Edition 9 January 2017 South Africa Economic Update Private Investment World Bank Group for Jobs South Africa Economic Update Edition 9 | Focus on Private Investment for Job Creation January 2017 | World Bank Group © 2017 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street NW Washington, DC 20433 USA All rights reserved All references to dollars are to U.S. dollars unless otherwise noted. Cover photos: Shutterstock & Sunday haircuts in Soweto – Craiglevebvre@Flickr. The report was designed and typeset by Shereno Printers, Gauteng, South Africa. Contents Foreword v List of Abbreviations vi Executive Summary 1 Chapter 1 – Recent Economic Developments 4 Global Economic Developments 4 Developments in South Africa’s Real Sector 7 Labor Market Developments 13 Inflation and Monetary Policy 16 Fiscal Developments 19 External Sector 23 Outlook 25 Chapter 2 – Private Investment for Job Creation 29 Introduction 29 Job Creation and Investment since 1994 32 Allocation of Private Capital across Sectors 35 Costs and Potential Benefits of Investment Tax Incentives 40 Conclusion 53 Endnotes 56 References 60 Tables Table 1.1: GDP growth (supply side) 10 Table 1.2: Expenditure projections and deviations from budget (2016/17 to 2019/20) 20 Table 1.3: Baseline annual growth forecast 26 Table 2.1: Long-term capital demand response to a change in the relative cost of capital, 2006–12 39 Table 2.2: Tax parameters used in marginal effective tax rate computations 43 Table 2.3: Marginal effective tax rate, by sector and class of investment assets 44 Table 2.4: Changes in the user cost of capital as a result of investment tax incentives, by sector 45 Table 2.5: Additional capital and labor demand due to investment tax incentives, by sector 46 Table 2.6: Impact and cost of investment tax incentives on the creation of jobs, 2012 48 iii SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Figures Figure 1.1: Global activity indicators 4 Figure 1.2: Real GDP growth in commodity exporters and importers 5 Figure 1.3: Global financial flows and commodity prices 6 Figure 1.4: Regional GDP growth in South Africa 8 Figure 1.5: GDP growth (demand side) 9 Figure 1.6: Finance, insurance, and real estate have been driving growth 11 Figure 1.7: Exchange rate volatility in 2016 12 Figure 1.8: Unemployment and economic inactivity 13 Figure 1.9: Labor productivity and unit labor costs 16 iv Figure 1.10: Labor productivity growth 17 Figure 1.11: Consumer price index inflation and contributions 19 Figure 1.12: Trajectory of net public debt across budgets and MTBPSs 21 Figure 1.13: Main components of the current account and effective exchange rate 24 Figure 1.14: Net foreign direct investment to selected countries 25 Figure 1.15: South African Reserve Bank business cycle indicators 26 Figure 2.1: Cumulative private sector job creation, 1994–2015 32 Figure 2.2: Job creation by sector, 1994–2015 33 Figure 2.3: Evolution of the capital-labor ratio within and between sectors, 1994–2015 34 Figure 2.4: Contribution of expansionary and technology investment to job creation, 1994–2015 35 Figure 2.5: Capital stock growth and reallocation across sectors, 1994–2015 36 Figure 2.6: Sector shares in total capital stock, 1994–2015 37 Figure 2.7: Economic gains of capital reallocation across sectors, 1994–2015 38 Figure 2.8: Impact of investment tax incentives on investment and job creation, 2012 47 Figure 2.9: Employment multipliers, 2012 48 Figure 2.10: Impact of employment tax incentives on job growth, 2014/15 49 Boxes Box 1.1: Education and labor market developments in South Africa 14 Box 1.2: Macroeconomic and poverty impacts of El Niño on Southern African Development Community countries 17 Box 1.3: The threat of a rating downgrade to “junk” 22 Box 1.4: Ideas in poverty reduction: Farming from rooftops in Johannesburg’s central business district 27 Box 2.1: Job creation and poverty reduction 31 Box 2.2: Promises and challenges of spatialized industrial policies 40 Box 2.3: Terms of trade, exchange rate movements, and new comparative advantages 51 Acknowledgments This Economic Update was produced by a team led by Sebastien Dessus and comprising Marek Hanusch, Yashvir Algu, Gerard Kambou, Charl Jooste, Mokgabo Molibeli, Sandra Gain, Victor Sulla, Zandile Ratshitanga, Olivia d’Aoust, Precious Zikhali, John Gabriel Goddard, Wayde Flowerday, Kobina Egyir Daniel, Ayanda Mavundla, Catiana Garcia-Kilroy, Indhu Raghavan, Aroop Chatterjee, Toby Linden, Paolo Belli, Thomas Farole, and Sebastian James. It was peer reviewed by Volker Treichel (Lead Economist, World Bank), Antonio Nucifora (Lead Economist, World Bank), and Jorge Maia (Head of the Strategic Research and Information Department, Industrial Development Corporation), and benefited from comments from Catherine MacLeod (Chief Director, National Treasury), Duncan Pieterse (Director, National Treasury), Konstantin Makrelov (Chief Director, National Treasury), Nigel Gwynne-Evans (Chief Director, Department of Trade and Industry), and Montfort Mlachila (Resident Representative, International Monetary Fund) and guidance from Guang Zhe Chen (Country Director for South Africa, World Bank), Catherine Tovey (acting Country Director for South Africa, World Bank), and Mark Roland Thomas (Practice Manager, Macroeconomics and Fiscal Management, World Bank). The team is also thankful to Wandile Zwane, Donny Phakwago, and Simon Motsusi, from the Municipality of Johannesburg’s Department of Social Development, for their introduction to the Metro’s Food Security and Agriculture Programme. Foreword The South Africa Economic Update is a in a context of shrinking fiscal space, where World Bank biannual publication that offers South Africa has had three consecutive years the country’s economic outlook against the of negative per capita economic growth. The global economic prospects. The Update also report argues that investment tax incentives provides evidence-based analysis on an aspect have been shown to limit job destruction in of the South African economy that is intended the industrial sector, despite the industrial to enhance ongoing policy debates to foster contraction that the country has experienced the country’s goals to increase growth, reduce since the establishment of democracy in 1994. poverty, lessen inequality, and make a dent As with previous editions of the South in the stubbornly high unemployment rate. Africa Economic Update, it is our hope that this These goals are in line with the World Bank’s report does not prescribe a way forward, but twin goals to help end poverty by 2030 and rather adds to the body of ongoing work that promote shared prosperity. will help policy makers and other stakeholders In this ninth edition, the focus of the in their efforts to find sustainable solutions to South Africa Economic Update is on private the highly complex challenges in improving investment for job creation. The pace of job the economy and bettering the lives of South creation has been too slow in the past decade Africans. Through its various financial and to meet South Africa’s national development advisory instruments, the World Bank Group goal to create some 600,000 new jobs every stands ready to contribute to these endeavors. year, to bring down the unemployment rate to The next edition of the South Africa Economic 6 percent by 2030, from 33.8 percent in 2015. Update will focus on innovation. This Update analyzes the central role that private investment plays in the government’s effort to accelerate the promotion of industrial development given the shortcomings of South Africa’s commodity-driven growth model. The report examines whether this emphasis has Paul Noumba Um yielded the desired outcomes, ultimately to World Bank Country Director drive labor demand and consequently increase for South Africa job creation. The report assesses the effectiveness, cost, and impact of investment tax incentives granted to the various economic sectors on additional investment, and on job creation. The report suggests that reorienting incentives toward the industrial sector would create additional jobs at no additional fiscal cost, an important point v SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS List of Abbreviations ABSA Barclays Africa Group Limited CGE Computable General Equilibrium CIT Corporate Income Tax DTI Department of Trade and Industry EBIT Earnings Before Interest and Taxes EM Emerging Market EMDES Emerging Markets and Developing Economies ETI Employment Tax Incentive FDI Foreign Direct Investment FEZ Food Empowerment Zone GDP Gross Domestic Product IDC Industrial Development Corporation IDZ Industrial Development Zone IT Information Technology ITI Investment Tax Incentive MPC Monetary Policy Committee MTBPS Medium-Term Budget Policy Statement METR Marginal Effective Tax Rate NAFTA North American Free Trade Area NDP National Development Plan NEER Nominal Effective Exchange Rate NIPF National Industrial Policy Framework OECD Organisation for Economic Co-Operation and Development OPEC Organization of the Petroleum Exporting Countries PBO Parliamentary Budget Office PIT Personal Income Tax PPP Purchasing Power Parity PMI Purchasing Managers Index Q/Q Quarter-on-Quarter Q1 First Quarter QLFS Quarterly Labor Force Survey REER Real Effective Exchange Rate S&P Standard and Poor’s SAAR Seasonally Adjusted Annualized Rate SADC Southern African Development Community SARB South African Reserve Bank SARS South African Revenue Service SBC Small Business Corporation SEZ Special Economic Zone StatsSA Statistics South Africa SUB-IG Sub-Investment Grade UCC User Cost of Capital Y/Y Year-on-Year ZAR South African Rand vi Executive Summary Global economic growth remained 1.3 percent in 2015. This slowdown was felt moderate at 2.3 percent in 2016. High-income across all regions in South Africa, Gauteng economies experienced an underlying weak included, and marks the third consecutive growth momentum (1.6 percent) and low year of negative per capita growth and inflation, although the U.S. Federal Reserve’s stagnating poverty. In retrospect, the decline decision to raise interest