERP are calculated as ERP = j
t j − ∑ aij t i i
1 − ∑ aij
where tj is the tariff on outputs, ti is the tariff on inputs
i
and aij is the quantity of intermediate input i used in the production of one unit of j. The Balassa (1965) approach is followed and services are given a zero output tariff. See Holden and Holden (1975), Greenaway and Milner (1993) and Holden (2001) for a critical evaluation of ERP. 6
These estimates are based on data obtained from the DTI and use a different approach to calculate ad valorem equivalents. Using this approach 2003 yields tariff rates of 10.7% for MFN, 9.6% for EU and 5.1% for SADC. The weighted average (using import values) nominal tariff in 1993 equals 10.1%.
5
Figure 1: Evolution of nominal tariff protection Evolution of nominal tariff protection 25.0 20.0 15.0 %
Surcharges Tariffs
10.0
All
Agriculture
Mining
2003
2000
1997
1994
2003
2000
1997
1994
2003
2000
1997
1994
2003
2000
1997
0.0
1994
5.0
Manufacturing
Note: The tariff rate for 2003 reflects the weighted average (using import values) of MFN, EU and SADC rates.
All aggregate sectors experienced a decline in nominal and effective protection between 1994 and 2003, but protection remains high in some sectors (Table 1). Relatively strong declines in nominal protection were experienced in textiles, footwear, wearing apparel and communication equipment. Despite these declines, nominal protection remains high in the textile, clothing and footwear sectors where average tariffs exceed 20%. Other highly protected sectors are tobacco (33%) furniture (17.4%) and motor vehicles (15.2%). The structure of effective protection rates is similar to nominal protection rates and therefore high ERP are also found in the tobacco (257%), textiles (76%), clothing (94%), footwear (51%), and furniture (46%) sectors. 7 These rates are substantially lower than in 1994 when ERP exceeded 100% for most of these sectors.
Table 1: Measures of sectoral protection
All Agriculture Mining Total Manufacturing Food Beverages Tobacco Textiles Wearing apparel Leather products
7
Scheduled rates 1994 2003 % Δb 21.9 10.6 -9.3 8.9 4.4 -4.1 2.8 0.6 -1.9 22.5 10.9 -9.5 18.8 11.5 -6.2 29.3 15.4 -10.8 41.7 32.9 -6.2 41.3 20.3 -14.8 75.1 33.4 -23.8 25.9 11.3 -11.5
ERP Anti-export biasa 1994 2003 1994 2003 38.6 18.9 2.0 1.4 7.3 3.8 1.2 1.1 3.8 -1.2 1.1 1.0 48.4 24.3 2.2 1.5 55.3 38.3 3.1 1.9 51.9 28.4 2.0 1.4 340.5 257.2 13.4 6.2 149.7 76.2 3.3 2.1 218.4 94.1 4.2 2.2 59.7 18.8 6.1 2.0
The correlation between ERP and nominal protection rates exceeds 0.79 in all cases.
6
Footwear Wood products Paper products Printing & publishing Coke & petroleum Basic chemicals Other chemicals Rubber products Plastic products Glass products Non-metallic minerals Basic iron & steel Non-ferrous metals Metal products Machinery & equipment Electrical machinery Communication equipment Professional & scientific Motor vehicles Other transport Furniture Other manufacturing
48.0 14.5 11.3 16.1 5.1 8.1 16.2 18.6 19.8 17.2 15.0 8.8 10.8 18.3 10.4 18.3 24.2 12.2 25.9 12.3 32.1 26.5
22.7 8.5 6.2 4.6 3.3 1.6 4.4 10.8 9.8 7.2 5.6 4.3 2.1 7.9 3.6 7.1 2.9 0.3 15.2 0.8 17.4 5.9
-17.1 -5.3 -4.7 -9.9 -1.8 -5.9 -10.2 -6.5 -8.4 -8.5 -8.2 -4.2 -7.9 -8.8 -6.2 -9.4 -17.1 -10.6 -8.5 -10.2 -11.2 -16.2
106.0 21.7 15.8 22.2 10.0 14.4 32.3 46.6 36.2 32.1 29.9 20.1 17.9 36.7 11.9 33.0 35.5 9.5 45.1 14.9 82.6 96.5
51.1 14.0 10.3 4.5 8.2 1.4 7.4 31.7 20.3 13.3 10.8 11.0 2.9 16.1 2.9 13.8 1.2 -5.9 32.3 -3.2 46.4 17.5
5.1 1.5 1.5 1.4 1.2 1.3 1.8 1.9 1.7 1.5 1.4 1.4 1.3 1.7 1.4 1.8 2.2 1.5 2.4 1.5 3.1 3.0
2.1 1.3 1.2 1.1 1.1 1.1 1.2 1.5 1.3 1.2 1.2 1.2 1.1 1.3 1.1 1.3 1.1 1.0 1.6 1.0 1.8 1.3
Notes: a. To capture the duty free credit system implemented in the clothing & textile and motor vehicle industries, zero tariffs were imposed on textile inputs in the production of clothing and textiles, and motor vehicles inputs used in the production of vehicles. b. Calculated as ((t1-t0)/(1+t0)-1). Tariffs include ad valorem equivalents for formula duties, specific duties and mixed duties. The AVE are calculated using collection rates. For formula duty and mixed duties the AVE are equal to the collection rates if these exceed the ad valorem component of the tariff. The average value for manufacturing is calculated as the weighted average of the 3-digit SIC codes using real output between 1988-2002 as weights. The tariff rate for 2003 reflects the weighted average (using import values) of MFN, EU and SADC rates. The values for the scheduled tariffs are simple averages at the HS8 digit level.
The reduction in tariffs has also contributed significantly towards raising the profitability of export production and reducing the anti-export bias (AEB) (Table 1). 8 Import taxes on intermediate goods were equivalent to 38% of value added in aggregate manufacturing in 1994 (not taking into account export subsidies), but fell to 19% by 2003. The combined effect of a reduced tax on intermediate goods and reduced effective protection was a reduction in the anti-export bias in aggregate manufacturing from 2.2 to 1.5 over this period. However, much of the improvement in the AEB from tariff liberalisation from 1994 to 1997 was offset by the removal of export subsidies under the General Export Incentive Scheme. Kuhn and Jansen (1997), for example, estimate that the removal of export subsidies led to an increase in the anti-export bias between 1993 and 1996. 9 In conclusion, South Africa has made considerable progress in reducing tariff protection during the 1990s, but there is still scope for further 8
The anti-export bias is calculated as (1+ERP)/(1-XRP) where XRP is the implicit export tax of tariffs ∑ aij ti calculated as XRP = i . j 1 − ∑ aij i
9
The lack of export subsidy data prevented a similar analysis in this paper.
7
simplification of the tariff structure. Average nominal protection is still higher than the average for developing economies and high effective protection rates remain in many manufacturing sectors. Nevertheless, tariff liberalisation has raised the profitability of export supply, which will have enhanced export growth during the 1990s.
Increased openness The reductions in tariffs, the re-integration of South Africa into the international arena and a real depreciation of the exchange rate has raised exports and imports as a share of GDP. As shown in Figure 2, merchandise exports as a share of GDP fell during the 1970s and the early 1980s in response to a decline in primary sector exports, particularly gold exports. From the mid-1980s manufactured export growth, spurred by a sizeable real depreciation and recession driven ‘vent-forsurplus’ exports (Fallon and Pereira da Silva, 1994), reversed this trend and total merchandise exports rose as a share of GDP. The role of manufacturing exports in driving this trend is reflected in the rise in its share of GDP from 4% in 1985 to 9.8% in 1994 (Figure 2). Manufacturing exports as a share of GDP continued to rise in the 1990s, despite the modest recovery in output growth. By 2000, manufactured exports as a share of GDP had risen to 15%. The rising importance of exports in aggregate manufacturing is also shown by the rise in export orientation (exports/gross output) from 12% in 1993 to 23% in 2000. This increased openness of manufacturing during the 1990s has been broad based, with export orientation rising in 25 of the 28 SIC 3digit sectors analysed (Table 2). Manufacturing firms have thus become firmly entrenched in the international market and no longer export primarily on a vent-forsurplus basis. Figure 2: Indices of openness, 1970-2001 Merchandise trade to GDP 30% 25% 20%
Merchandise exports/GDP
15%
Manufacturing exports/GDP
10%
Merchandise imports/GDP
5%
8
2000
1997
1994
1991
1988
1985
1982
1979
1976
1973
1970
0%
Growth in imports also recovered strongly in the 1990s in response to the recovery in output growth and the reduction in import barriers. Import growth had stagnated during the 1980s in response to the depreciation in the mid-1980s, the imposition of surcharges from 1988 and the onset of the domestic recession in 1989. However, from 1990 growth in merchandise imports recovered raising its share of share of GDP from 13% in 1990 to 21% in 2000 (Figure 2). Import penetration in manufacturing rose from 17% to 28% over the same period. The rise in import penetration has also been broad based, rising in 24 of the 28 SIC 3-digit sectors between 1990 and 2000. Particularly strong increases occurred in the textiles, apparel and footwear sectors as well as rubber products, other transport equipment and furniture. Table 2: South African exports by industry as a share total manufactured exports and export orientation Manufacturing (Rmill, 1995 prices) Manufactured exports/total merchandise trade Total Manufacturing Food products Textiles, apparel Wood Paper and printing Chemicals Non-metallic minerals Base metals Fab. metal and machinery Other Export orientation Manufacturing Food products Textiles, apparel Wood Paper and printing Chemicals Non-metallic minerals Base metals Fab. metal and machinery Other
1970
1980
1990
1994
2000
14,495
22,319
38,793
47,617
85,536
22%
29%
43%
47%
62%
100% 18% 4% 0% 9% 17% 4% 13% 21% 13%
100% 13% 6% 2% 5% 15% 3% 33% 12% 11%
100% 9% 5% 2% 7% 15% 1% 32% 18% 11%
100% 8% 6% 2% 6% 19% 2% 25% 21% 12%
100% 5% 3% 2% 6% 20% 1% 22% 32% 9%
8% 8% 5% 1% 11% 14% 8% 12% 5% 18%
8% 6% 7% 9% 7% 10% 7% 26% 3% 28%
12% 5% 9% 12% 11% 11% 5% 44% 8% 18%
15% 6% 12% 14% 12% 15% 8% 43% 12% 24%
23% 7% 14% 18% 18% 23% 11% 44% 26% 29%
Source: Own calculations based on data obtained from Quantech (2004). Export orientation is calculated as the share exports in gross output. Calculations based on real values.
The composition of exports has also changed with manufacturing displacing mining as the dominant export sector. During the 1970s and early 1980s mining sector exports accounted for between 60% and 65% of total exports (including services exports). The bulk of this was gold exports, which accounted for between 35% and 52% of total merchandise exports from 1972-1985 (Bell et al, 1999: Table
9
2). 10 With the collapse in the gold price in the 1980s and the declining grade of ore, the share of mining exports in total exports (gold in particular), fell dramatically, reaching 47% in 1990 and 29% in 2000. In contrast, in response to relatively strong export growth, the share of manufacturing in total exports rose from 25% in 1980 to 41% in 1994. During the 1990s, manufacturing overtook mining as the most important export sector, accounting for 53% of total exports by 2000.
Figure 3: Skill bias of manufactured export growth Skill bias of e xport growth, 1980s
0.30
0.20
0.25
0.15
0.20 0.15 0.10 0.05
y = -1.66x + 0.13
0.00
R = 0.27
-0.050.00
2
0.02
0.04
0.06
0.08
0.10
0.12
0.14
Average annual growth
Average annual growth
Skill bias of e xport growth, 1970s
-0.10
y = 0.65x + 0.02 2
R = 0.13
0.10 0.05 0.00 -0.050.00
0.05
0.20
Highly skille d/Le ss skille d ratio
Skill bias of e xport growth, 1990-94
Skill bias of e xport growth, 1994-2002
0.60 0.50
Average annual growth
Average annual growth
0.15
-0.15 Highly skille d/Le ss skille d ratio
0.40 0.30 y = -0.21x + 0.14
0.20
2
R = 0.01
0.10 0.00 -0.100.00
0.10
-0.10
0.05
0.10
0.15
0.20
0.25
0.20 0.15 0.10 0.05 0.00 -0.050.00 -0.10 -0.15 -0.20 -0.25
y = 0.66x - 0.03 2
R = 0.33
0.05
0.10
0.15
0.20
0.25
0.30
Highly skille d/Le ss skille d ratio
Highly skille d/Le ss skille d ratio
Manufacturing exports have also diversified during the 1990s, with relatively strong export growth in skill-intensive sectors. This is clearly shown in the scatter plots in Figure 3 where high export growth was experienced from 1994 in relatively skill-intensive sectors such as coke and refined petroleum products, other chemicals, motor vehicles, parts & accessories and other transport equipment. Particularly strong export growth was experienced in motor vehicles and other transport equipment sectors in response to the Motor Industry Development Programme (MIDP) introduced in 1995. The share of these two sectors in total manufacturing exports rose from 6% in 1990 to 19% in 2002. Poor export growth, combined with strong import growth, occurred in less-skill intensive sectors such as textiles, wearing apparel, leather and footwear has also contributed towards the skillbias of export growth. 11 12
10
Based on IDC (1995)—data in current prices.
11
Net exports show similar trends. However, no consistent trend is found when using the capital intensity (machinery and equipment capital per worker) of production within manufacturing sectors.
10
South Africa’s Dynamic Export Performance: A Cross-country Comparison The ways in which export patterns change over time has profound implications for the relationship between trade on the one hand, and industrialisation and economic growth on the other. “Success in entering lines of production with significant potential for global demand expansion, high value added and rapid productivity growth widens the scope for the exploitation of increasing returns from larger markets, and enhances the role of trade in economic growth” UNCTAD (2002: 52). In contrast, a high concentration of exports in sluggish global markets or activities with limited potential for productivity growth endangers the growth process. Middle-income countries like South Africa have additional reasons to urgently address the issue of upgrading their export structures. The entry of low wage labour abundant economies such as China and India into the world market have challenged middle-income economies’ comparative advantage in low-skill manufactures. As argued by (UNCTAD, 2002: 126): “… it is imperative that middle-income countries upgrade rapidly from low skill to more market-dynamic, technology-intensive products with a view to successfully competing with industrialised countries and the first-tier NICs. If not, they risk being squeezed between the bottom and top ends of the markets for manufactured exports”
This section, therefore, presents a cross-country analysis of South Africa’s dynamic export performance using two different dynamic indicators. Following UNCTAD (2002), ‘dynamic’ products are defined in terms of their global demand potential (market-dynamic products) and productivity potential (supply-dynamic products). The former classification identifies products that have a strong growth potential in world markets, while the latter identifies products characterised by strong productivity growth potential. 13
Supply-dynamic products Performance in supply-dynamic products is assessed using a technologybased product classification, developed by Lall (2000), and used in UNCTAD (2002) and UNIDO (2004). As shown in Table 3, exports are classified into primary products (PP), resource based manufactures (RB), low technology manufactures (LT), medium technology manufactures (MT) and high technology manufactures (HT). 12
Turning to imports, South African imports are heavily concentrated in capital and intermediate goods and reflect the high import dependency of the capital-intensive basic metals and chemicals subsectors. etailed discussion of import composition is contained in the longer version of this paper available from the author.
13
The data used in this sub-section differs in important ways from that presented in the rest of the paper. First, data for other countries is obtained from the UN Comtrade database and is classified according to SITC (Revision 2). For South Africa, export data at the 8-digit HS level are obtained from Customs & Excise and converted to 3-digit SITC level using a concordance files obtained from UN Comtrade. Secondly, to facilitate cross-country comparisons, all data are expressed in current US dollars rather than constant 1995 rands.
11
Primary products and resource-based manufactures tend to be unskilled-labour- and scale-intensive, and skill requirements tend to rise with the degree of technological complexity (see Lall (2000) for a complete description of each technology category). “Since increased application of human capital and technology tends to raise labour productivity, such a classification can be expected to provide a reasonably good guide to sectoral differences in the potential for productivity growth” UNCTAD (2002:66). The ability of a country to shift exports into high technology products therefore has important implications for long run output growth, as productivity growth becomes the primary source of output growth (and income per capita) once underutilised labour and natural resources are exhausted.
Table 3: The Technological Classification of Exports 14 PRIMARY PRODUCTS
Fresh fruit, meat, rice, cocoa, tea, coffee, wood, coal, crude petroleum, gas, metals
MANUFACTURED PRODUCTS Resource based manufactures RB1: Agro/forest-based products
Prepared meats/fruits, beverages, wood products, vegetable oils
RB2: Minerals-based products
Ores & concentrates, petroleum/rubber products, cement, cut gems, glass
Low technology manufactures LT1: ‘Fashion cluster’
Textile fabrics, clothing, headgear, footwear, leather manufactures, travel goods
LT2: Other low technology
Pottery, simple metal parts/structures, furniture, jewellery, toys, plastic products
Medium technology manufactures MT1: Automotive products
Passenger vehicles and parts, commercial vehicles, motorcycles and parts
MT2: Process industries
Synthetic fibres, chemicals and paints, fertilisers, plastics, iron, pipes/tubes
MT3: Engineering industries
Engines, motors, industrial machinery, pumps, switchgears, ships, watches
High technology manufactures HT1: Electronics and electrical products
Office/data processing/telecommunications equip, TVs, transistors, turbines, power generating equipment
HT2: Other high technology
Pharmaceuticals, aerospace, optical/measuring instruments, cameras
“SPECIAL” TRANSACTIONS
Electricity, cinema film, printed matter, art, coins, pets, non-monetary gold
Source: Lall (2000)
14
This study moves non-monetary gold (SITC 971) from “special transactions” into the primary products category, and precious and semi-precious stones (SITC 667) from resource-based manufactures to primary products.
12
There are also good market-dynamic reasons for shifting exports into high technology products as these products are the fastest growing in world trade. As shown in Table 4, the average annual export growth of high technology products (9.1%) between 1988 and 2002, was almost double that of low technology (5.6%), medium technology (5.7%) and resource-based (4.9%) products. The share of high technology products in world trade rose from 15.8% in 1988 to 23.6% in 2002. All other categories had a smaller share in world trade in 2002 than they did in 1988, although the medium technology group has regained some of its losses since 2000. An important feature of the changing pattern of world trade in the 1990s, is the rapid growth in developing country exports, particularly within high and medium technology products (Figure 4). Total developing country as a share of world trade rose from 19.1% in 1988 to 30.5% in 2002. Its share in high technology products doubled from 14.8% to 34.3% over the same period. By 2002, the value of developing country exports of medium and high technology products together exceeded the combined value of primary products, resource-based and low technology manufactures. In contrast, in 1988 medium and high technology exports equalled approximately half the value of the remaining categories.
Figure 4: Developing country market share gains in total exports, by broad technology category, 1988 & 2002 (%) 15 50
25
45 40
20
35 30
15
25 20
10
15 10
5
5 0
0 Total exports Share 88
Primary products Share 02
Resource-based manufactures Gain
Low tech manufactures
Medium tech manufactures
Developing country growth rate
High tech manufactures World growth rate
Source: as above Note: Share gains expressed simply in percentage points. Growth rates are annual averages.
15
Country groupings follow World Bank definitions. “Developing countries” include the four mature Tiger economies of East Asia (Hong Kong, Korea, Singapore and Taiwan).
13
In comparison to developing countries, however, South Africa’s total export growth during the 1990s (2% per annum) has been relatively poor. South African export growth has also lagged the average for the world (6%), developed countries (5%), and a selection of countries with similar shares of resource intensive products in total exports to South Africa, the Resource Group (6.1%) (Table 4). 16 An important source of this relatively poor growth is the negative growth in primary product exports (-1.1%) during this period. As a result of the poor export performance, South Africa’s share of world exports declined 0.89% to 0.52% between 1988 and 2002. Table 4: Annual average growth rates by broad technology category, 1988-2002 (%) 17
6.02
Developed countries 4.96
Developing countries 9.58
South Africa 2.02
Resources Group 6.14
Primary products
3.59
2.79
4.95
-1.14
4.18
Total manufactures
6.32
5.13
10.63
6.91
7.72
World Total exports
Resource-based
4.89
4.09
7.89
4.26
5.63
Pure manufactures
6.59
5.33
11.13
8.57
9.52
Low technology
5.63
4.37
7.94
5.57
8.57
Medium technology
5.67
4.77
11.07
9.67
8.51
High technology
9.10
7.14
15.83
11.53
14.95
Source: as above. As described in Appendix C, South African data sources comprise the following: Customs and Excise, the Minerals Production and Sales Statistics database, the South African Reserve Bank (SARB), and the South African Department of Trade and industry (DTI).
Within aggregate manufacturing, exports have grown marginally more quickly than the global average, but still slower than other developing countries and the Resource Group. 18 The growth performance vis-à-vis developing countries is relatively poor in all technology categories and is particularly weak in high technology products. South African exports also lag that of the Resource Group in all but the medium technology category. The relatively strong growth in Medium technology exports can be attributed to the very strong growth in motor vehicle exports under the MIDP programme.
16
The Resource Group consists of twenty-five countries, listed in Appendix A, selected on the basis of having similar levels of natural resource dependent products (primary products + resource-based manufactures) in total exports in the late 1980s. That is, on the basis of having broadly similar export structures. Most of the countries are low- and middle-income countries from Latin America and SubSaharan Africa, but also included were Australia, New Zealand, Norway and Indonesia.
17
Country groupings follow World Bank definitions. “Developing countries” include the four mature Tiger economies of East Asia (Hong Kong, Korea, Singapore and Taiwan).
18
There is substantial variation in export performance amongst the members of the Resource Group. The OECD economies, Australia, New Zealand and Norway, have growth rates more in line with developed economies. Indonesia, on the other hand, experienced very high growth rates, especially in high technology products. The growth rates of these ‘outliers’, however, tend to cancel each out, leading to similar growth rates for the group as a whole to those calculated when these four ‘outliers’ are excluded.
14
The relatively poor export growth has resulted in little or no gains to South Africa’s share of world manufacturing exports. This is shown in Table 5 which presents world market shares by region and technology sub-category. South Africa’s share of world exports of manufactures rose marginally from 0.3% to 0.32% between 1988 and 2002, while its share of pure manufactures (non-resource based manufactures) rose from 0.2% to 0.26%. 19 As can be seen, South Africa’s faster growth vis-à-vis the world in MT exports has been due to very strong performances in MT1 (automotive), whose world market share (WMS) increased seven-fold from 0.05% to 0.36%, and MT3 (engineering), whose WMS roughly tripled. In hightechnology products, South Africa has seen a small rise in its 1988 share of HT1 (electronics), and no change at all in HT2 (other). The relatively poor performance is in stark contrast to the pattern in East Asia, whose share of manufacturing exports rose from 12% to 18%, and share high technology exports rose from 13% to 27% over this period.
Table 5: South Africa’s changing share of developing country, Resource Group and world exports, 1988-2002 (%)
Total manufactures Resource based
Developing Countries % 1988 2002 change 1.7 1.1 -38.0
Resource Group % 1988 2002 change 12.8 11.5 -10.0
World 1988
2002
0.30
0.32
% change 6.7
4.3
2.6
-38.1
11.0
9.1
-16.7
0.76
0.7
-7.9
RB1: Agro-based
2.8
2.0
-26.8
7.9
7.0
-11.9
0.50
0.54
8.0
RB2: Minerals-based
6.6
3.5
-46.1
14.6
12.3
-15.8
1.16
0.94
-19.0
0.9
0.7
-26.7
12.8
8.7
-32.4
0.28
0.28
0.0
LT1: Fashion cluster
0.3
0.3
-7.2
5.4
3.9
-26.9
0.15
0.18
20.0
LT2: Other
1.9
1.1
-42.7
23.1
15.5
-32.9
0.39
0.36
-7.7
Medium technology
2.1
1.8
-16.3
18.2
21.1
16.0
0.23
0.39
69.6
MT1: Automotive
0.9
2.4
151.6
14.4
34.6
140.0
0.05
0.36
620.0
MT2: Process
4.6
2.6
-43.6
26.7
23.7
-11.1
0.74
0.73
-1.4
MT3: Engineering
0.7
1.0
50.3
7.6
13.6
80.6
0.08
0.24
200.0
High technology
0.3
0.2
-41.1
6.6
4.3
-34.5
0.04
0.06
50.0
HT1: Electronic
0.2
0.1
-32.4
6.0
4.1
-32.2
0.03
0.05
66.7
HT2: Other
1.3
0.8
-41.8
7.4
4.8
-35.3
0.07
0.07
0.0
Low technology
Source: as above. Note: Percentage changes calculated as (share02-share88)/share88*100.
South Africa’s share of developing country and Resource Group manufactured exports fell. The share of developing country exports fell from 1.7% to 1.1% and those of the Resource Group from 12.8% to 11.5% between 1988 and 19
South Africa’s share for world market share for total exports dropped alarmingly (0.89% to 0.52%), but again this is due to the abnormal influence of gold on South Africa’s market share in primary products.
15
2002. 20 Loss in share occurred in all technology categories apart from MT1 (automotive) and MT3 (engineering). Although overall export growth has been comparatively low, some progress has been made in diversifying manufacturing exports towards medium and high technology products. This diversification is mainly due to the strong growth in motor vehicle related exports (medium technology), with some minor diversification towards high technology products (Figure 5). The Resource Group also diversified its manufacturing exports, but this diversification was more evenly spread across high technology, medium technology and low technology products. The strongest restructuring towards high technology sectors occurred within the East Asian and Pacific region where the share of these sectors in total exports rose from 21% to 41% during the 1990s. Figure 5: Manufactured export structures, 1988-90 and 2000-02 100%
3 17
19
28
32
80%
5
21 41
13
25
34
27
46
27 60%
5
25 21
43
28
39 40%
16
17
14
40 27
19
20%
18
13
26 13
World
1988-90
2000-02 1988-90
Developing countries
Resource-based
38
33
2000-02 1988-90
EAP
Low tech
49
44
9
0% 1988-90 2000-02
22
24 38
21
20
2000-02
South Africa
Medium tech
1988-90 2000-02 Resources Group
High tech
Market-dynamic products The second indicator of South Africa’s dynamic export performance is its positioning in demand/market-dynamic products. Performance in market-dynamic products is measured by the change of a country’s world market share in products that are becoming steadily more important to global trade. Countries that are able to shift
20
Interestingly, the share reduction is almost twice as large if one removes the four mature Tiger economies are removed from the developing country group. This reflects the emergence during the 1990s of Mexico and China as manufacturing giants, but also Hong Kong’s (and to a lesser extent Singapore and Korea’s) de-industrialisation and shift into tertiary sector activities (Lall & KraemerMbula, 2005).
16
export production towards products with a strong global demand potential will reduce the risk of declining export growth and declining terms of trade. Previous analyses have noted weaknesses in South Africa’s export performance from the market positioning perspective (Van Seventer and Gibson (2004; Tsikata, 1999). 21 In this section, we extend these analyses and compare South Africa’s performance with East Asia’s and the Resource Group’s. The approach used here is similar to that of Tsikata (1999) and Van Seventer and Gibson (2004), but anchors the analysis around those products that contributed the most to South African export growth during the 1990s. Firstly, products (at 3-digit SITC level) are identified as dynamic products if world export growth exceeds the average for all products (i.e. the share of the product is rising in world trade). Secondly, South Africa’s top 20 products (at 3-digit SITC level) are then identified and their world market share (WMS) is calculated. 22 To assess whether these products are positioned optimally in terms of global demand, South Africa’s world market share (WMS) in each of them is compared to world export growth. We postulate that South Africa is positioned optimally in terms of global demand if its WMS in these products is rising and they are experiencing higher than average world export growth. This relationship is more clearly shown in Figure 6, which plots South Africa’s changing WMS of its top 20 export growth products against world export growth. The size of the bubble reflects South Africa’s export value of each product in 2002. The horizontal dotted line in the figure represents average world growth for all products (6.02% per year). Optimal market positioning occurs if a large proportion of the 20 products are situated in the top right hand ‘quadrant’ (above the dotted line and to the right of the vertical axis), i.e. South Africa’s WMS is rising in products experiencing above average growth in world demand. Using Tsikata’s (1999) terminology, these are ‘rising stars’. Poor market positioning occurs if a product’s WMS is rising in stagnating (in terms of world share) world markets (bottom right quadrant), or its WMS is falling in dynamic world markets (top left). The bottom left quadrant below average world growth reflects products in which the economy is rapidly retreating from stagnating world markets.
21
Van Seventer and Gibson (2004) find a low share of South African exports in the top 40 demanddynamic products identified by UNCTAD (2002). Tsikata (1999) finds that relative to a range of middle-income economies (Korea, Mexico, Taiwan, Malaysia, Thailand, Brazil), South Africa exports a relatively high proportion of products for which world markets are not growing very fast.
22
The top 20 export growth products accounted for 53% of South Africa’s total exports in 2002.