rates in December in commodity prices since 2012 may have 2016 signaled that the U.S. economy is picking cost at least 4 percentage points of GDP. In up steam. Among emerging markets and 2016, GDP growth was modestly driven by the developing economies (EMDEs), commodity financial, business, and real estate sectors, exporters were still grappling with low although this major engine of South African commodity prices, while gross domestic growth has also been slowing. Notably, South product (GDP) growth was broadly stable Africa has avoided a much-feared downgrade among importers. Although initially supported to sub-investment grade by an international in the first part of 2016 by a significant credit rating agency in 2016—which could acceleration in foreign capital inflows, GDP have reduced GDP per South African by about growth (at 3.4 percent in 2016) in EMDEs ZAR 1,000 by end-2017—by staying the fiscal suffered in the second part of the year from a consolidation course. Yet South Africa is not reversal of such capital inflows in anticipation out of the woods yet for 2017, with two agencies of tighter U.S. monetary policy. Meanwhile, now rating South Africa one notch above sub- commodity prices recovered slightly from investment grade with negative outlook. their low levels of January 2016. However, 2016 may mark the trough Overall, a modest global economic recovery of South Africa’s business cycle. A modest is expected in 2017 and 2018–19, with growth recovery is now foreseen for 2017 and 2018, at 2.7 and 2.9 percent, respectively. Pulled by driven by (modestly) rising commodity the U.S. economy, high-income economies’ prices, easing inflationary pressures (as the growth is foreseen at 1.8 percent in 2017–19, rand stabilizes and the effects of drought on and that of EMDEs at 4.2 percent in 2017, food prices dissipate) and a pickup in credit 4.6 percent in 2018, and 4.7 percent in 2019. stimulating household consumption demand. Risks nonetheless remain on the downside, By contrast, the continuation of the needed and emerging markets with preexisting fiscal consolidation efforts (mainly through vulnerabilities as a result of external imbalances, additional tax revenue) should not offer any large financing needs, and unsustainable debt significant stimulus to GDP growth. dynamics would likely be the most affected by As in 2016, private investment will be the financial market disruptions. main variable influencing GDP growth. On In this depressed environment, South the one hand, should investment remain Africa’s GDP growth is estimated to have weak, this would further undermine growth decelerated to 0.4 percent in 2016, down from prospects, raise again the likelihood of a 1 SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS rating downgrade, and perpetuate a vicious was not accompanied by a sufficient expansion circle of low growth–low investment. On the of productive capacity. Only the service sector other hand, accelerated investment could combined capital deepening with job creation. benefit from a still weak and more stable But the jobs that were created in services were rand, improving electricity capacity, and less among those offering the lowest wages. fractious labor relations, to boost exports The trends in capital allocation across and growth and stabilize the capital account. sectors were not only bad for job creation. Accelerating investment will require providing They were equally bad for GDP growth, a predictable business environment, not least generating significant losses in aggregate 2 through greater policy certainty. capital productivity. Indeed, since 2008, there Private investment not only matters for has been a significant deterioration in the growth, but also for job creation. In 2016, the South African economy’s capacity to direct unemployment rate reached a 13-year high, private investment toward sectors with growing especially for youth and the unskilled. Such economic potential, manufacturing sectors in stubbornly high unemployment constitutes particular. Although the delayed reactions the major hurdle for South Africa to overcome, of concentrated industries to changing to meet its national development goals of opportunities, and the long time needed by eliminating poverty and reducing inequality: large infrastructure projects to start generating each job created in South Africa lifts about returns may explain this negative trend, the one person out of poverty. analysis of the current investment tax incentive In retrospect, the growth model pursued framework suggests that the trend may have since democracy has not generated enough also strongly contributed to the misallocation jobs, even during the peak of the commodity of capital. Compared with industrial sectors, cycle. Cognizant of the shortcomings of South lower marginal tax rates for the mining and Africa’s commodity-driven growth model, construction sectors make private investment the authorities have accelerated efforts to in these sectors equally or more remunerative promote industrial development in the past despite much lower growth and job creation decade, through a suite of industrial policy returns for the economy at large. instruments, including tax incentives, public Nonetheless, by reducing the tax procurement requirements, provision of burden of firms, investment tax incentives infrastructure in support of specific sectors, have encouraged additional investment in lending to and equity participation in agriculture, construction, manufacturing, strategic sectors, and competition and trade trade, and other services. Overall, the policies. Such a policy orientation builds on additional investment generated by tax the expectation that it could generate several incentives exceeds the government’s positive outcomes, including higher wages foregone revenue from distributing the tax for workers, larger employment and growth incentives. Furthermore, the existence of large multipliers, increased domestic competition employment multipliers brings the fiscal cost and access to the larger world demand, and of job creation to a fraction of total labor costs, more stable growth underpinned by a less especially in the manufacturing sector. volatile capital account. Through appropriate Investment tax incentives have thus investments, industrial development can contained job destruction in industrial sectors, seek to expand production in sectors where and explanations for industrial contraction comparative advantage already exists, or since democracy must be found elsewhere, alternatively develop new comparative possibly among insufficient skills and advantages through technological upgrading. infrastructure, policy uncertainty, a volatile However, in recent years, the authorities’ rand, and complicated labor relations. efforts to promote industrial development were Moving forward, reorienting incentives not matched with a significant reallocation from mining toward industrial sectors would of private capital toward industrial sectors, create additional jobs at no additional fiscal or with higher industrial employment. In cost, the more so as industrial sectors would the agriculture, mining, and manufacturing benefit in the medium term from the new sectors, investments resulted in replacing jobs business opportunities from the depreciation with machines, as the technological upgrading of the rand, declining commodity prices, and the coming online of large additional power generation capacity. Over the longer term, efforts to raise workers’ skills and professional experience, and to foster spatial economic integration will be crucial to maintain this momentum. These efforts will include continued evaluation of education programs and youth employment initiatives; support to pilot initiatives in the domains of professional training and job postings at the local level; 3 and improved urban and regional planning, including special economic zones. SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS CHAPTER 1 Recent Economic Developments Global Economic Developments manufacturing, as industrial production Weaker trade and and trade growth returned to trend (Figure Global growth slowed notably in 2016, reflecting 1.1). The global manufacturing Purchasing investment, and weak growth in high-income economies Managers Index (PMI) increased further in Global growth is estimated to have slowed October and November, continuing to improve increasing policy from 2.7 percent in 2015 to 2.3 percent in in Q4. After contracting in Q2, global goods uncertainty kept 2016—its weakest performance since the trade recovered in Q3, helped by stronger global financial crisis. Stagnant global trade, import demand from high-income economies global growth low in subdued investment, and heightened policy and emerging market and developing 2016 uncertainty depressed global economic activity. economies (EMDEs). However, the number of After a weak first half, global growth gathered protectionist measures implemented by G20 economies continued to rise, and trade policy momentum in the third quarter (Q3) of 2016, uncertainty increased. supported by improving conditions in global Figure Global activity indicators 1.1 a. Global GDP, industrial production, b. Composite PMI of major commodity and trade exporters (percent, annualized q/q) (index) 55 3 2012Q1-2016Q3 average 2 50 1 0 45 Q3 average -1 Q4 average -2 40 Federation Russian -3 Nigeria Brazil Africa South Q1 Q2 Q3 Q1 Q2 Q3 Q1 Q2 Q3 GDP Industrial Trade production Source: World Bank. Note: PMI = Purchasing Managers Index; Q = quarter. PMI >50 is expansion. 