17
Figure 6: The market positioning of South Africa’s top 20 exports 10 784 - furniture and parts 781 - passenger cars
057 - fruit & nuts
743 - pumps & compressors
9 289 - precious metal ores Global total export 112 - alcoholic bev. growth rate
8
784 - car parts
Global growth rate, (%)
7
713 - engines
6
672 - iron ingots
667 - silver & platinum
034 - fish
5
684 - aluminium 4 667 - precious stones (diamonds)
641 - paper & paperboard
246 - pulpwood 3
782 - special purpose vehicles
281 - iron ore
2 322 - coal & lignite
1 671 - pig iron
522 inorganic chemicals 0
-15
-10
-5
0
5
10
Change in South Africa's world market share (WMS)
Note: Dotted line is world growth for all products. WMS changes expressed as percentage points.
As can be seen, the market positioning of South African exports is poor: The majority of its top 20 exports during the 1990s are in stagnating world markets (below the dotted line). These sectors in the diagram alone account for 43% of total South African exports in 2002. Very few of South Africa’s most important exports are found in the top right hand ‘quadrant’. The major sectors that do fall into this category are passenger cars (SITC 781), pumps & compressors (SITC 743), furniture (SITC 784) and precious metals (SITC 289). Together these sectors only account for 13% of total South African exports. The contrast in performance with East Asia is clearly shown in Figure 7, where 16 of the top 20 exports are in dynamic world markets. Thirteen of these fall in the top right hand quadrant. In other words, most of East Asia’s top exports are in markets growing faster than the global average, and it is gaining world market share in them.
18
15
Figure 7: The great market positioning of East Asia’s top 20 exports 14
771 - elec. power mach. 764 - telecomms. equip.
341 - natural & manuf. gas
776 - transistors & semiconductors
12 821 - furniture
655 - knitted fabrics
Global growth rate (%)
10 893 - articles of plastic 759 - computer parts
843 - women's outer garments
8 778 - elec. machinery
752 - computers
6 772 - switches & relays
894 - baby carriages 4
763 - sound rec. equip.
845 - knitted garments
793 - ships
Global total export growth rate
583 - polymerization products
2 781 - passenger cars
749 - non-elec. parts
775 - household equip.
0 -15
-10
-5
0
5
10
15
20
25
30
35
Change in EAP's world market share (WMS)
Note: Dotted line is world growth for all products. WMS changes expressed as percentage points.
The market positioning of the Resource Group also poor with a high proportion of the group’s top export performers located in the lower right quadrant (Figure 8). There is therefore some similarity in performance between South Africa and the Resource Group. A high concentration of exports in primary and natural resource-based products has negatively affected export growth for resource based economies similar to South Africa during the 1990s. Further, these economies have also been unable to restructure significantly into dynamic world markets. This similarity between South Africa and the Resource Group suggests that natural resource endowments are an important determinant of export performance and the ability to diversify. The ability of these economies to diversify into high technology products is constrained by the comparative advantage in resource-based products that the rich natural resource endowments provides them with. However, there are also important differences in export growth and diversification amongst countries within the Resource Group. These differences may highlight the role of country specific effects such as trade policy, infrastructure, skills, etc. The importance of many of these supply side determinants is assessed in Section 3.
19
40
Figure 8: The market positioning of the Resource Group’s top 20 exports 16 541 - pharma.
341 - natural & manuf. gas
14
764 - telecomms. equip.
12
759 - computer parts 821 - furniture
10
Global growth rate (%)
Global total export growth rate
843 - women's outer garments
8
112 - alcoholic bev.
781 - passenger cars
424 - veg. oils
6
057 - fruit & nuts
034 - fish
011 - edible meat
2 682 - copper
322 - coal & lignite
641 - paper
4
022 - milk
081 - animal feed
287 - base metal ores
0 -20
-15
-10
-5
0
971 - gold
5
10
041 - wheat
15
-2 Change in Resources Group's meorld market share (WMS)
Note: Dotted line is world growth for all products. WMS changes expressed as percentage points.
Concluding points The data analysis in this section highlights a number of important features of South African export performance since the late 1980s: •
Significant progress has been made in liberalising South Africa’s trade, although there is scope for further simplification of the tariff structure.
•
Manufacturing export growth has increased, but has been poor relative to the average for developing countries and a range of economies that had similar export structures to South Africa in the early 1990s. South Africa’s manufactured exports are being “squeezed” at both ends of the technology spectrum, and not only by better performing countries in East Asia.
•
The structure of manufacturing exports is highly concentrated in natural resource-based products, but there has been some diversification into skillintensive and medium technology products through increased exports of vehicles, chemicals and engineering products. However, South Africa lags other economies (including similar resource-based exporters) in its restructuring of exports towards high technology products.
•
The concentration of South African manufacturing (and total) exports in products with relatively weak world export growth has contributed towards the relatively poor performance of South African exports.
20
20
The evidence therefore indicates that while manufacturing export growth increased, it has not been sufficient to generate an export-led boom as has been experienced in many economies, particularly within East Asia. The question is why? One of the reasons for the relatively poor export growth appears to be the concentration of exports in natural-resource-based products, which experienced relative low growth in world markets. However, South Africa’s export performance was weak even in natural resource-based products. Further, export growth and diversification into high technology products was poor compared to a range of similar resource-based exporters. This suggests that there were important domestic constraints to export growth during the 1990s. In the following section we draw upon various techniques, including econometric estimations of export supply and demand function, to investigate the various determinants South African manufacturing export performance.
3. Determinants of South African Manufacturing Export Performance There is a diverse and growing empirical literature on the determinants of South Africa’s export performance. This literature includes cost or price competitiveness analyses through the use of real effective exchange rates (IMF, 1998, Kahn, 1998, Walters and de Beer, 1999, and Golub, 2000); Revealed Comparative Advantage studies (Valentine and Krasnik, 2000); shift-share analyses of the composition of exports (Nordas, 1996; Bell et al. 1999. Edwards and Schoer, 2002); market positioning studies (Edwards and Schoer, 2002; Van Seventer and Gibson, 2004); and econometric estimates of export supply and demand functions (Smal, 1996; Tsikata, 1999; Edwards and Wilcox, 2003; Edwards and Golub, 2004). To investigate the determinants of South African manufacturing export performance, this section estimates export demand and supply relationships using a panel of industry data from 1970-2002. The analysis extends existing empirical work in South Africa in two ways. Firstly, a fuller specification of the export supply relationship is estimated. Secondly, the endogeneity of export volumes and export prices is taken care of.
Specifying the Export Demand and Supply Relationships We use a variant of the imperfect substitution model outlined in Goldstein and Kahn (1985) and discussed further in Edwards and Wilcox (2003). 23 This model is represented as a system of equations for export supply (Xs) and export demand (Xd), which simultaneously determine the export price and the export
23
The model is an imperfect substitutes model where imperfect substitutability between domestic and export products enables domestic and export prices to differ from one another (Goldstein and Kahn, 1985).
21
quantity. The long run export demand (Xd) and supply (Xs) relationships are given by the following log-linear structures: X d = δ 0 − δ 1 PX + δ 2 e + δ 3 P ∗ + δ 4Y ∗ , δ i > 0
(1)
X s = α 0 + α 1 PX − α 2 P − α 3 C + ΨZ , α i > 0
(2)
and
where (all variables in logs): X Y* P* Px e P C Z
= volume of exports = real foreign income = foreign domestic price = domestic price of exports = domestic to foreign currency exchange rate = domestic price = nominal variable cost = vector of other variables that influence the supply of exports
Export demand is positively affected by foreign income (Y*) and the price of competing foreign goods (P*), but is negatively affected by the foreign price of domestic exports (Px*= Px/e). 24 The quantity of exports supplied is specified as a positive function of its own price and a negative function of the domestic price index and variable costs. 25 As export sales become profitably relative to domestic sales (Px/P rises) firms shift production towards the export market. Other supply side variables include tariff rates, import penetration, infrastructure costs, capacity utilisation and trend income. 26 Following Fallon and Pereira da Silva (1994), Tsikata (1999), Behar and Edwards (2003) and Edwards and Golub (2004), capacity utilisation is included to test the “vent-for-surplus” hypothesis. A negative coefficient is expected. Tariff liberalisation reduces the anti-export bias of production and thus positively affects export production. Trend income is included as a proxy for non-price improvements in competitiveness (infrastructure, total factor productivity, export supply networks, learning by doing) arising from increased economic activity. Finally, infrastructure constraints are expected to negatively affect export supply.
24
Normally it is assumed that the demand function is homogenous of degree zero in prices and the restriction -δ1(=δ2)+δ3 = 0 is imposed, i.e. Px/eP* is included on the right hand side.
25
Homogeneity on the supply function requires the restriction that α1+α2+α3=0. Alternatively, the supply function can be specified in terms of real variable costs and the relative price of exports to domestic prices (Px/P).
26
We also tested two concentration indices, the Gini coefficient and Rosenbluth index, obtained from Fedderke and Szalontai (1995). These variables were only available till 1996, thus restricting our sample size considerably. The coefficients were also mostly insignificant, although when significant they were positive. The specifications presented in this study therefore exclude the concentration indices.
22
An important consideration in estimating the export supply and demand functions is that Px and X are endogenous variables. Failure to account for this endogeneity will give rise to simultaneous equation bias when estimating either equation. 27 However, this is less of a problem in small price-taking economies where the export price is exogenous and the demand for exports is infinite. As export prices are no longer endogenous, the export supply function can be estimated independently of the demand equation. Two approaches to the estimation of the export demand and export supply functions are followed in this study. Firstly, we first estimate the export demand function and test whether the small country assumption holds in the case of South Africa. Following Riedel (1988) export demand (equation 1) is normalised on Px to obtain
Px =
δ0 1 d δ2 δ δ − X + e + 3 P∗ + 4 Y ∗ . δ1 δ1 δ1 δ1 δ1
(3)
In a small price taking country, the export price elasticity of demand (δ1) tends towards negative infinity and the coefficient on Xd and Y* therefore tend towards zero. 28 Equation (3) then becomes the standard PPP relationship in which export prices, measured in domestic currency, equal foreign prices multiplied by exchange rate. If price homogeneity holds, the coefficients on the exchange rate and foreign prices equal one, i.e. δ2/δ1 = δ3/δ1 = 1. Secondly, we estimate the reduced form function for export volumes. 29 Imposing the homogeneity assumptions and expressing the export demand and supply functions in terms of relative prices (i.e. δ1=δ2=δ3=δp and α1 = α2 = αp) and real unit labour costs (RC), the reduced form equation for exports is expressed as: X=
[
(
]
)
1 λ 0 + λ 2 e + P ∗ − P + λ3Y ∗ − λ 4 RC + λ5 Z , λi > 0 1 + λ1
(4)
where
λ0 = α 0 +
α 1δ 0 α αδ , λ1 = 1 , λ2 = α 1 , λ3 = 1 4 , λ4 = α 3 , λ5 = Ψ . δ1 δ1 δ1
27
This arises because the export volume and price in the demand and supply relationship are correlated with the error terms. Domestic prices, wages and the exchange rate may also be endogenous. Export growth can affect the exchange rate, which in turn affects inflation and wages. This problem may be particularly problematic for South Africa during the 1970s and 1980s when the gold price rose and then fell.
28
However, rising world demand for a particular product will affect export supply through its impact on world prices.
29
Initial estimates of the export supply function (2) gave a negative instead of positive coefficient for the relative price of exports to domestic products (PX/P); a result also found by Fallon and da Silva (1994). It was thus decided to concentrate on the reduced form results.
23
(e+P*-P) is Real Effective Exchange rate measuring the price of foreign products relative to South African products, valued in a common currency. A real depreciation (e+P*-P rises) positively affects exports. Note that in a small price-taking economy, the reduced form equation effectively becomes the export supply equation (2). We draw on various data sources to construct a panel of data for 28 manufacturing sectors over the period 1970-2002. The data are mostly obtained from Quantec (2004), the World Development Indicators, the UNIDO INSTAT database, Statistics South Africa and the IMF International Financial Statistics. Further details are presented in the data Appendix. To capture the short-run dynamics, the long-run relationships (equations (2, 3 and 4) are embodied in an autoregressive distributed lag (ARDL) model. To capture the influence on exports of industry specific variables that are constant over time, sector specific effects (ηi) are included. The export functions are estimated using two estimators: A dynamic fixed effects (DFE) estimator and the “system” General Methods of Moments (GMM) estimator developed by Arrelano and Bover (1995) and Blundell and Bond (1998). 30 When using the GMM estimator, export volumes, prices and the exchange rate are treated as endogenous variables. In estimating the functions, the data are pooled and homogeneity is imposed for all parameters other than the sector fixed effects (ηi). A potential limitation arising from the pooling of the data is that it imposes common long-run relationships and short run dynamics across all sectors, which can give rise to misleading estimates of the coefficients (Pesaran and Smith, 1995). 31 To deal with the possible biases arising from parameter heterogeneity, export functions were estimated for a number of broadly defined manufacturing sub-sectors: Beneficiated, natural resource-based, metal and labourintensive products. 32
Empirical Results Export demand To test the sensitivity of the export demand function to the selection of foreign prices two data sources are used: (a) US import prices, obtained from the
30
See Pesaran and Smith (1995) for a discussion of the DFE and other estimators.
31
Alternative estimators are the Mean Group estimator of Pesaran and Smith (1995: 80) and the Pooled Mean Group Estimator of Pesaran, Shin and Smith (1999). The former allows for heterogeneity in short and long run coefficients, while the latter constrains the long run coefficients to be the same for each sector, but allows for short-run heterogeneity across sectors.
32
Beneficiated consists of iron & steel, chemicals and non-ferrous metals. Natural-resource based includes beneficiated products, paper products and food products (food, beverages & tobacco). Metal products include metal products, machinery & equipment, electrical machinery, motor vehicles and other transportation equipment. Labour-intensive products include textiles, wearing apparel, footwear, leather and furniture.
24
Bureau of Labour Statistics, and (b) a weighted average output deflator for developed countries, constructed from the UNIDO INSTAT (2001) database. The estimations using the foreign output price deflators are estimated over the periods 1970-99 and 1980-99 to enable comparisons with the results using the US import prices. Table 6 presents the estimated long-run average coefficients for the full sample of manufacturing industries and Table 6 to Tables 7-9 present the results for the subgroupings.
Table 6: Long-run average export demand coefficients for manufacturing Using US Import Price DFE
Using Foreign output deflator
GMM
DFE
1982-99
GMM
DFE
1970-99
Export
0.084 ns
0.000 ns
Exchange rate
1.000 ***
1.077 ***
-0.103 ** 1.205 ***
GMM 1980-99
-0.048 * 1.414 ***
Foreign price
0.934 **
0.435 *
0.948 ***
0.941 ***
Foreign output
-0.005 ns
0.077 ns
0.455 ***
0.581 ***
Adjustment term
-0.104 ***
-0.109 ***
-0.173 ***
-0.048 ***
-0.021 ns
-0.008 ns
0.916 ***
1.081 ***
1.186 ***
0.259 *
-0.136 ns -0.123 ***
0.035 ns -0.152 ***
Tests (H0) Exchange rate = 1
ns
ns
**
***
ns
ns
Foreign price =1
ns
**
Ns
ns
ns
***
Erate=Pfor
ns
ns
Ns
ns
ns
***
R2 F Obs AR(1)
0.57 16.06 *** 428
0.40 20239 ***
13.61 ***
428
0.54 33138 ***
812 ***
16.70 ***
11685 ***
560 ***
AR(2) ns ns Note: lag limit set to 10 for GMM estimations, except for the period 1970-99 where it is set to 5 (to solve problem of autocorrelation).
*** ns
The estimation results present supportive evidence for the “small country” hypothesis for total manufacturing when the sample is restricted to the 1980s and 1990s. The coefficient on export volumes and foreign output during this period is insignificantly different from zero and the result is robust to changes in the foreign price variable or the estimator (DFE or GMM). The results are less robust when the sample is extended to include the 1970s. The long-run average coefficients on exports and foreign output estimated using the DFE are significantly different from zero and imply a price elasticity of export demand of -10 and an income elasticity of export demand of 4, respectively. However, the GMM estimator suggests that when the endogeneity of export volumes and the nominal effective exchange rate are accounted, the export volume coefficient is only significant at the 10% level.33 Thus the small country hypothesis holds once endogeneity problems are dealt with. There is also strong evidence that the long run average coefficients on the nominal effective exchange rate and the foreign price variable are insignificantly 33
The implied price elasticity of demand and income elasticity of demand are -21 and 12, respectively.
25
different from each other and are equal to 1. The null hypothesis of equality of coefficients is only rejected in the GMM estimator results when using foreign output deflators over the period 1980-1999. In most cases it is also not possible to reject the hypothesis that the long-run coefficient on the exchange rate or foreign price equals 1. This provides strong evidence that domestic exporters are price-takers in the international market and hence that export prices rise by the full increase in foreign prices or the depreciation of the exchange rate. 34 These results are consistent with those found by Edwards and Wilcox (2003) for aggregate non-gold merchandise exports. Although some variation is found, the results for the sub-groups are broadly consistent with those for total manufacturing. 35 There is strong evidence that the metal product and labour-intensive industries are price-takers in the international market and therefore face an infinite demand for their products. In both these sub-groupings, the coefficient on export volumes is mostly insignificant. Amongst natural resource based industries, a significant negative coefficient on export volumes is found during the 1970-99 period (giving an implied price elasticity of export demand of -4 to -10), but this becomes insignificant once the endogeneity of export volumes and the exchange rate are dealt with and the sample is restricted to the 1980s and 1990s (see the GMM results). Similarly, in most cases the equality of the exchange rate and foreign price coefficients cannot be rejected. In cases where they diverge (mainly metal products) export prices are generally found to be more strongly affected by the exchange rate (equal to or greater than 1) than foreign prices. Overall, therefore, the results provide strong evidence that South African industries are price-takers in the international market. On average, export prices rise by the full depreciation of the rand and the increase in the foreign price. Various implications arise from these results. •
Firstly, export growth is not predominantly dependent on the economic prosperity of South Africa’s trading partners or on their ability to compete in the export market on the basis of price.
•
Secondly, this implies that export volumes are determined by the profitability of export supply. Factors that raise the output price received by exporter and reduce their cost of production will therefore enhance export performance.
•
Thirdly, exchange rate depreciations on average positively affect export performance by raising the profitability of export supply, and not by increasing the cost competitiveness of South African products. Exporters
34
The adjustment lag is relatively slow, suggesting that export prices adjust to correct 10% to 17% of the disequilibrium in the long-run equilibrium each year. However, the short-run coefficients suggest that between 21% and over 100% of the adjustment takes place within the same year, implying a relatively quick adjustment period.
35
The results for beneficiated products have been omitted as they are similar to that of natural resource based products.
26
raise their prices by the depreciation rate and do not, on average, lower the foreign currency price of their products in order to capture market share. 36 These implications do not imply that world demand and foreign market access are unimportant. While world demand does not directly affect export performance via the demand relationship, it affects export supply via its impact on world prices. Similarly, preferential reductions in foreign tariffs and market access will improve export performance if they raise the price received by exporters.
Table 7: Long-run average export demand coefficients for natural resource products Using US Import Price DFE
Using Foreign output deflator
GMM
DFE
-0.040 *
-0.250 **
1982-99 Export
-0.262 *
GMM
DFE
-0.097 ***
-0.278 **
1970-99
GMM 1980-99 -0.048 ns
Exchange rate
0.440 ns
0.575 **
0.809 ***
0.964 ***
0.421 ns
0.797 ***
Foreign price
1.227 ***
0.936 ***
0.763 ***
0.838 ***
1.275 ***
0.640 **
Foreign output
1.513 *
1.060 **
1.314 ***
1.140 ***
1.280 *
0.622 **
-0.111 ***
-0.137 ***
-0.109 ***
Adjustment term
-0.129 ***
-0.129 ***
-0.129 ***
Tests (H0) Exchange rate = 1
**
ns
ns
ns
**
ns
Foreign price =1
ns
ns
ns
ns
ns
ns
Erate=Pfor
ns
ns
ns
ns
ns
ns
R2
0.57
F
0.44
11.57 ***
Obs
177
61.78 *** 177
0.55
11.40 ***
99 ***
319.00
12.87 ***
773 ***
220.00
AR(1)
**
***
***
AR(2)
ns
ns
ns
Table 8: Long-run average export demand coefficients for metal products Using US Import Price DFE
Using Foreign output deflator
GMM
DFE
1982-99 Export
0.073 ns
0.055 ***
Exchange rate
1.398 ***
1.279 ***
Foreign price
-0.394 ns
GMM
DFE
1970-99
-0.537 *
Foreign output
-0.837 ns
-0.549 ns
Adjustment term
-0.209 ***
-0.195 ***
-0.106 ns
GMM 1980-99
0.038 ns
0.074 ns
1.096 ***
0.716 ***
0.935 ***
1.066 ***
1.270 ***
0.742 ns
0.551 ***
0.967 ***
-0.409 ***
-0.219 ***
0.029 ns 1.325 *** -0.792 ns
-0.079 ns
-0.699 *
-0.231 ***
-0.220 ***
Tests (H0) Exchange rate = 1
ns
ns
ns
**
ns
ns
Foreign price =1
**
***
ns
ns
ns
***
Erate=Pfor
**
ns
ns
**
ns
***
36
A further implication of this point is that a depreciation of the currency will improve (or rather will not worsen) the trade balance.
27
R2
0.62
F
0.50
11.38 ***
Obs
11.96 ***
126
126
0.58
10.81 ***
22 ***
203.00
10.74 ***
5 **
140.00
AR(1)
**
**
**
AR(2)
ns
*
**
Table 9: Long-run average export demand coefficients for labour-intensive products Using US Import Price DFE
Using Foreign output deflator
GMM
DFE
1982-99 Export
-0.012 ns
Exchange rate
1.620 ***
GMM
DFE
1970-99 0.027 ns
-0.158 ***
1.384 ***
1.163 ***
Foreign price
-0.119 ns
-0.072 ns
2.771 ***
Foreign output
-0.137 ns
-0.323 ns
0.254 ns
Adjustment term
-0.230 **
-0.266 ***
-0.587 ***
GMM 1980-99
-0.042 * 1.027 *** 1.232 ***
-0.019 ns
-0.004 ns
1.215 ***
1.293 ***
1.342 *
0.205 ns
-0.234 ns
-0.332 ns
-0.256 **
-0.423 ***
-0.378 ***
-0.493 ***
Tests (H0) Exchange rate = 1
*
ns
*
ns
ns
**
Foreign price =1
ns
ns
***
**
ns
**
Erate=Pfor
ns
ns
ns
ns
ns
**
R2
0.58
F
7.23 ***
Obs
0.48 5.42 *
72
72
8.29 ***
0.61 22 ***
145.00
10.28 ***
3 ns
100.00
AR(1)
*
**
**
AR(2)
ns
ns
ns
Export supply The export demand analysis identifies the importance of analysing factors that affect export supply. Some of the most important results arising from the estimation of the reduced form equation are presented here. Table 10 presents the results for total manufacturing, while Table 11 presents those for the industry subgroupings. Only the reduced form results using the foreign output deflator over the period 1970-99 are presented for the sub-groups, as the results using US import prices are similar. Table 10: Reduced form average coefficients for manufacturing Using US import prices 1982-1999 Coef. Long run Relative price Foreign output ULC
2.05 *** 1.38 *** 0.37 ns
Using foreign output deflators 1970-1999 1980-1999 Coef. Coef. 2.45 *** 1.20 *** -0.51 ns
28
1.81 *** 1.61 *** 0.11 ns
Output Output deviation Import penetration Rail capacity Fuel P Adjustment term Short run Relative price Foreign output ULC Output Import penetration Output deviation Rail capacity Other K/L Skill/Unskill Sanctions Adj R-squared F( 41, 410) Number of obs
0.54 -0.77 0.37 2.05 -1.38 -0.20
ns ns ** * *** ***
0.33 -0.28 0.23 1.95 0.20 -0.19
ns ns ns * ns ***
0.57 ** -0.75 ns 0.55 *** 2.34 *** -1.17 *** -0.25 ***
0.37 ***
*** *** * ** *** ** ***
0.32 ***
0.42 *** -0.61 *** 1.10 ***
0.77 1.22 -0.19 0.80 0.42 -0.47 1.78
0.50 *** -0.68 *** 1.29 ***
-0.07 ns 2.00 *** -0.06 *
-0.76 ** 3.02 *** -0.02 ns
-0.40 ns 1.98 *** -0.09 ***
0.27 5.05 *** 452
0.24 5.91 *** 784
0.30 6.35 *** 560
Note: The error correction parameterization of the ARDL model is estimated. Dummy variables are included for each sector, the MIDP programme (1995-2002), political turmoil in 1976 and the debt crisis (1984-86). Insignificant short run coefficients have been eliminated
Table 11: Reduced form average coefficients for sector groupings, 1970-99 using foreign output deflators Beneficiated Coef. Long run Relative price Foreign output ULC Output Output deviation Import penetration Rail capacity Fuel P Adjustment term Short run Relative price Foreign output ULC Output Import penetration Output deviation Rail capacity Fuel P Other K/L Skill/unskill Sanctions Adj. R2 F-statistic Obs
2.33 1.01 -0.18 0.28 -0.47 0.46 0.61 0.69 -0.20
Natural resource Coef.
Metal products Coef.
Labour-intensive Coef.
*** ns ns ns ns ns ns ns ***
1.51 1.48 -0.16 0.48 0.23 0.31 0.92 -0.26 -0.24
*** * ns ns ns ns ns ns ***
0.79 ns -0.14 ns 0.03 ns 0.62 * -1.27 *** 0.90 ns -0.79 ns -0.98 ** -0.36 ***
4.13 2.27 -1.00 0.57 -0.63 0.98 5.74 2.06 -0.36
*** *** ns ns *** ns ** *** ***
0.99 *** 1.08 *** -0.14 ns 0.70 ns 0.32 *** -0.54 * 1.81 *** -0.10 ns
0.66 1.18 -0.12 0.45 0.27 -0.51 1.59 -0.09
*** *** ns ns *** ns *** ns
0.24 ns 1.00 *** 0.00 ns 1.42 ** 0.79 *** -0.79 ** 1.36 *** -0.01 ns
1.39 1.20 -0.63 -0.19 0.61 0.08 1.87 0.22
*** ns * ns *** ns ** ns
-0.82 ** 2.98 ** -0.03 ns
-0.71 * 2.58 ** -0.04 ns
0.32 4.21 *** 196
0.20 3.38 *** 308
29
-3.35 ns 5.39 *** 0.02 ns 0.42 5.82 *** 196
-23.62 ns 10.25 * -0.08 ns 0.38 4.33 *** 140
Resource endowments and comparative advantage As shown earlier, the dependence on natural resource-based products and the failure to diversify sufficiently into products with growing world markets is an important source of South Africa’s poor aggregate manufacturing export performance during the 1990s. Resource-based products are declining as a share of world trade. These products are also characterised by a long term decline in prices relative to other products, although the world economy is currently experiencing a commodity price boom. The net effect of the existing trade structure and the inability to diversify sufficiently is that South Africa’s share of developing economy exports declined dramatically from 1.73% in 1988 to 1.07% in 2002. The dominance of natural resource-based products in South African exports and the failure to diversify significantly during the 1990s in part reflect South Africa’s rich endowment of natural resources, rather than a failure of policy. Wood and Mayer (2001), for example, show that Africa’s land abundance per worker, combined with its low level of skills, is an important determinant of its high share of primary products in total exports. Dependence on primary and natural resource-based exports also inhibits diversification as it can lead to a ‘Dutch Disease’ effect (Wood and Mayer, 2001), and results in volatile exchange rate movements (Sachs and Warner, 1997; Collier, 2001). 37 This volatility is detrimental to manufacturing and agricultural processing, which have a low share of factor costs in total costs and are therefore unable to absorb the negative impact of large price decreases (Collier, 2001). With its large natural resource endowments, South Africa is not immune from these effects. The entry of highly populated, low-income economies into the world market from the mid-1980s has also inhibited the diversification of middleincome countries such as South Africa into labour-intensive exports. UNCTAD (2002: 126) estimates that the share of low-skilled labour (adults with schooling up to secondary level) in world trade rose from 64% to 68% due to the participation of these economies in the world market. In addition, rapid skills accumulation in the Newly Industrialised Economies (NIEs) and the abundance of skilled labour in developed economies has reduced options for diversification into skill-intensive exports. South Africa’s natural resource endowment has thus led to a squeeze in competitiveness by both developed and NIEs and labour abundant developing economies. While natural resource endowments may inhibit diversification, the availability of skilled labour is shown to be an important determinant of diversification into manufacturing. The econometric estimates in Table 10 and Table 11 show a strong positive relationship between the skill-intensity of production
37
The rents generated by primary commodities have also been associated with poor governance and substantially higher risk of civil war, both of which negatively affect economic growth (Sachs and Warner, 1997)
30
and export performance, with coefficients for aggregate manufacturing in excess of 2. The coefficient on the skill-unskilled ratio for metal products, which fall in the high technology group of products, is a high 5.4. The ability to diversify into high technology sectors is therefore strongly influenced by the availability of well educated labour. The importance of skilled labour for the diversification of exports into manufacturing within Africa has also been shown by Wood and Mayer (2001). Relative prices and competitiveness of manufacturing Foreign prices, domestic prices and the exchange rate have a strong impact on manufacturing export performance in South Africa. This is shown by the positive and significant coefficient on the relative price variable (the real effective exchange rate) in the reduced form results presented in Table 10 as well as Figure 9 38 39 A 1% increase in the relative price of exports is estimated to raise average manufacturing export volumes by 1.8% to 2.5% in the long-run. The very elastic response of export volumes to changes in relative prices found in these estimates is much larger than the estimate (0.63) by Fallon and da Silva (1994), but is similar to the results (1.6 to 2.8) of Edwards and Golub (2004). Support for the vent-for-surplus hypothesis is also found, but its importance in determining export growth is diminishing. Declines in output from the long-run trend are found to positively affect aggregate manufacturing exports, at least in the short run. As shown in Table 10, the short-run coefficients on the output deviations variable range from -0.47 to -0.68. The vent-for-surplus relationship helps to explain why export growth continued during the late 1980s (see Figure 9), despite the real appreciation of the currency. The rapid growth in exports from the early 1990s reflects a diminished importance of vent-for-surplus exports, as well as a onceoff adjustment to the ending of sanctions and the re-integration of South Africa into the world economy. 40
38
Note that in this specification, price homogeneity is imposed and foreign prices (P*), domestic prices (P) and the exchange rate (e) are combined into a single relative price variable (eP*/P). In many cases, but not all, the restriction of homogeneity could not be rejected.