4 High-income economies continued to Developments were mixed in other high- struggle with subdued growth and low income economies. Growth in Japan remained inflation in a context of increased uncertainty subdued in 2016, estimated at 1.0 percent, about policy direction. Growth in advanced compared with 1.2 percent in 2015. Investment economies slowed from 2.1 percent in 2015 to and exports were generally weak, while an estimated 1.6 percent in 2016. Growth in private consumption showed some signs of the United States weakened notably, from 2.6 improvement. Growth in the United Kingdom percent in 2015 to an estimated 1.6 percent was more resilient than expected. Inflation in in 2016. Activity in the United States picked November rose to its highest level since April up in Q3, supported by a rebound in exports, 2014, reflecting the substantial depreciation 5 a positive contribution from inventories, and of the pound sterling since June. The diminished drag from private investment. government announced additional spending The labor market strengthened, with the for infrastructure in November’s Autumn unemployment rate dropping to 4.6 percent Statement. The U.K. High Court’s ruling that in November. Amid tighter labor market the government should seek parliamentary conditions, actual and expected inflation rose. approval before triggering Article 50 to start The U.S. Federal Reserve increased policy Brexit negotiations could potentially delay the Among emerging rates by 25 basis points in December, and process. markets, commodity signaled that further rate hikes were likely in EMDEs grew by 3.4 percent in 2016, 2017. broadly the same pace as in 2015 (3.5 percent). importers Growth in the Euro Area slowed as well, Commodity exporters continued to expand outperformed from 2.0 percent in 2015 to 1.6 percent at a markedly slower pace than commodity in 2016, as domestic demand and exports importers (Figure 1.2), although there was commodity remained weak. However, in November, the notable heterogeneity within each group. exporters composite PMI reached its highest level in Growth in commodity exporters is estimated 2016, and economic sentiment continued to at 0.3 percent in 2016. Gross domestic product improve gradually. Headline inflation ticked (GDP) in Brazil, the largest commodity- up to 0.6 percent (y/y, year-on-year), while core exporting EMDE, contracted 3.3 percent inflation remained at 0.8 percent, and market- quarter-to-quarter (q/q) at a seasonally based, long-term inflation expectations stayed adjusted and annualized rate (saar) in Q3, appreciably below the European Central the seventh consecutive quarter of negative Bank’s target. At its December meeting, the growth, led by a large decline in exports and European Central Bank kept interest rates investment. In October, the central bank cut unchanged. It extended its bond-buying interest rates for the first time in four years. program until the end of 2017, but monthly In the Russian Federation, GDP contracted 0.4 asset purchases will be reduced from 80 billion percent (y/y) in Q3—a modest improvement to 60 billion starting in April 2017. from the contraction of 0.6 percent in Q2. Figure Real GDP Growth in commodity exporters and importers 1.2 (percent) Commodity exporters 10 Commodity importers Commodity importers ex. China 8 6 4 2 0 -2 2011 2012 2013 2014 2015 2016 Source: World Bank. SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Meanwhile, commodity importers are Global financing conditions tightened and estimated to have expanded by 5.6 percent in capital flows to EMDEs slowed following the 2016, reflecting resilient domestic demand, U.S. elections Since reaching historical lows in August, low commodity prices, and accommodative global bond yields have risen, initially driven macroeconomic policies. In India, growth by more supportive economic data and continued to be strong in Q3, at 7.3 percent prospects of global reflation. The U.S. election (y/y), but the unexpected demonetization was followed by a 62 basis-point jump in U.S. initiative will likely restrict consumption and 10-year Treasury yields, on prospects of higher 6 inflation and U.S. interest rate levels over the activity. Momentum in Eastern Europe and medium term. Rising long-term interest rates Central Asia slowed in Q3, with growth rates in the U.S. have put upward pressure on bond falling in Poland and Romania, likely reflecting yields in other high-income economies and a continued slowdown of investment. GDP in EMDEs, while contributing to a significant Turkey contracted 1.8 percent (y/y), the first appreciation of the U.S. dollar. decline since 2009, policy uncertainty and the The U.S. election results were followed Rising U.S. interest by a moderate sell-off in emerging market coup in July weighed on confidence. rates put pressure equities, pairing gains made earlier in the Growth in China slowed slightly to 6.7 year. EMDE currencies dropped collectively on bond yields percent in 2016, as the economy continues by around 2 percent, with markedly more globally to rebalance from industry to services. In pronounced declines in the Turkish lira, November, the official manufacturing PMI Mexican peso, Argentine peso, Malaysian ringgit, and Brazilian real. There were large increased to 51.7—the highest reading outflows from EMDE exchange-traded funds since July 2014—while exports and imports in the immediate aftermath of the U.S. rebounded. Industrial production, retail sales, elections, but they have generally moderated and investment growth stayed broadly stable. since then (Figure 1.3, panel a). Higher bond Credit growth remained robust despite tighter yields, concerns about more restrictive trade regulation, and continues to outpace nominal policies, and higher policy risk premiums have all contributed to shifting sentiment. In GDP growth. Foreign reserves declined for the October, EMDE Eurobond sales reached their fifth straight month in November, showing the highest monthly level since April 2014, but the biggest monthly decline since January. momentum slowed sharply in November. Figure Global financial flows and commodity prices 1.3 a. Net flows into emerging market funds b. Change in Commodity prices (US$, million) (percent) 50 6,000 EM equity funds EM bond funds 30 4,000 2,000 10 0 -10 -2,000 -30 2014-16 -4,000 -50 2017-19 -6,000 -70 -8,000 Dec-15 Feb-16 May-16 Aug-16 Nov-16 Energy Metals Agriculture Sources: JP Morgan; World Bank. Note: EM = emerging market. Foreign direct investment (FDI) flows to to recover marginally to an average pace of 1.8 EMDEs remained subdued throughout 2016, percent in 2017–19, reflecting strengthening with significant differences across commodity of activity in the U.S.. Growth in EMDEs is importers and exporters. Among commodity projected to accelerate to 4.2 percent in 2017, exporters, persistently low commodity prices and to an average of 4.7 percent in 2018–19, have reduced the attractiveness of investment as gradual recovery in commodity prices helps in mining and exploration. FDI growth is now boost activity in commodity-exporting EMDEs, well below long-term averages in commodity- and growth in commodity-importing EMDEs importing and commodity-exporting regions. remains solid. Subdued FDI flows to commodity exporters Downside risks to global growth include 7 add to external financing needs at a time rising policy uncertainty, particularly in the when fiscal and current account positions are U.S. and Europe; financial market disruptions; under pressure. FDI flows to large commodity and growth disappointments in major importers were generally resilient in 2016. economies. In contrast, fiscal stimulus in major economies—particularly the U.S.—represents Commodity prices recovered an important upside risk. In EMDEs, large Commodity prices stabilized during 2016, investment gaps amid limited fiscal resources Commodity prices and are expected to gradually recover in remain important challenges. are expected 2017–19 (Figure 1.3, panel b). Crude oil prices recovered from a low of US$30 per barrel at Developments in South Africa’s Real to recover, the start of 2016, but are still half their pre- Sector 2014 levels. Crude oil prices are projected supporting growth to rise to US$55 per barrel in 2017, from an South African growth continued to slow in 2016 in commodity annual average of US$43 per barrel in 2016. In 2015, the South African economy grew At its November 30 meeting, the Organization by 1.3 percent, less than half what it registered exporters of the Petroleum Exporting Countries after its rebound from the global financial (OPEC) announced a cut of 1.2 million crisis, reaching a high of 3.3 percent in 2011. barrels a day (the first cut in eight years), The growth rate in 2015 was also a mere quarter higher than the 0.5 million to 1 million barrels of the five-year average of 4.7 percent before a day envisaged two months earlier. The cuts the financial crisis (2003–07). By contrast, were reaffirmed at a subsequent meeting while the impacts of the crisis continued to between OPEC members and various non- linger across the world, global growth in 2015 OPEC oil producers on December 9. OPEC’s was back at almost two-thirds of its pre-crisis crude output, which averaged 33.5 million average, led by high-income economies, which barrels a day during 2016 Q3, is expected to were at three-quarters; emerging markets fall to 32.5 million barrels a day, if countries and Sub-Saharan African economies were adhere to the agreed cuts. Meanwhile, prices at over half. This means that South Africa’s of metals and energy commodities are settling economy has been struggling to return to past in line with fundamentals. Metal and mineral performance, finding it particularly difficult prices are projected to rise by 4.1 percent y/y to gather steam in a weak global economy, in 2017, reflecting supply tightness for some compounded by domestic challenges. metals and mine closures. Agriculture prices One of the global impacts that have are projected to remain broadly stable in 2017. affected South Africa particularly hard Supplies for most commodities are adequate. was the end of the commodity super cycle, The likelihood of supply disruptions due to La aggravating headwinds that continued to Niña has diminished. linger from the global financial crisis of 2008. Yet it would be wrong to blame low growth Global growth is projected to recover on commodities alone. Electricity shortages moderately in 2017–19, but risks remain on have also kept growth low during that period, the downside as did disruptions to production from the Global growth is expected to rise to 2.7 South Africa’s fractious labor relations. percent in 2017, and to 2.9 percent in 2018– Domestic politics have increasingly been on 19, mainly reflecting a recovery in EMDEs. investors’ minds, holding back investments Growth in high-income economies is projected that are needed to restructure the economy SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS in response to the commodity downturn (see contributed 0.7 percentage points to national Chapter 2). This was true in 2015 and arguably growth. Although it is still South Africa’s intensified in 2016. South Africa is not alone powerhouse, and much like in most other in this. Some members of the BRICS group— provinces, Gauteng province saw its growth which includes Brazil, Russia, India, China, slowing in 2015, compared with the 2007– and South Africa—share similar experiences: 14 average (Figure 1.4). A similar trend is commodity prices, structural constraints, and seen in South Africa’s eight other provinces, politics have also adversely impacted Brazil with KwaZulu-Natal and Western Cape (the and Russia, arguably worse than South Africa, two fastest growing provinces in 2007–14) 8 both seeing GDP contract in 2015, without experiencing a significant decline in growth in recovering in 2016. 