39
The use of the real effective exchange rate as a measure of South African export competitiveness is widespread in South African empirical analysis (IMF, 1998, Kahn, 1998, Walters and de Beer, 1999, and Golub, 2000). Despite the sensitivity of this measure to the choice of price indices and weights, most estimates suggest that South Africa’s competitiveness improved from the mid 1990s. The recent appreciation of the Rand, however, has reversed much of this gain.
40
The coefficient on the sanctions dummy (1986-92) is significant and negative for aggregate manufacturing when the period is restricted to the 1980s and 1990s (Table 10).
31
Figure 9: Real effective exchange rate and exports in aggregate manufacturing
Export orientation
0.25
140 120 100 80 60
0.20 0.15 0.10
40 20 0
0.05
Export orientation
2001
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
1977
1975
0.00
REER (eP*/P , 1995=100)
Aggregate manufacturing
REER
Note: The REER is constructed using the weighted average foreign output price deflators, the weighted average nominal exchange rate and the domestic producer price index. Export orientation is the value of exports in gross output valued in 1995 prices.
The highly elastic response to relative prices for aggregate manufacturing is also found within all sub-groups apart from metal products (Figure 10). 41 The long-run coefficient on relative prices in the reduced form results for beneficiated and natural-resource based products range from 1.5 to 2.3 (Table 11). Much of this appears to be driven by the positive impact of exchange rate depreciations on exports of these products. Labour-intensive products appear to be particularly sensitive to relative prices in the long-run (4.13). In contrast, exports of metal products are not sensitive to changes in relative prices. This is most clearly shown in Figure 10 during the period 1987 to 1995 when the REER fell, but export orientation rose very strongly. Three factors may explain the insensitivity of metal product exports to relative prices. •
41
Firstly, exports of these products are very sensitive to domestic demand conditions. The level of exports is weakly correlated with the level of output (the long-run coefficient of 0.67 is significant at the 10% level) and is very strongly affected by business cycles. A decline in output below its long-run trend raises exports by 1.3% in the long-run. The correlation with output growth
We also find strong evidence that the sectoral composition of exports is determined by relative prices across sectors. This provides support for the argument of Bell et al. (1999) that commodity price booms raise the share of natural resource-based products in total exports.
32
may reflect the development of technological capabilities required for the production of these products prior to entering the export market. 42 •
Secondly, exports of metal products are strongly biased towards the regional market and reflect advantages arising from regional proximity and the ability to offer after sale service (repair, operator training and engineering support) rather than cost competitiveness in production. 43 44 Thus, many of the regional exports contain little domestic content as is reflect in the very high import content of these products (60% for machinery and over 80% for scientific and professional and communication equipment). Additional evidence of this effect is shown by the high positive short run coefficient on import penetration (0.79) for metal products in Table 11.
•
Thirdly, export performance within the motor vehicle industry during the 1990s has largely been driven by industrial policies such as the MIDP, rather than adjustments in response to international price movements.
Figure 10: Export performance and the Real Effective Exchange Rate by subgroup Metal products
0.20 0.15 0.10 0.05
REER
2001
1999
1997
1995
1993
1991
1989
1987
REER
Export orientation
2001
1999
REER (eP*/P , 1995=100)
0 1975
2001
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
1977
20 0.00 1997
0
0.00
40
1995
20
1993
0.05
60
0.05
1991
40
80
0.10
1989
60
1987
0.10
100
1985
80
1981
0.15
120 0.15
1979
100
1977
0.20
140
0.20 Export orientation
120
REER (eP*/P , 1995=100)
140
1975
REER
Labour-intensive products
0.25 Export orientation
1985
Export orientation
Natural resource-based products
Export orientation
1983
1981
1979
1977
1975
0.00
REER (eP*/P , 1995=100)
Export orientation
REER (eP*/P , 1995=100)
160 140 120 100 80 60 40 20 0
0.25
1983
Export orientation
0.30
2001
0 1999
0.00 1997
20 1995
40
0.05 1993
0.10
1991
60
1989
0.15
1987
80
1985
100
0.20
1983
0.25
1981
120
1979
140
0.30
1977
0.35
1975
Export orientation
Beneficiated products
REER
42
Goldstein and Kahn (1985:1060) note that, “secular changes in the level of aggregate real output will be accompanied by advances in factor supplies, infrastructure, and total factor productivity that will lead to an increase in export supply at any given level of export prices.”
43
Africa accounts for 31% of total exports of these products, but only 18% of total SA manufacturing exports.
44
After sale service has become an important competitive tool involving repair, operator training, and engineering support (Gourevitch et al., 2000).
33
Firm level survey data provides further evidence of the effect of exchange rate depreciations on export supply. Drawing on a survey of large firms (more than 50 employees) located in the Greater Johannesburg Metropolitan Area (GJMA), Chandra et al. (2001) find that a high percentage of firms raised exports in response to the Rand crisis of 1998. The positive response was particularly strong amongst firms employing between 200 and 10,000 workers, where 80% of firms exported more. Less than 45% of small firms (50 to 100 employees) raised exports. Many of these firms (30%) were uncertain about the exchange rate depreciation and decided to wait and see prior to altering production or substituting away from imports. Exchange rate volatility therefore appears to have negative impact on export supply, particularly within small firms who are unable to bear the cost of fluctuations in profitability. These results thus provide considerable support for the importance of a stable and competitive real effective exchange rate in driving export performance. Although the real depreciation of the Rand during the 1990s has contributed extensively towards growth in manufacturing exports during this period, the volatility of exchange rate fluctuations has potentially deterred new entrants and given rise to a more muted response than otherwise would have been the case. This may have contributed to the poor export response relative to other developing economies. Variable costs of production The profitability of export supply is dependent both on output prices as well as the variable costs of production. In the econometric analysis of the determinants of export supply, variable production costs are captured by unit labour costs (ULC) and producer prices. The negative impact of domestic price inflation on export performance has already been shown in the analysis of real exchange rates. Domestic price inflation causes a real appreciation of the currency and hence a decline in export performance. An implication drawn from this is that nominal depreciations of the currency do not lead to long-term increases in export performance if the gains in competitiveness are eroded by domestic inflation and wage increases (Edwards and Wilcox, 2003). ULC is found to be insignificant in all but labour-intensive sectors where a long-run coefficient of -1.2 is estimated in the export supply function (not shown). This is consistent with expectations as natural-resource based products and metal products are less dependent on labour costs as a determinant of the profitability of export supply. Edwards and Wilcox (2003) also find a significant negative coefficient on ULC in their estimates of the export supply function. The importance of labour costs competitiveness has also been shown by Edwards and Golub (2004) who analyse South Africa’s unit labour costs relative to a range of developed and developing economies. They find that South African labour cost competitiveness improved over the 1970-1998 period for most sectors, 34
with particularly large improvements in the late 1990s (see the unit labour cost bar Figure 11). 45 Their econometric results show that relative unit labour costs, relative productivity and relative wages are important determinants of South African manufacturing export performance. As found in this study, labour-intensive sectors are more strongly affected by labour cost competitiveness than natural resource intensive sectors. 46 Despite the gains in labour cost competitiveness during the 1990s, South African relative unit labour costs in aggregate manufacturing still exceed those of other developing economies. This will have contributed towards the slow growth in manufacturing exports relative to developing economies. Improved export growth, particularly within labour-intensive sectors, can be enhanced by policies that improve labour productivity combined with wage moderation.
45
Much of the gain in competitiveness during the 1990s is due to the depreciation of the rand, although South Africa’s relative labour productivity also improved in the late 1990s (Edwards and Golub, 2004).
46
They estimate a long-run coefficient on Relative Unit Labour Costs of -1.9 for natural resource intensive, -2.8 for labour-intensive and -1.78 for chemical products.
35
Figure 11: South African wages, productivity and unit labour costs in total manufacturing a. Against Developed Countries 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 1970-79
1980-89
1990-94
1995-98
b. Against Developing Countries 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 1970-79
1980-89 wages
productivity
1990-94
1995-98
unit labor cost
Source: Edwards and Golub (2004).
Infrastructure Economic infrastructure such as transport, communication, power, water and sanitation systems provide the foundation for economic activity within an economy. The provision of infrastructure also has important consequences for an economy’s export performance. •
Firstly, the provision of economic infrastructure raises economic growth (DBSA, 1998) and therefore indirectly exports. This indirect relationship is supported by the positive relationship between exports and trend output found in this study.
•
Secondly, good infrastructure lowers the transaction costs associated with exporting.
•
Thirdly, good infrastructure facilitates the diversification of export production (Elbadawi, 1999; Collier, 2002).
•
Fourthly, lower transport costs can give rise to powerful forces for agglomeration (Redding and Venables, 2004) and contribute towards the 36
development of international production networks in world trade (Mayer, 2003). Six different variables are used to assess the impact of infrastructure availability on export performance. These variables are rail carrying capacity (tonnes), public-sector infrastructure fixed capital stock (R million, 1995 prices) 47 , roads paved (km), the share of total roads paved, electricity generated (gigawatts), electricity, water and gas fixed capital stock, communication fixed capital stock and telephone mainlines per 1000 people. 48 Figure 12 to Figure 14 present trends in these various indicators of infrastructure from 1970. Many of the infrastructure variables are expressed as a ratio to real manufacturing GDP. As shown in Figure 12, public sector capital stock, which rose strongly from the 1960s to the early 1980s, has failed to keep pace with the recovery in manufacturing output growth during the 1990s. The downward trend during the 1990s has occurred despite the expansion programmes by Telkom and Eskom to extend telephone lines and electricity to areas which were under-serviced. Similarly, the collapse in railway infrastructure investment from the 1980s led to a decline in the number of locomotives, coaching stock and goods stock and an eventual decline in goods stock carrying capacity from the late 1980s. The paving of national and provisional roads, however, roughly kept pace with manufacturing GDP growth. The strongest growth in infrastructure capital stock has occurred in communications infrastructure. Telephone mainlines have grown consistently from the early 1900’s, rising from 41 telephones per 1000 people in 1970 to 112 in 2001. The growth in the roll-out of fixed lines accelerated in 1997 with the 5-year capital expansion programme required in terms of Telkom’s service licence, but nonpayment of telephone bills and competition from mobile market has led to many of these new connections being disconnected. Not presented in Figure 14 is the very rapid rise in mobile during the 1990s. Despite the rapid growth in communication infrastructure, South Africa’s telecommunications prices compare very poorly with international best practice (Truen, 2005). Compared to its developing-country peers, South African telecommunication prices are found to be the most expensive in seven of 10 product categories analysed. SA’s cost disadvantage also extends to cell telephony with business cellphone calls the second most expensive in the sample analysed by Truen (2005). 49 South Africa also lags many middle-income countries in terms of the number of fixed lines per 1000 people (139 for lower middle income and 208 for
47
Public-sector investment consists of investment by the government and public corporations such as Transnet (rail and air transport services), Eskom (electricity) and Telkom.
48
The data are sourced from Perkins (2003), Quantech (2004) and the World Development Indicators Paved roads and share roads paved are updated using data obtained from World Development Indicators.
49
The countries include: Canada, Hong Kong, Israel, Norway, Singapore, Sweden, South Korea, US, Brazil, India, Malaysia, Morocco, the Philippines and Thailand.
37
upper middle income countries in 2001) and the number of telephone faults per fixed lines (40 faults per 100 fixed lines in 2000) (Perkins, 2003).
Figure 12: Public-sector infrastructure capital stock
Capital stock/Manufacturing GDP
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
Infrastructure capital stock (Rbill, 1995 prices)
100 50 0
1980
0.0 0.0 0.0
1978
350 300 250 200 150
1976
0.0 0.0 0.0 0.0 0.0
1974
500 450 400
1972
0.0 0.0 0.0
1970
Capital stock/Manufacturing GDP
Public-sector infrastructure capital stock
Capital stock, level
Note: Public-sector infrastructure fixed capital stock includes transport, communication, power, water and sanitation systems.
Figure 13: Transport infrastructure to manufacturing GDP
90
0.6
80 70
0.5
60
0.4
50
0.3
40 30
0.2
20
Paved roads
Rail carrying capacity
38
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
1980
1978
0
1976
0.0
1974
10 1972
0.1
Rail carrying capacity (Tonnes/GDP)
0.7
1970
Paved roads (km/GDP)
Transport infrastructure
Figure 14: Communication infrastructure
140.00
1.00
120.00 100.00
0.80
80.00 0.60
60.00
0.40
40.00
Communication K stock
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
1980
1978
0.00
1976
0.00
1974
20.00 1972
0.20
Telephone mainlines (per 1000 people)
1.20
1970
Communication capital stock (Rmill/manufacturing GDP)
Communications infrastructure
Telephone mainlines
The econometric analysis shows that access to good rail infrastructure is an important determinant of manufacturing export performance. The short-run coefficient on rail carrying capacity is positive and significant for all the reduced form regression results shown in Table 10 and Table 11. The long-run coefficients are positive and significant for aggregate manufacturing and range from 2 to 4.7, suggesting a high responsiveness in manufacturing exports to rail carrying capacity. Public-sector infrastructure fixed capital stock, paved roads, electricity generated and electricity, gas and water fixed capital stock also positively affect export performance, but primarily in the short run (Table 12). A 1% increase in public-sector infrastructure capital relative to manufacturing GDP is shown to raise average manufacturing exports by 2.4% in the long-run. An increase in the price of fuel relative to producer prices is also found to negatively affect export performance, particularly during the 1980s. This relationship, however, is not consistent across the industry groupings. No significant results are found for communication related infrastructure. Table 12: Impact of various infrastructure variables on export performance, reduced form results Infrastructure variables as ratio of total manufacturing GDP Rail carrying capacity (tonnes) Long-run Short-run Public-sector infrastructure fixed capital stock (Rmill 1995 prices) Long-run Short-run Roads paved (km) Long-run Short-run Roads paved, share Long-run Short-run
39
Coefficient 1.95 * 1.78 ** 2.37 ** 1.89 *** 1.62 ns 1.70 *** -0.70 ns -1.62 *
Electricity generated (gigawatt hours) Long-run Short-run Electricity, gas and steam capital stock (Rmill 1995 prices) Long-run Short-run Communication capital stock (Rmill 1995 prices) Long-run Short-run Telephone mainlines (Unit: per 1,000 people) Long-run Short-run
0.66 ns 2.07 *** 0.75 * 1.27 *** 0.03 ns 0.16 ns 0.12 ns 0.36 ns
Note: All variables, except telephone mainlines, are expressed relative to total manufacturing GDP. All infrastructure variables are in natural logarithms. The coefficients are obtained from the reduced form specification using foreign output deflators over the period 1970-99. As shown in Table, the long-run coefficient on rail carrying capacity is significant at the 5% level when the sample is restricted to the 1980s and 1990s. The coefficient on electricity capital stock (1.77) is also significant in the long-run during the period 1980-99.
The importance of infrastructure for export performance is also revealed in manufacturing firm-level data. Figure 15 and Figure 16 present the percentage of firms finding telecommunications, postal services and transport services an obstacle to their operations. The data are obtained from the National Enterprise Survey conducted in late 1999 (Gelb, 2002). The NE survey is national in coverage and consists of 941 manufacturing firms, 39 % of which are large firms consisting of more than 50 employees. A relatively high percentage of exporters relative to domestic-oriented firms find that the cost, reliability and speed of communication, postal and transport services are an obstacle to their operations. This is particularly evident in relation to freight handling in harbours, airports, airlines and railways where up to 44% of exporters found the cost and reliability of these services problematic (Figure 16). 50 Poor port infrastructure is primarily a problem for manufacturing exports as primary exports (coal and iron ore) are channelled through specialised ports such as Saldahna and Richard’s Bay. These results provide strong evidence that poor infrastructure investment from the mid-1980s has curtailed the growth of manufacturing exports. Noncommunication infrastructure investment has failed to keep pace with manufacturing GDP growth. Export supply functions and firm surveys suggest that exporters are more susceptible to poor transport infrastructure than domestic-oriented firms. Improvement in infrastructure is therefore essential if manufacturing export growth is to be raised.
50
The differences are significant at the 5% level.
40
Figure 15: Percentage firms finding telecommunications and postal services an obstacle to their operations, National Enterprise Survey, 1999
70 60 50 40
Non-exporters
30
Exporters
20
Reliability/speed of internet access
Cost of internet access
Reliability/speed of telephone services
Cost of international telephone calls
Cost of local/national telephone calls
Cost of power supplies
Reliability of power supplies
0
Cost of postal services
10 Reliability/speed of postal services
% firms experiencing obstacles to operation
Telecommunications, power and postal infrastructure constraints
Note: Responses are significantly different for the reliability/speed of postal services and the cost of power supplies, international telephone calls and internet access.
Figure 16: Percentage firms finding transport services an obstacle to their operations, National Enterprise Survey, 1999
% firms experiencing obstacles to operation
A. Reliability and speed of freight handling 45 40 35 30 25
Non-exporters
20
Exporters
15 10 5 0 Harbours
Airports
Airlines
Railways
Road trans port
% firms experiencing obstacles to operation
B. Cost of freight or freight handling 50 45 40 35 30
Non-exporters
25
Exporters
20 15 10 5 0 Handling s ervices at harbours
Sea freight
Airlines
Railways
Road trans port
Notes: Responses are significantly different from each other for all but the reliability and speed of road transport. Firms could select between four options: obstacle, neutral, beneficial and not relevant.
41
Tariffs and export support measures The decline in nominal and effective protection rates during the 1990s will have had two important effects on export production. •
Firstly, lower tariffs reduce the cost of imported intermediate and capital goods and thus raise the profitability of export production.
•
Secondly, lower effective protection rates reduce the incentive to produce for the domestic market relative to the export market. Together lower intermediate costs and lower effective protection rates reduce the antiexport bias of the trade regime.
We find some evidence that tariff liberalisation has improved export performance. Using simple cross-sector regressions, we find that sectors experiencing high average protection (nominal and effective), implicit export taxes from tariff protection and anti-export biases have on average lower levels of export orientation. Only the relationship between export orientation and effective protection rates is significant at the 5% level. These relationships do not show causality as other factors such as comparative advantage may account for the high protection rates and low levels of export orientation in some sectors. Nevertheless, they are suggestive of a negative impact of tariff protection on export performance. To assess the impact of changes in protection and the anti-export bias on export performance over time, we re-estimate the reduced form export functions using a static fixed effects model over the period 1988-2002. Significant (at 5% level) negative coefficients are found on the export tax variable (-0.57) and the effective protection rate variable (-0.003). 51 Tariff reductions have therefore boosted export performance by reducing intermediate input costs as well as lowering the incentive to produce for the domestic market. Import penetration, which can be interpreted as a proxy for the impact of tariff liberalisation on export performance, is also found to positively affect aggregate manufacturing exports. Import penetration affects exports by (a) improving access to imported intermediate inputs, (b) enhancing productivity growth (Jonnson and Subramanian, 2000) and lowering mark-ups and hence the relative incentive to produce for the domestic market (Fedderke et al. 2003, Edwards and van de Winkel, 2005). The econometric estimates show that a 1% rise in import penetration raises aggregate manufacturing exports by 0.37% to 0.55% in the longrun. 52 Although the long-run coefficients are not significant for the sub-groupings, the short-run coefficients are significant and suggest that a 1% rise in import penetration is concurrently associated with a 0.27% to 0.79% increase in export growth.
51
Most of the other coefficients were insignificant. Trend output and deviation from trend output were generally significant and of the expected sign.
52
This result is consistent with that of Kahn and Knight (1988) who find a negative relationship between the cost of imported intermediate inputs and export performance in a pooled sample of 34 developing countries (including South Africa).
42
4. Conclusions and Policy Recommendations This study yields a number of insights regarding the composition and determinants of manufacturing export growth in South Africa South African export growth during the 1990s is shown to be mediocre compared to many developing economies and other similar natural resource abundant economies. Exports also remain concentrated in natural resource-based products and products with a declining share of world markets. In contrast, East Asian economies have successfully restructured production towards dynamic high technology products. The inability to re-structure exports towards these dynamic high technology products is one explanation for the relatively poor export performance of South Africa during the 1990s. The relatively poor manufacturing export growth can, in part, be attributed to South Africa’s comparative advantage in natural resource-based products that have declined as a share of world trade. However, South African manufacturing export performance has been poor even relative to other similar resource-based economies. This suggests that South Africa has not fully exploited all opportunities to expand exports. This study finds that foreign export demand in terms of market access is not a constraint to export growth. South African manufacturing exporters are predominantly price-takers in the international market and face an infinite demand for their products. Changes in foreign demand, however, impact upon export supply through changes in the international price. The constraint to growth lies on the side of export supply. Manufacturing exporters have responded positively to the real depreciation of the currency and trade has dampened the response by exporters and inhibited further diversification of exports. Declining infrastructure investment, particularly in transport infrastructure (ports, railways and roads), has dampened the response of manufacturing exports to the more favourable trade environment in the 1990s. Other factors such as the availability of skills are also shown to be important determinants of export growth, particularly within the high technology metal products sectors. These findings carry various policy implications if South African policymakers wish to accelerate manufacturing export growth. (i)
First, improving the rail, harbour and road infrastructure is of paramount importance if export growth, particularly of high technology products, is to improve. The proposed government strategy to invest R180-billion on transport logistics, electricity and water resources is a positive step in this direction. Recent work on infrastructure and 43
growth in South Africa (e.g., Bogetić and Fedderke, 2005a,b) also suggests that there are substantial infrastructure investment needs in the coming years. (ii)
Second, human capital accumulation needs to be fostered as this both enhances export growth and diversification into high technology products. Investments in education, especially in terms of quality (particularly maths and science), will also boost productivity and output growth (Fedderke, 2005) and thereby export performance. Much has been accomplished to raise the enrolment rates, but learning outcomes and skills build-up have not matched expectations relative to the needs of the needs of the formal and export oriented sectors, especially those in high technology products.
(iii)
Third, investments in R & D deserve more scrutiny by both public and private sectors. As shown elsewhere (Fedderke, 2005), total factor productivity gains, while rising, have been insufficient to offset sluggish contributions of factor accumulation and propel South Africa’s growth to a higher trajectory. Increased R & D and more favourable environment for technological innovation with appropriate incentives schemes could, potentially, help develop new products, technologies and markets. We believe that this is an area that South Africa’s private and public sectors could look more systematically, and in partnership, as an important source of longer term growth (see below).
(iv)
Fourth, a stable, competitive real exchange rate is an important tool to encourage manufacturing export growth. The increasing credibility of the central bank’s inflation targeting framework and very low and stable inflation in the past two years have established a solid policy basis for a more stable nominal and real exchange rates in the future.
(v)
Fifth, further simplifications of the tariff regime are possible. Although tariffs are already relatively low, the number of different tariff rates (102 in 2004, including non-ad valorem tariffs) is high and considerable variation of tariff rates at the 8-digit HS level remains. Although the ability to reform the tariff structure is constrained by the numerous trade agreements currently being negotiated, it is possible to at least simplify the MFN schedule by unifying tariff levels within the 4-digit HS code and reducing the number of tariff rates to 6, as agreed with the WTO. This simplification is particularly important, as future bilateral trade negotiations may entrench the existing complexity of the tariff structure.
(vi)
Sixth, thought could be given to carefully designed incentives provided only to “new” products and activities aimed at further diversifying the economic export base. Any such incentives would need to be crafted carefully towards specific activities, not sectors, with clear benchmarks/criteria for success or failure, and built-in sunset clauses, and partnerships between the public and private sector in order to facilitate the process of “discovery” of new export opportunities and new products (Rodrik, 2004). Importantly, this is not to argue for the traditional industrial policy for “picking winners”. It just stresses the need for the public and private sector in South Africa to engage in a more innovative and intensive processes of learning by doing in order to develop new 44
products and technologies, and to enter new markets in a similar fashion to Chile and the salmon industry, for example, or Taiwan and the orchid industry. South Africa could learn from these and similar international success stories of export diversification and development of new markets, which sometimes may require more than just relying on stable real exchange rates and trade liberalization alone.
References Allenye, T. and Subramanian, A. 2001. What does South Africa’s Pattern of Trade say about its Labor Markets? International Monetary Fund Working Paper WP/01/48. Altman, M. 1994. An Industrial Strategy for the Clothing Sector, University of Cape Town Press, Cape Town. Arellano, M. and Bover, O. 1995. Another look at the instrumental variable estimation of error-components models. Journal of Econometrics, 68: 29-51. Arellano, M. and Bond, S. 1991. Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The Review of Economic Studies, 58: 277-97. Arize, A. 1987. The Supply and Demand for Imports and Exports in a Simultaneous Model. Applied Economics, 19, 1233-1247. Balassa, B. 1965. Tariff Protection in Industrial Countries: An Evaluation. Journal of Political Economy, 73: 573-594. Baldwin, R. Determinants of the commodity structure of U.S. trade. American Economic Review, 61 (March): 126-46. Behar, A and Edwards, L., 2004. Estimating elasticities of demand and supply for South African manufactured exports using a vector error correction model. Centre for the Study of African Economies Working Paper Series, WPS/200404. Bell, T., Farrell, G. and Cassim, R. 1999. Competitiveness, International Trade and Finance in a Minerals-rich Economy: The case of South Africa. Paper presented at 1999 TIPS conference. Bell, T. 1992. Should South Africa further liberalise its foreign trade? Economics Trends Working Paper no. 16, Department of Economics and Economic history, Rhodes University. Bell, T., 1997. Trade policy, in Michie. J. and V. Padayachee, V. (eds.), The Political Economy of South Africa’s Transition, Dryden press, London. Bell, T. and Madula, N. 2001. Where has all the growth gone? South African manufacturing 1970-2000, paper presented at the Trade and Industrial Policy Strategies Annual Forum, Muldersdrift, September 2001. Belli, P., Finger, M. and Ballivian A. 1993. South Africa: A review of trade policies. World Bank Informal Discussion Papers on Aspects of the South African
45
Economy no. 4. The World Bank Southern Africa Department: Washington DC. Berman, E., Bound, J. and Machin, S. 1998. Implications of skill-biased technological change: International evidence. Quarterly Journal of Economics, 113, 4: 124579. Bhorat, H. and Hodge, J. 1999. Decomposing shifts in labour demand in South Africa. South African Journal of Economics, 67(3): 348-380. Bhorat, H. 1998. Income and price Elasticities in Manufactured Exports. Working Paper 13, Development Policy Research Unit. Blundell, R. and Bond, S. 1998. Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics, 87: 115-43. Bogetić, Ž and J. W. Fedderke (2005a). International Benchmarking of South Africa's Infrastructure Performance, Journal of Development Perspectives (South Africa), Vol.1, No.2, 2005, forthcoming. Bogetić, Ž and J. W. Fedderke (2005b). Infrastructure and Growth in South Africa: Benchmarking, Productivity, and Investment Needs, bi-annual conference of the Economic Society of South Africa (ESSA), Durban, September 7-9, 2005. Chandra, V., Moorty, L., Rajaratnam, B. and Schaefer, K. 2001. Constraints to growth and employment in South Africa: Statistics from the large manufacturing firm survey. World Bank Informal Discussion Papers on Aspects of the South African Economy no. 14. The World Bank Southern Africa Department: Washington DC. Collier, P. 2002. Primary Commodity Dependence and Africa’s Future, World Bank, Washington. DTI. 1997. Annual report 1996-97. Pretoria: Department of Trade and Industry. Edwards, L. and Golub, S. 2004. South Africa’s International Cost Competitiveness and Productivity in Manufacturing. World Development, 32, 8: 1323-1339 Edwards, L. 2001. Globalisation and the occupational structure of employment in South Africa. South African Journal of Economics, 69(1): 40-71. Edwards, L. and Schoer, V. 2002 Measures of competitiveness: A dynamic approach to South Africa’s trade performance in the 1990s. South African Journal of Economics, 70, 1008-1046. Edwards, L. and van de Winkel, T. 2005. The Market Disciplining Effects of Trade Liberalisation and Regional Import Penetration on Manufacturing in South Africa. Paper prepared for Trade and Industry Policy Strategies. Edwards, L. and Wilcox, O. 2003. Exchange rate depreciation and the trade balance in South Africa. Paper prepared for the National Treasury, 2003. Elbadawi, I. 1999. Can Africa Export Manufactures? The Role of Endowment, Exchange Rates and Transaction Costs. Development Research Group and Africa Region, World Bank, Washington. Elbadawi, I. and Helleiner, G. 1998. African Development in the Context of New World Trade and Financial regimes: The Role of WTO and Its Relationship to
46
the World Bank and the IMF, Presented at AERC Workshop on Africa and the New World Trade System, Mombasa, Kenya, April 27-29. Fallon, P. and Pereira da Silva, L. 1994. South Africa: Economic performance and policies, World Bank Informal Discussion Papers on Aspects of the South African Economy no. 7, The Southern Africa department, The World Bank. Fallon, P. 1992. An analysis of employment and wage behaviour in South Africa, World Bank Informal Discussion Papers on Aspects of the South African Economy, The Southern Africa department, The World Bank. Fedderke, J. 2005. South Africa: Sources and constraints of long-term growth. Mimeo. Fedderke, J. and Vase, P. 2001. The nature of South Africa’s trade patterns by economic sector, and the extent of trade liberalisation during the course of the 1990’s. South African Journal of Economics, 69, 3: 436-473. Fedderke, J. and Vaze, P. 2004. Response to Rangasamy and Harmse: Trade liberalisation in the 1990s. South African Journal of Economics, 72, 2: 407412. Fedderke, J., Kularatne, C. and Mariotti, M. 2003. Mark-up Pricing in South African Industry, ERSA, University of the Witwatersrand. Fedderke, J. and Szalontai, G. 1995. Industry concentration in South African manufacturing: Trends and consequences, 1972-96. Mimeo Gelb, S. 2002. Foreign companies in South Africa: Entry, performance and impact. Report by the Edge Institute, South Africa. Goldstein, M. and Khan, M.S. 1985. Income and price effects in foreign trade, in R. Jones and P. Kenen (eds), Handbook of International Economics, vol. II, North-Holland, Amsterdam. Golub, S.S. 2000. South Africa’s international cost competitiveness. Paper presented at the TIPS annual forum, 18-20 September, 2000. Golub, S.S. and Ceglowski, J. 2001. South African real exchange rates and manufacturing competitiveness. South African Journal of Economics, 70, 6: 1047 – 1075. Gourevitch, P., Bohn, R. and McKendrick, D. 2000. Globalization of production: Insights from the hard disk drive industry. World Development, 8, 2, 301-317. Greenaway, D. and Milner, M. 1993. Trade and Industrial policy in developing countries, Macmillan Press, Basingstoke, Hampshire. Hayter, S., Reinecke. G. and Torres, R. 1999. South Africa: Studies on the social dimensions of globalisation. Task force on country studies on globalisation, International Labour Office: Geneva. Hodge, J. and Miller, J. 1996. Information technology in South Africa. Paper presented at the IT and Exclusion in Developing Countries conference, UNU/INTECH, Maastricht, October 1996. Holden, M. 2001. Effective protection revisited: How useful a policy tool for South Africa? South African Journal of Economics, 69, 4: 659-673.