2015. Because of their large share in total GDP Of South Africa’s nine provinces, Gauteng (66.7 percent), 79 percent of growth in 2015 province—accounting for over a third of was accounted for by Gauteng, KwaZulu-Natal, the country’s total gross value added— and Western Cape. Figure Regional GDP growth in South Africa South African growth 1.4 (annual GDP growth, %) slowed due to 3.0 both external and domestic headwinds 2.5 2.0 1.5 1.0 0.5 0.0 South KwaZulu Western Gauteng Northern Limpopo Mpu- Eastern Free North Africa Natal Cape Cape malanga Cape State West -0.5 2007-2014 avg. 2015 -1.0 Sources: StatsSA; World Bank staff calculations. At the national level, growth in South Private consumption has been weak Africa in 2015 was driven largely by rebound in 2016, growing at 0.9 percent y/y in the effects, not least from extensive industrial first three quarters. Household spending action in the first half of 2014, as well as has been constrained by weak labor market frequent periods of load-shedding. The year developments, including rising unemployment ended with the abrupt and opaque dismissal and a significant weakening in real wage of well-respected Minister of Finance Nene, growth from 2015 levels. Private sector credit which shocked markets, sending the rand and extension in 2016 slowed considerably, with investor confidence to new lows. This event its remaining driver being corporate credit; was expected to take a toll on fixed investment, meanwhile, credit to households contracted by which plummeted by -10.0 percent q/q saar an average rate of -3.7 percent y/y in real terms in Q1 2016, the sharpest decline since the between January and October 2016, a further global financial crisis. On a y/y basis, gross deterioration of the -0.7 percent y/y in the same fixed investment contracted from 2015 Q4 period the previous year. Yet, although coming through 2016 Q3, by -3.0 percent on average, down, household indebtedness remained see Figure 1.5. elevated, at 75.1 percent of disposable income in 2016 Q2. Household wealth has also come momentum could not be sustained in Q3, with under pressure as property markets have export growth contracting by -3.9 percent y/y, softened: the ABSA House Price Index fell by largely due to weak performance in exports -2.6 percent in real terms between January and of precious metals and transport equipment. October 2016. Combined, these factors help Overall, exports shaved 0.1 percentage point explain the weak household consumption in off GDP growth in the first three quarters of 2016. That said, the dynamics have improved 2016. over the year, as inflationary pressures Low domestic demand—and a weak have been easing, undermining household rand by historical standards—translated purchasing power increasingly less. On a q/q into a decline in imports, which fell by -3.3 9 saar basis, private consumption accelerated percent y/y in the first three quarters of the from -1.7 percent in 2016 Q1, to 1.4 and 2.6 year, the first three-quarter decline since the percent in Q2 and Q3, respectively. global financial crisis. Overall, absorption Public consumption grew at 2.1 percent (consumption and investment) plus exports in 2016 Q1, but decelerated to 1.5 and 1.1 contributed negatively to GDP growth, shaving percent y/y, respectively, over the following 0.5 percentage point off headline growth two quarters, as the government continued in the first three quarters of 2016, while the Consumption has on its fiscal consolidation path. Overall, final strong contraction of imports added 1.1 been driving growth consumption expenditure grew by an average percentage points in a strict accounting sense. of 1.1 percent y/y in the first three quarters A more meaningful interpretation is that, to in 2016, while of 2016, which, although a low rate, was the some extent, growth in 2016 was driven by investment fell fastest growing expenditure category in the a rebalancing from imports to consuming year, adding 0.8 percentage points to overall domestically produced products, which GDP growth. chimes with the change of relative foreign Exports performed well in the first half of and domestic prices induced by the rand’s 2016, growing at 0.7 percent y/y in 2016 Q1, depreciation (more on this in Chapter 2). accelerating markedly to 2.8 percent in Q2, Nonetheless, growth in 2016 was low—a supported by a large positive production shock fact that comes out even more starkly when to manganese and iron ore. Automotives, measuring GDP per head, which points to a tourism, as well as vegetables and prepared third year of falling GDP per capita growth, food and beverages also contributed to export making South Africans poorer compared with growth in the first half of 2016. Yet the positive previous years. Figure GDP growth (demand side) 1.5 (contribution to year-on-year growth, %) Private consumption expenditure 7 Government consumption expenditure Investments 6 Exports of goods & non-factor services Imports of goods & non-factor services 5 4 GDP by expenditure approach 3 2 1 0 -1 -2 -3 -4 2013 Q2 2013 Q3 2014 Q2 2014 Q3 2015 Q2 2015 Q3 2016 Q2 2016 Q3 Source: StatsSA. SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS On the production side, some positive percent y/y in the first three quarters of 2016. results are emerging from agriculture, where Although agriculture continued to decline a drought associated with the global climate in the first three quarters of 2016 y/y (Table phenomenon El Niño (Box 1.2) shaved 0.2 1.1), it did so at a slowing rate, and the sector percentage point off 2015 growth and turned is about to turn around again—not least South Africa into a net food importer, with supported by strong precipitation in late agricultural imports increasing (in nominal 2016. Nonetheless, although it is expected to terms) by 9.7 percent y/y in 2015 and 29.4 rebound, agricultural output has fallen back 10 Table GDP growth (supply side) 1.1 (contribution to year-on-year growth, %) Quarter Agri- Mining Manu- Elec- Con- Trade, Trans- Finance, Gen- Per- Net GDP culture, and factur- tricity struc- cater- port real es- eral sonal indirect forestry, quarry- ing and tion ing, and tate, and govern- ser- taxes and ing water and com- business ment vices fishing accom- munica- services services moda- tions Low domestic tion demand resulted in falling imports 2013 Q1 6.0 4.1 -0.1 -2.8 4.4 1.9 2.5 1.7 3.2 1.9 1.3 2.0 2013 Q2 7.9 -1.4 2.6 -0.1 5.4 2.0 2.4 2.4 3.0 1.9 2.7 2.4 2013 Q3 -1.6 3.4 -0.3 0.4 5.1 1.7 3.1 2.5 2.6 2.2 1.4 2.0 2013 Q4 0.2 10.0 1.1 -0.1 3.4 1.8 3.0 3.7 2.8 2.8 1.4 2.9 2014 Q1 -0.6 -1.1 1.9 0.0 4.0 1.9 2.5 2.5 2.8 2.0 0.7 1.8 2014 Q2 7.8 -0.7 -1.3 -2.2 3.4 0.9 3.1 2.3 3.1 1.6 -0.1 1.5 2014 Q3 13.0 -2.4 -0.2 -2.5 3.3 1.6 2.9 2.5 2.8 1.7 0.7 1.7 2014 Q4 5.3 -1.4 0.2 -0.2 3.7 1.1 3.7 2.2 2.0 1.4 1.6 1.5 2015 Q1 11.7 6.8 0.2 2.1 2.4 1.5 2.5 2.7 1.4 1.1 4.0 2.5 2015 Q2 -6.7 5.2 -1.8 -0.4 2.3 1.1 1.6 3.2 0.7 1.4 1.6 1.2 2015 Q3 -18.7 1.5 1.5 -2.7 1.9 1.2 1.1 3.2 0.4 1.1 0.6 0.8 2015 Q4 -5.0 -0.5 -1.2 -2.9 1.6 1.7 0.5 2.4 0.3 0.7 0.1 0.5 2016 Q1 -2.9 -8.8 -0.9 -4.5 2.5 1.2 -0.6 2.3 1.8 0.9 -2.0 -0.1 2016 Q2 -12.0 -3.4 3.6 -2.6 0.3 1.8 0.1 2.2 1.7 1.0 -0.1 0.7 2016 Q3 -3.6 -0.1 -0.4 -1.8 1.4 0.5 0.1 1.8 1.7 1.4 0.6 0.7 Source: StatsSA. to 2012 levels, and regaining lost production finance, real estate, and business services in potential will take time. El Niño also adversely particular. Mining continued to restructure in affected the electricity and water sectors, as response to the end of the commodity super water restrictions were put in place due to cycle. The sector shaved 0.3 percentage point reservoirs running at critically low levels. (y/y) off headline GDP growth. Yet this was Mining and manufacturing, which have largely due to an unusually strong performance been identified by the government through in the first half of 2015, as the sector rebounded the National Development Plan and Industrial from extensive industrial action. On a q/q Policy Action Plans (at least in the case of basis, the sector made headway toward manufacturing) as strategic drivers of growth potential production in Q2 and Q3 of 2016. and job creation (see Chapter 2), have been Manufacturing added positively to GDP, 0.1 moving sideways (Figure 1.6). Jointly, the percentage point y/y, for the first time since sectors account for about one-fifth of GDP, 2013 due to strong performance in Q2, largely although their share has been falling, as the driven by the petroleum and automotive sectors have been outperformed by services— sectors. Yet, in spite of some positive (although still mixed) signals from the mining and of many of the companies. The Top 40 index manufacturing sectors, manufacturing is likely closed the year 4.1 percent below the level at to remain outperformed by the service sector. which it