47
Holden, M. 1992. The structure and incidence of protection in South Africa, in Black. P. and B. Dollery (eds), Leading Issues in South African Microeconomics, Southern book publishers, Johannesburg. Holden, M. and Holden, P. 1975. An intertemporal calculation of effective rates of protection for South Africa. South African Journal of Economics, 43, 3: 37079. IDC. 1990. Ondersoek na die Tariefbeskermingsbeleid: Ontleding fan die Tariefstruktuur, Industrial Development Corporation, South Africa IMF. 1998. Country report - South Africa. International Monetary Fund, Washington, D.C. Jenkins, C., Bleaney, M., Holden, M. and Siwisa, N. 1997. A review of South Africa’s trade policy, paper presented at the Trade and Industrial Policy Annual Forum, Muldersdrift, September 1997. Jonsson, G. and Subramanian, A. 2000. Dynamic gains from trade: Evidence from South Africa. International Monetary Fund Working Paper no. 00/45. Kahn, B. 1998. Assessing South Africa’s Competitiveness: Is the Reserve Bank’s Measure Misleading, London School of Economics Centre for Research and Finance in Southern Africa Quarterly Review, no. 4. Kahn, B. 1998. Assessing South Africa’s competitiveness: Is the Reserve Bank’s real exchange rate measure misleading? LSE Centre for Research and Finance in Southern Africa Quarterly Review, number 4, 1998. Kahn, S.B. 1987. Import penetration and Import Demands in the South African Economy. South African Journal of Economics, 55(3), 238-247. Kaplinsky, R. 1995. Capital intensity in South African manufacturing and unemployment, 1972-90. World Development, 23, 2: 179-192. Khan, M. and Knight, M. 1988. Import compression and export performance in developing countries. The Review of Economics and Statistics, 70 (2): 315321. Kuhn, G. and Jansen, R. 1997. The effective protection rate and anti-export bias. IDC Research Paper Series, (TS1/97), Industrial Development Corporation, Sandton. Lall, S. 2000. The Technological Structure and Performance of Developing Country Exports, 1985-1998. QEH Working Paper 44, Queen Elizabeth House, Oxford University. Lall, S. and Kraemer-Mbula, E. 2005. Is African Industry Competing? QEH Working Paper 121, Queen Elizabeth House, Oxford University. Lewis, J. 2001. Reform and opportunity: The changing role and patterns of trade in SA and SADC, Africa Region Working Paper Series, no. 14. World Bank. Mayer, J. 1997. Is having a rich natural-resource endowment detrimental to export diversification? UNCTAD Discussion Paper, No. 124 (Geneva: UNCTAD), March. Mayer, J. 2003. Export dynamism and market access. WIDER Discussion paper No. 2003/42. 48
Nordas, H. 1996. South African manufacturing industries - Catching up or falling behind? The Journal of Development Studies, 32, 5: 715-733. Perkins, P. 2003. An analysis of economic infrastructure investment in South Africa. Unpublished Masters thesis, University of the Witwatersrand. Pesaran, M.H., and Smith, R.P. 1995. Estimating long run relationships from dynamic heterogeneous panels. Journal of Econometrics, 68, 79-113. Pesaran, M.H., Shin, Y., and Smith, R. J. 1999. Pooled mean group estimation of dynamic heterogeneous panels. Journal of the America Statistical Association, 94: 621-634. Pissarides, C. 1997. Learning by trading and the returns to human capital in developing countries, The World Bank Economic Review, 11, 1: 17-32. Rangasamy, L. and Harmse, C. 2003. Revisiting the extent of trade liberalisation in the 1990s. South African Journal of Economics, 70, 4: 705-728. Redding, S. and Venables, V. 2004. Economic Geography and International Inequality. Journal of International Economics, 62(1), pp. 53-82 Republic of South Africa. 1996. Growth, Employment and Redistribution: A Macroeconomic Strategy, South African government printers, Pretoria. Riedel, J. 1988. The demand for LDC exports of manufactures: Estimates from Hong Kong. The Economic Journal, 98, March: 138-148. Rodrik, D. 2004. Industrial Policy for the Twenty-First Century, Paper prepared for UNIDO, John F. Kennedy School of Government (September). Sachs, J. and Warner, A. 1997. Natural resource abundance and econmic growth. Center for International Development and Harvard Institute for International Development. Senhadji, A. & Montenegro, C. 1998. Time Series Analysis of Export Demand Equations: A Cross-Country Analysis. IMF Working Paper, WP/98/149. Smal, M.M. 1996. Exchange Rate Adjustments as an Element of a Development Strategy for South Africa. South African Reserve Bank Quarterly Bulletin, 200, 30-39. Smith, R. 2001. Estimation and inferences with non-stationary panel time-series data. Paper prepared for the RC33 conference, Cologne, October 2000. Statistics South Africa. 2003. Final Supply and Use Tables, 2000: An Input-Output Framework (Report no. 04-04-01), Government Printers, Pretoria. Truen, S. 2005. Price complaints ring true. Web posting on Business Day (www.businessday.co.za), 22 April 2005. Tsikata, Y. 1999. Liberalisation and trade performance in South Africa, World Bank Informal Discussion Papers on Aspects of the South African Economy no. 13, The Southern Africa department, The World Bank. UNCTAD. 2002. Trade and Development Report. United Nations Conference on Trade and Development, Geneva.
49
UNIDO. 2004. Industrial Development Report: Industrialisation, Environment, and the Millenium Development Goals in Sub-Saharan Africa. United Nations Industrial Development Organisation, Vienna. Valentine, N. and Krasnik, G. 2000. SADC trade with the rest of the world: Winning export sectors and revealed comparative advantage ratios. South African Journal of Economics, Vol. 68, 2: 266-285. Van Seventer, D. and Gibson, K. 2004. SA’s Absence from Global Trade in Dynamic Products. Trade and Industry Monitor, volume 29, Trade and Industrial Policy Strategies, Johannesburg. Walters, S. and De Beer, B. 1999. An indicator of South Africa’s external competitiveness. South African Reserve Bank Quarterly Bulletin, September 1999, pp. 54-65. Wood, A. and Mayer, J. 2001. Africa’s Export Structure in a Comparative Perspective. Cambridge Journal of Economics, Vol. 5, No. 3: 369-394. Wood, A. and Owens, T. 1995. Export-oriented industrialisation through primary processing? World Development, Vol. 25, pp. 1453-70. Wood, A. and Berge, K. 1997. Exporting manufactures: Human resources, natural resources, and trade policy. Journal of Development Studies, Vol. 34, pp. 3559. Wood, E. 1995. Essays on trade and manufacturing industry in South Africa, Unpublished PhD, University of Cambridge. WTO, 1998. Trade Policy Review: Republic of South Africa. World Trade Organisation Secretariat, Geneva. WTO, 2002. Trade Policy Review: Republic of South Africa. World Trade Organisation Secretariat, Geneva.
50
Appendix A: The “Resource Group” of countries To construct the Resource Group, countries with populations over one million and consistently recorded export data were ranked according to average share of primary products and resource based manufactures in total exports between 1988 and 1990. 10% of the sample on either side of South Africa’s mean value were then selected as the the comparator group. 53 The process yielded the list of countries in Table A. It is worth mentioning that these countries were not the most resource intensive exporters in 1988-1990 (by this measure). Twenty-six countries, mainly from Africa, had ratios higher than that of Malawi, the most resource intensive country in the selected group. Table A: Similar export structures, 1988-1990
Malawi Senegal Paraguay Chile Algeria Tanzania Kenya Benin Madagascar Australia Ethiopia New Zealand South Africa
(RB+PP)/ TX 0.95 0.94 0.94 0.94 0.91 0.91 0.89 0.88 0.88 0.86 0.84 0.83 0.83
Peru Guatemala Nigeria Indonesia Argentina Colombia Costa Rica Zimbabwe El Salvador Uruguay Venezuela Morocco Norway
(RB+PP)/ TX 0.81 0.79 0.78 0.76 0.72 0.71 0.71 0.71 0.68 0.65 0.61 0.61 0.58
Source: as before Note: ‘PP’ = primary products; ‘RB’ = resource based manufactures; TX = total exports. All data represent 1988-1990 averages.
Appendix B: Industrial descriptions and data sources
sectors,
variable
Table B1: Industrial classification 1 2 3 4 5
53
Industry Food [301-304] Beverages [305] Tobacco [306] Textiles [311-312] Wearing apparel [313-315]
15 16 17 18 19
Industry Plastic products [338] Glass and glass products [341] Non-metallic minerals [342] Basic iron and steel [351] Basic non-ferrous metals [352]
Alternative approaches have been followed in the international literature. UNIDO (2004) classify economies as resource rich if natural resource based exports as a percentage of GDP exceeds 20%. Some further modifications are made to ensure the inclusion of mineral-rich economies into this sample. Wood and Berge (1997), Mayer, (1997) and Wood & Mayer (1998) all use land area (km2) per capita and classify economies as resource rich if their ratio is above the sample median.