opened in January 2016. The financial, real estate and business Other growing sectors in the service services sectors have been South Africa’s industry include construction, not least as drivers of growth for a while. They account a consequence of continued additions to for just over a fifth of GDP, extending their South Africa’s electricity supply, including share as South Africa continues on its path by independent power producers. The of structural transformation (see Chapter 2). construction sector grew by 1.4 percent y/y in At 2.1 percent y/y growth in the first quarters the first three quarters of 2016. Trade, catering, 11 of 2016, the sectors maintained their role as and accommodation also experienced growth, South Africa’s engine of growth—but the of 1.1 percent y/y, at least partly supported engine has been slowing from previous levels, by a buoyant tourism sector. Public services not least due to a weak domestic economy, grew by an average 1.7 percent, driven by an including slowing consumer credit and a increase in spending on goods and services weakening real estate market. This slowing and also employment—where the payment Services have been was amplified by the weaker performance of of temporary electoral staff for the municipal companies listed at the Johannesburg Stock elections in August 2016 contributed to the carrying growth Exchange, which was also a reflection of weak increase in government service growth. in 2016, although global growth, given the international reach Finance, the engine Figure Finance, insurance, and real estate have been driving growth of growth, has been 1.6 (index, 2008 Q1 = 100) slowing 130 125 ors 120 sect The drought a l e state th re hit agricultural and grow 115 cial iving hard but h e finan been dr T effects are have 110 dissipating 105 100 95 Mining and maufacturing are moving sideways 90 85 2008 Q1 2009 Q1 2010 Q1 2011 Q1 2012 Q1 2013 Q1 2014 Q1 2015 Q1 2016 Q1 Agriculture, Forestry Mining & Manufacturing Finance, real estate and & Fishing Quarrying business services Source: StatsSA; World Bank staff calculations. Uncertainty on investors’ minds investors who were surveyed in the second Investment is urgently needed to propel half of 2014 identified the volatility of the South African growth in the medium to longer rand as a major constraint to doing business term, and help the country meet its aspirations in South Africa. In 2016, the rand continued enshrined in the National Development Plan. to be one of the most volatile currencies in The volatility of the rand is a major concern emerging markets (Figure 1.7). A recent study of foreign investors—and South African by the International Monetary Fund suggests exporters and importers—and has been so for that the volatility of the rand is mostly driven a while. For example, 91 percent of European by commodity price shocks (which affect the SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Figure Exchange rate volatility in 2016 1.7 (standard deviation of daily exchange rate, normalized) 0.09 0.08 0.07 12 0.06 0.05 0.04 0.03 Greater Policy 0.02 certainty is key 0.01 to encourage 0 Brazil Russian Turkey South Japan Mexico Poland Euro Zone China Indonesia India investment (Real) Federation (Lira) (Ruble) Africa (Rand) (Yen) (Peso) (Zloty) (Euro) (Yuan) (Rupiah) (Rupees) Sources: Haver Analytics; World Bank staff calculations. profitability of many South African companies positive developments, giving investors cause and thus financial flows), global market for optimism. On the political front, South volatility (shifting global financial flows to African institutions proved their resilience riskier or less risky assets), as well as, notably, and maturity yet again in 2016, not least in domestic policy uncertainty.1 the smooth holding of municipal elections, Indeed, investors also fret about politics, which saw opposition parties take over several ranging from concerns around state major cities in August. On the economic capture, investigated by the Public Protector, front, unreliable electricity supply, which has to uncertainty around certain pieces of been making load-shedding a South African legislation. Mining investment hinges on the regularity since 2007, is being addressed and passing and modalities of the Mineral and showing results: 2016 did not experience Petroleum Resources Development Act, as major periods of load-shedding, partly well as a new Mining Charter; agricultural because of suppressed demand in a weak investment depends on the details of land economy, but also due to new capacity coming reform and the settlement of restitution online (including one unit of Eskom’s coal- claims on land. Although a trade spat with the fired Medupi power plant—with the next U.S. around imports of certain agricultural one expected in March 2017—and several products, especially poultry, was settled earlier independent power producers, many of in the year, it continues to simmer, raising them producing renewable energy) (see also concerns over South African duty-free access Chapter 2). to the U.S. market under the African Growth Moreover, the year 2016 witnessed relatively and Opportunity Act. A tightening of visa few strikes. A wage agreement in the automotive regulations for tourists for certain countries sector was struck in August, unaccompanied has been relaxed, stimulating tourism inflows by industrial action for the first time in seven again. The government undertook road shows years. Other than weak corporate profitability, in 2016, aiming to unlock foreign investment which renders pay increases less realistic, the that has been held back, largely due to political reduction in the number of strikes is illustrative uncertainty. of the government’s efforts to improve labor There have also been some noteworthy relations. Further efforts, such as rules around secret strike ballots and improvements in labor manufacturing. mediation systems, if implemented, could make significant contributions to mending Labor Market Developments South Africa’s fractious labor relations. The introduction of a national minimum wage has The specter of unemployment continues to become more likely, and is suggested to be set haunt workers at ZAR 3,500 per month. South African policy Unemployment is notoriously high in South makers are treading a careful line, endeavoring Africa. In 2016 Q3, the unemployment rate to ensure that the minimum wage makes a edged up by 1.6 percentage points compared meaningful difference for low-wage earners with Q3 the previous year, touching 27.1 13 while minimizing the impact on employment percent, the highest recorded level in 13 years.2 (for example, by providing exemptions to On the one hand, this Q3 development reflects agriculture and domestic work and phasing in an increase in labor force participation, which the minimum wage over a period of time). signals that more South Africans are willing to On balance, investor confidence remains work. Yet, according to the Quarterly Labor at low levels, although early 2016 witnessed Force Survey (QLFS), only 5,000 additional a modest increase in optimism. The Bureau jobs were created in net terms over the year Unemployment of Economic Research Business Confidence leading up to 2016 Q3, while the number of reached a 13-year Index edged up 10 points from a six-year low unemployed South Africans increased by of 32 in 2016 Q1, to a more positive reading of 455,000, which includes net job losses and new high 42 in Q3, a level last seen in 2014. Optimism labor market entrants who do not find jobs. remains more cautious in manufacturing, The unemployment rate of youths aged however, chiming with the mixed performance 15–24 years was 54.2 percent in 2016 Q3, up of the sector over the year. While the global 4.3 percentage points from the first half of PMI was trending (modestly) above the 50 2015. Just over a quarter of this age group mark, cautiously pointing toward expansion participates in the labor market, with others (see Figure 1.1), in South Africa the seasonally continuing in the education system. The adjusted PMI only registered readings youth unemployment rate underscores the above 50 in March through July, pointing importance of acquiring skills to be employed in to a more persistent sense of pessimism in one of Africa’s most technologically advanced Figure Unemployment and economic inactivity 1.8 (percent of labor force or working age population, by education, 2016, Q3) a. Unemployed b. Economically inactive 60 60 50 50 National average: 40.9% 40 40 National average: 27.1% 30 30 20 20 10 10 0 0 Less than Secondary Tertiary Less than Secondary Tertiary secondary completed secondary completed Unemployed Economically inactive Sources: StatsSA; World Bank staff calculations. SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS economies. Unemployment is highest this is still high by most standards—and a likely among those who do not hold a high school reason for the continued student protests in diploma, at 31.6 percent (Figure 1.8), nearly 2016—it is less than half the national average. 