51
6 Leather and leather products [316] 7 Footwear [317] 8 Wood and wood products [321-322] 9 Paper and paper products [323] Printing, publishing and recorded media [32410 326] 11 12 13 14
Coke and refined petroleum products [331-333] Basic chemicals [334] Other chemicals and man-made fibers [335-336] Rubber products [337]
Metal products excluding machinery 20 [353-355] 21 Machinery and equipment [356-359] Electrical machinery and apparatus [36122 366] Television, radio and communication 23 equipment [371-373] Professional and scientific equipment 24 [374-376] Motor vehicles, parts and accessories 25 [381-383] 26 Other transport equipment [384-387] 27 Furniture [391] 28 Other manufacturing [392-393]
Table B2: Variables descriptions and sources Variable output export px mpenet imports ppidom
Description Source output, R mill, 1995 prices QUANTECH exports, R mill, 1995 prices QUANTECH export price, 1995=100 QUANTECH import penetration, 1995 prices QUANTECH imports, R mill, 1995 prices QUANTECH PPI of output for South African consumption, 2000=1 SSA Weighted output price deflator (1995=100) of developed economies (US, UK, Spain, UNIDO pfor Netherlands, Japan, Canada). Uses total weights in Walter & De Beer (1999). INSTAT 2001 Weighted real output index (1995 = 100) of developed economies (US, UK, Spain, UNIDO qfor Netherlands, Japan, Canada). Uses total weights in Walter & De Beer (1999). INSTAT 2001 SSA [P0142.1: ppifuel PPI: Petrol, 93 octane - Witwatersrand (Unit: Index : 2000=100) Table 16] neer Nominal effective exchange rate. Calculated using Walter and De Beer (1999) weights IMF IFS PPI: Consumption in SA (Unit: Index : 2000=100), includes domestic produced and SSA [P0142.1: ppicons imported goods Table 10]. QUANTECH ulc wage nominal/lprod data ulcr ulc/ppidom skill_unsk Highly skilled/(skilled+semi+unsk) QUANTECH K/L Machinery & equipment capital stock R mill in 1995 prices per worker QUANTECH LNoutputHP Hodrick Prescott filter of log(output), lambda = 7 dLNqHP LNoutputHP-log(output) userate R/US$ exchange rate IMF IFS uspm US import price index, constructed from SITC classified data BLS WITS, Government tar_s Scheduled rates including surcharges Gazettes erp_s Effective rates of protection, calculated using tar_s and 2000 SU table aeb_s anti-export bias (aeb = (1+erp)/(1-xtax)), incl surcharges xtax_s export tax (tax on intermediates/va at Pw), incl. surcharges tel Telephone mainlines (Unit: per 1,000 people) WDI rcapq Rail carrying capacity/manufacturing GDP Perkins (2003) ifcsq Infrastructure Capital stock /manufacturing GDP (R mill in 1995 prices) Perkins (2003) road Paved roads (km) Perkins (2003) elect Electricity generated (gigawatts) Perkins (2003) electk Electricity, water and gas fixed capital stock (R million, 1995 prices) QUANTECH SSA [P0142.1: relpfuel PPI: Petrol, 93 octane - Witwatersrand (Unit: Index : 2000=100)/domestic PPI Table 16] relppfor pfor*neer/ppidom relppusm uspm*neer/ppidom
52
Africa Region Working Paper Series Series # ARWPS 1
ARWPS 2
ARWPS 3
ARWPS 4
ARWPS 5
ARWPS 6
ARWPS 7
Title
Date
Progress in Public Expenditure Management in Africa: Evidence from World Bank Surveys Toward Inclusive and Sustainable Development in the Democratic Republic of the Congo Business Taxation in a LowRevenue Economy: A Study on Uganda in Comparison with Neighboring Countries
C. Kostopoulos January 1999 Markus Kostner March 1999 Ritva Reinikka June 1999
Pensions and Social Security in Sub-Saharan Africa: Issues and Options October 1999 Forest Taxes, Government Revenues and the Sustainable Exploitation of Tropical Forests The Cost of Doing Business: Firms’ Experience with Corruption in Uganda On the Recent Trade Performance of Sub-Saharan African Countries: Cause for Hope or More of the Same
January 2000
Luca Barbone Luis-A. Sanchez B. Juan Zalduendo Jacob Svensson
June 2000
August 2000
Foreign Direct Investment in Africa: Old Tales and New Evidence November 2000
ARWPS 9
The Macro Implications HIV/AIDS in South Africa: Preliminary Assessment
of A November 2000
ARWPS
Revisiting Growth and Convergence: Is Africa Catching Up? December 2000
ARWPS
Spending on Safety Nets for the Poor: How Much, for
11
Duanjie Chen
Luca Barbone
ARWPS 8
10
Author
Francis Ng and Alexander J. Yeats
Miria Pigato
Channing Arndt Jeffrey D. Lewis C. G. Tsangarides William J. Smith
January 2001
How Many? The Case of Malawi ARWPS
Tourism in Africa
Iain T. Christie
12
February 2001 ARWPS
Conflict Diamonds
D. E. Crompton Louis Goreux
13 53
February 2001 ARWPS 14
Reform and Opportunity: The Changing Role and Patterns of Trade in South Africa and SADC
ARWPS
The Foreign Direct Investment Environment in Africa
ARWPS
Choice of Exchange Rate Regimes for Developing Countries
15 16 ARWPS 18
Rural Infrastructure in Africa: Policy Directions
ARWPS 19
Changes in Madagascar: 1993-1999
ARWPS 20
ARWPS 21
ARWPS 22
Poverty
Jeffrey D. Lewis March 2001 March 2001
Fahrettin Yagci April 2001 Robert Fishbein June 2001
in
S. Paternostro July 2001
Information and Communication Technology, Poverty, and Development in sub-Saharan Africa and South Asia
Miria Pigato
J. Razafindravonona David Stifel
Miria Pigato August 2001
Handling Hierarchy in Decentralized Settings: Governance September 2001 Underpinnings of School Performance in Tikur Inchini, West Shewa Zone, Oromia Region Child Malnutrition in Ethiopia: Can Maternal Knowledge Augment The Role October 2001 of Income?
Navin Girishankar A. Alemayehu Yusuf Ahmad
Luc Christiaensen Harold Alderman
ARWPS 23
Child Soldiers: Preventing, Demobilizing and Reintegrating November 2001
Beth Verhey
ARWPS 24
The Budget and Medium-Term Expenditure Framework in Uganda December 2001
David L. Bevan
ARWPS 25
Design and Implementation of Financial Management Systems: An African Perspective
ARWPS 26 ARWPS 27
Guenter Heidenhof H. January 2002 Grandvoinnet Daryoush Kianpour B. Rezaian
What Can Africa Expect From Its Traditional Exports? February 2002 Free Trade Agreements and the SADC Economies February 2002
54
Francis Ng Alexander Yeats Jeffrey D. Lewis Sherman Robinson
Karen Thierfelder ARWPS
Medium Term Expenditure Frameworks: From Concept to Practice. February 2002 Preliminary Lessons from Africa
ARWPS
The Changing Distribution of Public Education Expenditure in Malawi
28
29
February 2002
P. Le Houerou Robert Taliercio Samer AlSamarrai Hassan Zaman
ARWPS 30
Post-Conflict Recovery in Africa: An Agenda for the Africa Region
Serge Michailof April 2002
Markus Kostner Xavier Devictor
ARWPS 31
ARWPS 33
Efficiency of Public Expenditure Distribution and Beyond: A report on Ghana’s 2000 Public Expenditure Tracking Survey in the Sectors of Primary Health and Education Addressing Gender Demobilization and Programs on
Issues in Reintegration
ARWPS 34
Putting Welfare Madagascar
the
Map
ARWPS 35
A Review of the Rural Firewood Market Strategy in West Africa
May 2002
XiaoYe S. Canagaraja
N. de Watteville August 2002
in August 2002
JohanA.Mistiaen BerkSoler T.Razafimanantena J. Razafindravonona Gerald Foley
August 2002
Paul Kerkhof Djibrilla Madougou
Patterns of Governance in Africa
Brian D. Levy
ARWPS 36
September 2002 Stephen Golub
ARWPS 37
Obstacles and Opportunities for Senegal’sInternational Competitiveness: September 2002 Case Studies of the Peanut Oil, Fishing and Textile Industries
ARWPS 38
A Macroeconomic Framework for October 2002 Poverty Reduction Strategy Papers : With an Application to Zambia
S.Devarajan Delfin S. Go
ARWPS 39
The Impact of Cash Budgets on Poverty Reduction in Zambia: A Case November 2002 Study of the Conflict between Well Intentioned Macroeconomic Policy and Service Delivery to the Poor Decentralization
in
Africa:
55
A
Ahmadou Aly Mbaye
Hinh T. Dinh Abebe Adugna Bernard Myers Stephen N. Ndegwa
ARWPS 40 ARWPS 41 ARWPS 42
ARWPS 43
ARWPS 44
Stocktaking Survey
November 2002
An Industry Level Analysis of Manufacturing Productivity in Senegal
Professor A. Mbaye December 2002
Tanzania’s Cotton Sector: Constraints and Challenges in a Global December 2002 Environment Analyzing Financial and Private Sector Linkages in Africa
John Baffes
Abayomi Alawode January 2003
Modernizing Africa’s Agro-Food System: Analytical Framework and February 2003 Implications for Operations
Steven Jaffee Ron Kopicki Patrick Labaste Iain Christie
ARWPS 45
Public Expenditure Rwanda
Performance
in
Hippolyte Fofack March 2003
C. Obidegwu Robert Ngong Elizabeth Crompton
ARWPS 46 ARWPS 47
Senegal Tourism Sector Study
March 2003
Reforming the Cotton Sector in SSA
Louis Goreux March 2003
ARWPS 48
ARWPS 49
HIV/AIDS, Human Capital, Economic Growth Prospects Mozambique
Iain T. Christie
and for
April 2003
John Macrae
Channing Arndt
Rural and Micro Finance Regulation in Ghana: Implications for Development and Performance of the Industry
June 2003
David O. Andah
ARWPS 50
Microfinance Regulation in Benin: Implications of the PARMEC LAW for Development and Performance of the Industry
June 2003
K. Ouattara
ARWPS 51
Microfinance Regulation in Tanzania: Implications for Development and Performance of the Industry Regional Integration in Central Africa:
56
William F. Steel
June 2003
BikkiRandhawa Joselito Gallardo Ali Zafar
ARWPS 52
Key Issues
ARWPS 53
Evaluating Africa
ARWPS 54
AWPS 55
AWPS 56
AWPS 57
AWPS 58a
AWPS 58b
June 2003 Banking
Supervision
in
Keiko Kubota Abayomi Alawode
June 2003
Microfinance Institutions’ Response in Conflict Environments: Eritrea- Savings and Micro Credit Program; West Bank and Gaza – Palestine for Credit and Development; Haiti – Micro Credit National, S.A. Malawi’s Tobacco Sector: Standing on One Strong leg is Better than on None Tanzania’s Coffee Sector: Constraints and Challenges in a Global Environment The New Southern AfricanCustoms Union Agreement How Far Did Africa’s First Generation Trade Reforms Go? An Intermediate Methodology for Comparative Analysis of Trade Policies How Far Did Africa’s First Generation Trade Reforms Go? An Intermediate Methodology for Comparative Analysis of Trade Policies
Marilyn S. Manalo June 2003
Steven Jaffee June 2003 John Baffes June 2003
June 2003
Robert Kirk Matthew Stern
June 2003
Lawrence Hinkle A. Herrou-Aragon Keiko Kubota
June 2003
Lawrence Hinkle A. Herrou-Aragon Keiko Kubota
AWPS 59
Rwanda: The Search for Post-Conflict Socio-Economic Change, 1995-2001 October 2003
AWPS 60
Linking Farmers to Markets: Exporting October 2003 Malian Mangoes to Europe
Morgane Patrick J-M. Voisard
AWPS 61
Evolution of Poverty and Welfare in Ghana in the 1990s: Achievements and October 2003 Challenges
S. Canagarajah Claus C. Pörtner
AWPS 62
Reforming The Cotton Sector in SubSaharan Africa: SECOND EDITION November 2003
AWPS 63 (E) AWPS 63 (F)
Republic of Sector Study
Madagascar:
C. Obidegwu
Danielou Labaste
Louis Goreux
November 2003
Iain T. Christie D. E. Crompton
République de Madagascar: Etude du November 2003 Secteur Tourisme
Iain T. Christie D. E. Crompton
Tourism
57
AWPS 64
AWPS 65
AWPS 66
Migrant Labor Remittances in Africa: Reducing Obstacles to Development Novembre 2003 Contributions Government Revenues and Expenditures in Guinea-Bissau: Casualty and Cointegration How will we know Development Results when we see them? Building a Results-Based Monitoring and Evaluation System to Give us the Answer
AWPS 67
An Analysis of the Trade Regime in Senegal (2001) and UEMOA’s Common External Trade Policies
AWPS 68
Bottom-Up Administrative Reform: Designing Indicators for a Local Governance Scorecard in Nigeria
AWPS 69
Tanzania’s Tea Sector: Constraints and Challenges
AWPS 70
Tanzania’s Cashew Sector: Constraints and Challenges in a Global Environment
AWPS 71
An Analysis of Chile’s Trade Regime in 1998 and 2001: A Good Practice Trade Policy Benchmark
AWPS 72
AWPS 73
AWPS 74
Regional Trade Integration inEast Africa: Trade and Revenue Impacts of the Planned East African Community Customs Union Post-Conflict Peace Building in Africa: The Challenges of Socio-Economic Recovery and Development An Analysis of the Trade Regime in Bolivia in2001: A Trade Policy Benchmark for low Income Countries
Cerstin Sander Samuel Maimbo Francisco G. Carneiro
January 2004 Joao R. Faria Boubacar S. Barry Jody Zall Kusek June 2004
Ray C. Rist Elizabeth M. White
Alberto Herrou-Arago June 2004
Keiko Kubota Talib Esmail
June 2004
Nick Manning JanaOracGalia Schechter John Baffes
June 2004 Donald Mitchell June 2004 Francesca Castellani July 2004
A. Herrou-Arago Lawrence E. Hinkle Lucio Castro
August 2004
Christiane Kraus Manuel de la Rocha Chukwuma Obidegwu
August 2004 Francesca Castellani August 2004
Alberto Herrou-Aragon Lawrence E. Hinkle Vincent da Cruz
AWPS 75
Remittances to Comoros- Volumes, October 2004 Trends, Impact and Implications 58
M.
Wolfgang Fendler
Adam Schwartzman AWPS 76
Salient Features of Trade Performance in Eastern and Southern Africa
Fahrettin Yagci October 2004
AWPS 77
Alan Gelb Implementing Based Aid in Africa
AWPS 78
AWPS 79
AWPS 80 AWPS 81
AWPS 82
Enrique Aldaz-Carroll
Performance- November 2004
Xiao Ye
Poverty Reduction Strategy Papers: Do they matter for children and Young December 2004 people made vulnerable by HIV/AIDS? Experience in Scaling up Support to Local Response in Multi-Country Aids December 2004 Programs (map) in Africa What makes FDI work? A Panel Analysis of the Growth Effect of FDI in February 2005 Africa Earnings Differences between Men and Women in Rwanda February 2005 The Medium-Term Framework
Brian Ngo Rene Bonnel Miriam Temin Faith Tempest Jean Delion Pia Peeters Ann Klofkorn Bloome Kevin N. Lumbila Kene Ezemenari Rui Wu
Expenditure April 2005
Chukwuma Obidegwu
The Challenge of Budget Integration in SSA countries AWPS 83
AWPS 84
Rules of Origin and SADC: The Case for change in the Mid Term Review of the Trade Protocol Sexual Minorities, Violence and AIDS in Africa
Paul Brenton June 2005
Frank Flatters Paul Kalenga
July 2005
Chukwuemeka Anyamele Ronald Lwabaayi Tuu-Van Nguyen, and Hans Binswanger
AWPS 85
Poverty Reducing Potential Smallholder Agriculture in Zambia:
of July 2005
Opportunities and Constraints AWPS 86
Infrastructure, Productivity and Urban Dynamics in Côte d’Ivoire An empirical analysis and policy implications
59
Paul B. Siegel Jeffrey Alwang
July 2005
Zeljko Bogetic Issa Sanogo
AWPS 87
Poverty in Mozambique: Unraveling Changes and Determinants
Louise Fox August 2005
Elena Bardasi, Katleen Van den Broeck
AWPS 88
AWPS 89
AWPS 90
Operational Challenges: Community Home Based Care (CHBC) forPLWHA in Multi-Country HIV/AIDS Programs (MAP) forSub-Saharan Africa
Framework for Forest Resource Management in Sub-Saharan Africa
August 2005
Nadeem Mohammad Juliet Gikonyo
August 2005
Giuseppe Topa
Kenya: Exports Prospects and Problems
September 2005
Uganda: How Good a Trade Policy Benchmark for Sub-SaharanAfrica
September 2005
AWPS 92
Community Driven Development in South Africa, 19902004
October 2005
David Everatt Gwagwa
AWPS 93
The Rise of Ghana’’s Pineapple Industry from Successful take off to Sustainable Expansion
November 2005
Morgane Danielou
South Africa: Sources and Constraints of Long-Term Growth, 1970-2000
December 2005
AWPS 91
AWPS 94
AWPS 95
South Africa: Determinants of Export Supply
60
Francis Ng Alexander Yeats Lawrence E. Hinkle Albero Herrou Aragon RangaRajanKrishnamani Elke Kreuzwieser Lulu
Christophe Ravry
December 2005
Johannes Fedderke
Lawrence Edwards Phil Alves