5 percentage points higher than the national Box 1.1 discusses the challenges faced by the average. For those with tertiary education, the education sector in raising the skills of job unemployment rate is 13.2 percent. Although seekers. Box Education and labor market developments in South Africa 1.1 South Africa’s education and training system faces multiple challenges in ensuring that all children have the knowledge, 14 skills, and attributes they need to be successful workers and citizens. Key challenges include the low—by international standards—and inequitable learning levels of children. Only 24 percent of grade 5 children can answer correctly the following question: Pam has ZAR 40. She spends ZAR 28. How much money does she have left? There are high repetition and drop-out rates in secondary schooling. The proportion of Black Africans who obtain a matric qualification (high school diploma) has been stagnant at approximately 52 percent since 1994, while pass rates for Coloureds have been falling since 2000, see Box 1.1 figure 1; in both cases, there is a wide and persistent gap vis-à-vis the white population. Those who do not complete secondary education are finding it more difficult to obtain work: between Education holds 2008 and 2015, there were significantly fewer jobs for those in this category. In contrast, new jobs increasingly went to those with secondary or tertiary education. the key to reducing Box 1.1 figure 1: Inequalities are still persistent in groups obtaining a matric qualification unemployment and (proportion) poverty 1.0 0.9 0.8 0.7 0.6 Black Africans 0.5 Coloureds Asians 0.4 Whites 0.3 0.2 0.1 0.0 1940 1950 1960 1970 1980 1990 2000 2010 Year Source: National Department of Basic Education, based on StatsSA General Household Survey, 2002–14. Low participation and high unemployment rates among workers with low education attainment translate into a life- long poverty trap: of the population with no schooling attainment, over 50 percent has no income, see Box 1.1 figure 2. Box 1.1 figure 2: Relationship between schooling and income status at different ages (no income) 80% 70% 60% No schooling 50% Grade 11 40% Grade 12 30% Diploma 20% 10% Bachelor degree 20 30 40 50 60 70 80+ Age Source: National Department of Basic Education, based on StatsSA General Household Survey, 2002–14. Box Education and labor market developments in South Africa 1.1 Tackling these challenges requires making the following priority investments. First, it requires a focus on ensuring fluency in reading and mathematics by grade 3, which is part of the government’s plans. Rigorous evaluations have demonstrated the importance of teaching at the competency level students have (rather than focusing on teaching the grade level material). Successful models should be tried and evaluated in South Africa. Second, through secondary education, the focus should be on a general set of foundational skills and aptitudes. This focus will help to provide what is required by most employers, and these skills and aptitudes will give young people greater resilience in the labor market, since specialized technical and vocational skills fade much more quickly during periods of unemployment or labor market inactivity. (Sadly, recent experience suggests that many young South Africans will not be able to find work immediately after completing education and training.) 15 Third, workplace experience is the best place to learn about the working world and acquire new skills that are relevant to today’s job market. Work experience opportunities for young people, before they leave education and training, and for the long-term unemployed, is important. These opportunities should be made available to all students, not just those in vocational education and training programs. Fourth, as South Africa continues to meet its commitment to minimum standards in its education infrastructure, it will have to take into account the overall reduction in the size of the student cohorts and the highly mobile student population (with many students living with extended family members to attend a particular school). Adjusting the school network will be a long-term and complex process. It will have to balance local decision making to meet local dynamics Skills shortages and enable a more rapid response, improving efficiency but informed by the full costs of school consolidation (capital and recurrent spending, transportation, and so forth) and measures of the impact on the learning and engagement of students are also a major in academic and nonacademic areas. constraint for The skills shortage is mirrored by the 2016 the largest declines in 2016, behind skilled businesses and Talent Shortage Survey,3 as the number of South agriculture. Indeed, manufacturing is a sector African employers surveyed who had difficulty that is saturated with respect to employment: growth filling positions increased 3 percentage points just under 19 percent of capacity was since 2015 (and 26 percentage points since underutilized in the sector in the first half of 2014), to 34 percent. Among the most difficult 2016 (a 1 percentage point improvement over skills to find are those in skilled trades and the same period the previous year). Of that management and executive positions. This underutilization, skilled labor only accounted finding reflects the need for more talented for 1 percent and semiskilled and unskilled individuals, as the main reasons for having labor even less than that, 0.2 percent. difficulties filling positions include lack of High-skill services thus created the most experience (27 percent), lack of hard skills jobs. According to the QLFS, finance and (26 percent), and lack of available applicants related services added 7.6 percent y/y to their (9 percent). At the other end, lower-income employee base in 2016 Q3, equivalent to more South Africans—those who tend to be less than 163,000 jobs, followed by more modest skilled—more often than not have to resort to increases in construction (2.1 percent y/y, or bribes to obtain jobs. A recent survey by the 31,000 jobs) and transport (1.9 percent y/y, Ethics Institute shows that obtaining a job is or 17,000 jobs). Most other sectors shedded the second most important reason for bribery jobs, led by manufacturing (91,000 jobs lost)— in South Africa, whereby those who earn less the backbone of South Africa’s industrial than ZAR 100,000 a year are 17 percent more economy—and followed by community and likely to pay a bribe for a job than those in social services (83,000 jobs lost), agriculture high-income groups are.4 (16,000 jobs lost), and mining (8,000 jobs The need for skills is also reflected in the lost). Thus, the increase in the financial sector professions that experienced the highest job only modestly offset job losses in other sectors, growth. According to the QLFS, 10.3 percent explaining the low rate of net employment more professionals were hired on average growth. The fall in agricultural employment, in the first three quarters of 2016 over the a consequence of lingering drought effects, same period the previous year. Professions spells bad news especially for poorer associated with lower skills, such as clerks households. and elementary and domestic workers, saw Despite the comparably strong employment decline. Blue collar jobs, such performance of the finance, real estate, as plant and machine operators, also saw and business service sector in 2016, its wage employment decline, by -3.8 percent, one of growth has been moderate compared with SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS several other sectors that experienced a more of -0.5 percent, putting pressure on working mixed performance, according to the QLFS. households’ budgets. Wages per worker in the sector grew by 5.5 At the same time, however, there appear to percent y/y in nominal terms, lower than in have been redundancies in sectors that were mining (10.4 percent), community and social relatively inefficient, as productivity growth in services (6.8 percent), and manufacturing the nonagricultural sectors increased in early (5.8 percent). Overall, wages per worker in 2016, for the first time since 2014, meaning nonagricultural sectors grew by 5.9 percent that South Africa managed to produce more y/y in the first three quarters, below the rate output per employed person (Figure 1.9). 16 of inflation. This resulted in real wage growth The manufacturing sector has been Figure r Labor productivity and unit labor costs 1.9 (index, 2011 Q1 = 100, seasonally adjusted) 160 Labor productivity Unit labor cost 150 As productivity increases, falling 140 unit labor costs 130 make hiring more 120 attractive 110 100 90 2011/01 2012/01 2013/01 2014/01 2015/01 2016/01 Source: South African Reserve Bank. Figure r Labor productivity growth 1.10 (growth in output per worker, %) Total (including agriculture) Total (non-agriculture) Mining Agriculture Utilities Construction Finance Community and social services Transport Trade Manufacturing -30% -25% -20% -15% -10% -5% 0% 5% 10% 15% 2011 (Q1-Q3) - 2016 (Q1-Q3) 2015 (Q1-Q3) - 2016 (Q1-Q3) Sources: StatsSA; World Bank staff calculations. restructuring, improving its labor productivity Consumer Price Index—which includes core compared with 2011 and 2015 (Figure 1.10). inflation as well as food, and nonalcoholic The growth in productivity in 2016 Q2, beverages, petrol, and energy—was pushed combined with more modest wage growth over the South African Reserve Bank’s (SARB) compared with previous years, also lowered upper target of 6 percent, especially because unit labor costs (in q/q and seasonally adjusted of food prices, which soared in 2016 (Figure terms), the first such reduction since 2014 and 1.11). One reason was the 2015 drought, the strongest since 2007. Harnessing such whose effects lasted well into 2016. Food prices gains in competitiveness will be important to increased by an average 10.6 percent between raise job prospects and improve livelihoods January and November 2016, more than twice 17 for more unemployed South Africans, and the average of the same period the previous to bring down the high unemployment rate year. Food inflation peaked at 12 percent in of 27.1 percent in 2016 Q3—or 36.3 percent October. These developments hit the poor when including the discouraged, in other particularly hard, given that food accounts for words, those who have given up looking for the largest part of their consumption basket. work. The drought was a consequence of shifts in global weather patterns attributed to the The poor were Inflation and Monetary Policy El Niño phenomenon. While El Niño hit particularly affected Southern African agriculture hard (Box Inflation remains above target but is easing, 1.2), Latin American farmers experienced a by high food reducing pressure on policy rates bumper harvest, helping meet South Africa’s Core inflation hovered between 5.4 and import needs (traditionally, South Africa is a inflation 5.7 percent during 2016. Inflation in the net food exporter to the subregion). Box Macroeconomic and poverty impacts of El Niño on Southern African 1.2 Development Community countries Using the LINKAGE global computable general equilibrium (CGE) model, World Bank (2016c) estimates the short-term macroeconomic and poverty impacts of El Niño–related droughts. Absent sufficient information on projected hydroelectricity production, cattle destocking, and the impact of the drought on other crops, World Bank (2016c) concentrates on the short-term impact of reduced maize production (an 18 percent decline in 2015/16 at the regional level, from a situation of self-sufficiency in 2014/15). Except for Madagascar, maize constitutes Southern African households’ main staple, and it is believed that the impact on maize production constitutes the largest channel through which El Niño affects households’ welfare. Impacts are measured through the comparison of a business as usual (“baseline”) scenario with a scenario that assumes reduced total factor productivity in the maize sector to reproduce anticipated countries’ maize production in 2015/16. Through modeling the supply and demand sides at the country level, and trade relationships among Southern African Development Community (SADC) countries and with the rest of the world, the CGE is employed to capture orders of magnitude of El Niño impacts on prices, household incomes, and demand. Combined with household surveys, these estimates are also used to compute poverty impacts. The simulation results suggest that, at the regional SADC level, gross domestic product (GDP) growth decelerated by 0.1 percent in 2016 because of droughts, and by 0.05 percent in South Africa, the largest economy in SADC. Given the relatively high importance of maize in GDP, the growth impact of El Niño was felt to be much larger in Malawi (-2.2 percent); Tanzania (-1.4 percent); and Zimbabwe, Lesotho, and Swaziland (-0.6 percent). Given the assumed low elasticity of households’ maize demand to income, the simulations suggest that households reduced their demand for other goods and services to satisfy their maize consumption. This resulted in a surge in maize prices in 2016 that the CGE model anticipated in the range of 15–45 percent for the SADC countries.a In turn, countries such as Botswana, Lesotho, Namibia, and Swaziland, which traditionally rely to a large extent on imports of maize from other SADC countries (Zimbabwe, South Africa, and Malawi), were hard hit on the consumption side. The combination of these effects is measured by the induced cost of maintaining pre-crisis real household consumption levels at post-crisis prices and GDP levels (compensating variations).b Poor households were particularly hit by price shocks, as they devote larger shares of consumption to maize (consistent with income elasticities lower than unity). The simulation results suggest that the real per capita consumption of the bottom 40 percent of households residing in SADC countries declined by 1.7 percent, against only 0.1 percent for the population as a whole (box 1.2 table 1). The negative impact was particularly pronounced in Malawi (-11.8 percent), Tanzania (-7.5 percent), Lesotho (-6.2 percent), Swaziland (-4.7 percent), and Zimbabwe (-3.5 percent). It was only -0.5 percent in South Africa. Sadly, the countries that were most affected by El Niño are also those in the SADC region facing the tightest fiscal and external situations. Although foreign assistance is needed to address the effects of the current crisis, there is a need to integrate short- and longer-term mitigation solutions. Mitigation solutions can be grouped into three categories, responding to three complementary objectives: available financial resources, continued food supply, and protection of the poorest segments SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Box Macroeconomic and poverty impacts of El Niño on Southern African 1.2 Development Community countries (continued) of population. SADC countries could usefully (i) adopt countercyclical fiscal policies, (ii) rely on ex ante contingent financial instruments, (iii) develop regional risk-sharing mechanisms, (iv) reduce regional barriers to agricultural trade, (v) encourage market-based storage solutions, (vi) introduce physical and financial hedging mechanisms in food procurement, (vii) develop targeted and swiftly scalable safety nets, and (viii) facilitate the introduction of micro-level insurance programs. Box 1.2 table 1: Macroeconomic and social impacts of reduced maize production in 2016 Compensating variation Variation in real private consumption 18 (% deviation w.r.t Baseline 2016) (%of GDP 2016) Bottom 40% Botswana 0.14% -2.45% Lesotho 0.54% -6.18% Madagascar -0.11% 0.56% Malawi 1.90% -11.78% Mauritius -0.02% 0.10% Mozambique -0.27% 1.53% Namibia 0.08% -1.03% El Niñ o hit South Seychelles -0.14% 0.76% South Africa 0.03% -0.52% Africa hard but Swaziland 0.54% -4.73% Tanzania 1.43% -7.48% less so than its Zambia -0.13% 1.61% Zimbabwe 0.72% -3.54% neighbouring Southern Africa Development Community 0.07% -1.66% countries Source: World Bank (2016f). Facing the current crisis, several SADC countries have approached the World Bank for short- and longer-term solutions. Demands have ranged from post-crisis needs assessments and budget support, to advice on risk mitigation options and agriculture insurance programs. Initial responses from the World Bank have concentrated on (i) portfolio screening to identify opportunities to support national emergency plans, (ii) development of emergency response financial operations, and (iii) analytical overviews of potential longer-term disaster risk management instruments that could meet SADC countries’ specific needs. a. Between March 2015 and March 2016, nominal retail prices of white maize recorded the following variations: Zimbabwe: +20 percent; Zambia: +41 percent; Tanzania: +64 percent; Swaziland: +54 percent; South Africa: +98 percent; Mozambique: +121 percent; Malawi: +152 percent; and Lesotho: +33 percent. However, changes in food prices observed on markets cannot necessarily be fully attributed to lower food supplies. Indeed, since 2015, SADC countries have seen their overall macroeconomic situation deteriorate under the influence of two main factors: the decline in commodity prices, and the reversal in capital flows due to rising global uncertainty. As a result, GDP growth in SADC decelerated from 3.8 percent in 2013 and 3.0 percent in 2014, to 2.3 percent in 2015, and is projected at 2 percent in 2016. Reversal in capital flows and lower export receipts led to sharp currency depreciation in several SADC countries, pressures on the balance of payments, and accelerated inflation from exchange rate pass-through. b. For the sake of simplicity and to focus on the supply-side effects in SADC, these simulations do not factor in the decline in the world price of maize recorded in recent years, from US$178 per ton in November 2014 to US$152 in November 2016 (World Bank 2016b). Thus, it is likely that the simulations marginally overestimate the welfare costs of El Niño. In U.S. dollar terms, food grain prices fell 20 percent of urban residents experienced by 9.9 percent in 2016, but the significant average inflation of 7.3 percent, while the depreciation against the dollar sent the rand richest 20 percent experienced inflation that value of food imports soaring. In addition, was 1.2 percentage points lower. Although the falling U.S. dollar–denominated food prices exchange rate had a large effect on the price of triggered increases in agricultural import food, fuel, a major South African import, did tariffs for wheat in May, which further not drive inflation significantly, as the decline increased food prices for consumers. The in U.S. dollar–denominated fuel prices was main staple foods, such as bread and cereals, partly offset by the weaker rand. accordingly experienced the highest inflation, Inflation outside its upper target peaking at 17.2 percent in November 2016. preoccupied the SARB through much of Inflation affected households in different ways, 2016, resulting in cumulative rate hikes of 75 given their spending patterns. The poorest basis points in January and March. The SARB maintained its tightening bias in subsequent temporary. Inflation expectations for 2017 meetings of the Monetary Policy Committee have come down from an annual 6.2 percent, (MPC). The food price trajectory was a major which was held early in the first half of 2016, concern, but the SARB also showed concern to 6 percent—thus moving into the SARB’s over many downside risks (such as the Brexit), inflation bound. Accordingly, the MPC did not keeping MPC members cautious. Yet inflation decide to implement further rate hikes. surprised on the downside over the year, A fall in expected inflation raises expected partly due to the rand strengthening more real interest rates. The real repo rate in 2016 than expected through early 2016, lowering Q3 stood at 0.8 percent (7 percent nominal) predictions for the peak of food prices and the prime lending rate at 4.3 percent 19 through the first half of the year. Although (10.5 percent nominal). Although this makes inflation crept above the SARB’s 6 percent rate hikes less likely, a higher real interest rate upper target again in September and through (now due to disinflation) further deters the November, this breach was considered investment that South Africa urgently needs. Figure Consumer price index inflation and contributions Inflation was met 1.11 (year-on-year growth, %) with rising policy 8 rates 7 6 5 4 3 2 1 0 -1 -2 Jan-14 Jul-14 Jan-15 Jul-15 Jan-16 Jul-16 Core inflation Food & nonalcoholic beverages inflation Petrol inflation Electricity & other fuels inflation Headline inflation Inflation target band Sources: South African Reserve Bank; World Bank staff calculations. Fiscal Developments effort is increased by another ZAR 16 billion (100 percent from expenditure). On the Fiscal policy continues to consolidate expenditure side, the wage bill will see further In the 2016 budget and the October cuts (Table 1.2), of 0.4 and 0.7 percent below Medium-Term Budget Policy Statement what was foreseen in the budget for 2017/18 (MTBPS), the government remains and 2018/19, respectively. This would help committed to gradual fiscal consolidation. stabilize the public wage bill, as a percentage Given lower growth projections, the MTBPS of GDP and of government expenditure. proposed to increase the consolidation effort. Reducing government positions and limiting Although the 2016 budget originally foresaw promotions thus counters pressures from a an additional fiscal effort of ZAR 25 billion (60 three-year wage agreement the government percent of which from revenue) for 2017/18, entered into with public sector unions in the MTBPS adds another ZAR 23 billion (56 2015, which had put the wage bill on an percent from revenue). For 2018/19, the fiscal unsustainable path. SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Given persistent under-execution, lower above 10 percent in 2017/18 and 2018/19. levels of government are also hit hard This is a significant upward revision from the by expenditure restraint. For 2017/18, budget, given that debt will increase faster than allocations remain almost constant in real expected. Yet the fastest growing spending terms, and only pick up modestly in 2018/19, category is now tertiary education, giving in, and more strongly (growing at 10.4 percent in to an extent, to demands by the #FeesMustFall nominal terms) in 2019/20. Compared with student movement. The budget already foresaw the original budget allocations, in 2017/18 an increase in tertiary education spending lower levels of government experience cuts of of more than 11 percent for 2017/18, and 20 1.3 percent, and even higher, at 3.8 percent in further allocations were made for 2018/19 and 2018/19. 2019/20. These allocations at least partly come With respect to spending areas, debt at the expense of spending on basic education service costs continue to be one of the fastest (which is arguably more pro-poor), where growing budget items going forward, growing allocations have been cut. Within a tightening Table Expenditure projections and deviations from budget (2016/17 to 2019/20) budget, the MTBPS 1.2 (percent) 2016 MTBPS w/y change 2016 MTBPS % deviation from 2016 increased allocations 2016/17 2017/18 2018/19 2019/20 2015/16 2016/17 2017/18 2018/19 for universities. Current payments 9.4 7.5 7.5 7.4 -0.9 -0.2 -0.2 -0.3 Compensation of employees 8.9 6.7 6.8 7.3 -0.8 -0.3 -0.4 -0.7 Raising additional Goods and services 7.2 7.4 6.8 7.3 -1.5 -0.1 -1.0 -0.9 Interest and rent on land 14.2 10.5 10.4 9.1 0.0 0.1 1.2 1.5 revenue will be key of which: debt-service cots 14.7 10.8 10.5 9.1 -0.2 0.0 1.1 1.3 Transfer and subsidies 6.5 8.2 6.5 7.2 -1.4 -1.2 -0.1 -1.8 to stay the course of Provinces and municipalities 4.6 6.4 7.6 10.4 -0.5 0.0 -1.3 -3.8 Departmental agencies and accounts -4.7 8.2 6.3 4.6 -1.6 -2.6 -1.2 -1.1 fiscal consolidation Higher education institutions 11.0 22.9 5.8 13.5 -0.1 3.9 13.0 12.7 Foreign goernments and international organisations 3.3 -8.6 6.5 26.2 1.8 1.3 -12.5 -13.4 Public corporations and private enterprises 0.3 11.3 7.6 -5.0 -8.1 -9.1 -7.3 -11.2 Non-profit institutions 7.0 3.8 4.5 5.6 -0.8 2.0 2.9 3.6 Households 9.1 7.4 6.2 6.8 -1.1 -1.5 -0.4 -0.7 Payments for capital asests -1.4 3.2 4.6 7.9 4.8 0.7 -1.1 -1.1 Buildings and other capital assests -2.9 1.6 5.0 6.4 6.5 -1.0 -3.1 -2.8 Machinery and equipment 4.2 9.2 3.5 13.0 -1.2 7.4 5.9 4.8 Payments for finacncial assets -79.6 -16.2 -2.2 5.6 2.0 17.8 0.0 0.0 Total 5.7 7.3 6.9 7.4 -0.6 -0.4 -0.3 -0.8 Contingency reserve Consolidated expenditure 5.7 7.8 7.2 8.0 -6.0 -0.8 -0.5 -1.1 Sources: National Treasury; World Bank staff calculations. On the revenue front, the MTBPS proposes potentially an increase in the value-added tax. to raise another ZAR 13 billion in 2017/18, Overall, expenditure is expected to in addition to the ZAR 15 billion proposed remain stable as a percentage of GDP between in the 2016 budget (and an additional ZAR 2016/17 and 2019/20, while revenue is 15 billion in 2018/19). However, neither the expected to increase from 29.7 to 30.4 percent budget nor the MTBPS identifies what specific of GDP over the same period. Revenue is thus measures will yield this additional revenue. doing the heavy lifting to reduce the budget About ZAR 5 billion can be expected from a deficit, which is now projected to fall from 3.7 partial tax amnesty (the Voluntary Disclosure percent of GDP in 2015/16, to 3.4 percent Programme), yet this would only account for in 2016/17, to 2.7 percent in 2018/19, and about 12 percent of the ZAR 43 billion that 2.5 percent in 2019/20, roughly maintaining needs to be raised. Likely measures are limited the consolidation path foreseen in the 2016 fiscal drag for personal income tax, further budget. Yet, given the increase in expected excise taxes, increases in the fuel levy, and interest payments, the budget balance in the Figure Trajectory of net public debt across budgets and MTBPSs 1.12 (percent) 50 48 46 21 44 42 40 South Africa 38 defended its 36 investment grade 2014/15 2015/16 2016/17 2017/18 2018/19 2019/20 2015 Budget 2015 MTBPS 2016 Budget 2016 MTBPS credit rating Sources: National Treasury; World Bank staff calculations. Note: MTBPS = Medium-Term Budget Policy Statement. MTBPS deteriorated markedly compared with the two rating agencies affirmed the rating yet the budget, by 0.2 percentage point of GDP again, although Fitch joined S&P in putting in 2016/17, and 0.3 percentage point of GDP South African on a negative rating outlook. in 2017/18 and 2018/19. This deterioration The same analysis suggests that the first shifts the debt stabilization into the future. rating downgrade to “junk” carries particular Debt stabilization will now occur two years weight for borrowing costs, while the second later than foreseen in the budget (in 2019), downgrade adds to it (Box 1.3). The two rating and is 1.9 percentage points of GDP higher agencies seeing South Africa at BBB- with a than envisioned in the budget (Figure 1.12). negative outlook means that the possibility of South African foreign currency– a downgrade will continue to occupy South denominated debt was reaffirmed in African policy makers and investors alike in December, with a rating one notch above 2017. Yet, the fact that South Africa managed “junk” by Standard and Poor’s (S&P) and to avert a downgrade to “junk” in 2016 speaks Fitch in June. However, an analysis conducted to the fine balancing act the National Treasury by World Bank and SARB economists has been delivering, through the budget and suggested that one downgrade had already the MTBPS. been priced in, at least partly. In December, SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Box The threat of a rating downgrade to “junk” 1.3 South Africa has come a long way since the advent of democracy in 1994. The democratic government embarked on impressive efforts to make the economy inclusive, considerably expanding social spending and extending public services to those who had been underserved under apartheid. The government managed to balance its new spending priorities with fiscal prudence—a remarkable achievement rewarded by two of the world’s major credit rating agencies, Standard and Poor’s (S&P) and Fitch, with the highly prized BBB- rating in 2000, judging South African debt to be of investment grade. By 2005, South Africa’s rating had improved to BBB+. A decade on, the trend had reversed, as growth had slowed and public debt increased markedly since the global financial crisis, bringing South Africa back to its BBB- rating by S&P and 22 Fitch, to which S&P added a negative outlook in December 2015, with Fitch following suit in December 2016. South Africans and investors were closely watching the reviews of rating agencies in 2016, and the markets partly priced in at least one downgrade to sub-investment grade (sub-IG). To understand what the potential impact of South Africa losing its investment-grade credit rating might be, a team of World Bank and SARB economists looked at the history of past downgrades in 20 countries and estimated the impact on short-term borrowing costs. The study employs annual data and only looks at short-term Treasury bills (where risk premia are more pronounced among longer-term debt instruments), and there are still relatively few cases to study (most of them derive from the post–financial crisis period). Yet, given these important caveats, the findings suggest that a downgrade to sub-IG by one major rating agency increased Treasury bill yields by 138 basis points on average in the sample under study. When a second rater followed suit with such a A rating downgrade downgrade, Treasury bill rates increased by another 56 basis points (although this effect is not statistically significant). Since would have raised the markets already partly priced in a downgrade for South Africa in 2016, the impact would have been less, estimated at about 60 basis points. borrowing costs What would this downgrade mean for the economy? The World Bank modeled the impact of a risk premium shock that increases the Treasury bill rate by 60 basis points in December 2016 (the time of the S&P and Fitch decisions) on the South and reduced growth African economy. This shock represents uncertainty and capital flight. The uncertainty shock materially affects investment (Bloom 2009) and consumption decisions through higher cost of capital and an endogenous interest rate response. The further flight of capital causes the nominal exchange rate to depreciate by 6 percent initially, which tapers back to its baseline as the economy rebalances. The more competitive currency will improve trade, assuming the historical empirical estimates of the exchange rate and trade hold. The trade response does not offset the fall in investment and consumption. However, the pass-through to inflation would have been significant (the maximum impact on inflation occurs approximately two quarters after the shock). The response of gross domestic product (GDP) factors is a reaction of the SARB to higher inflation. The impact on GDP would have been most pronounced within two quarters of the shock (Box 1.3 figure 1), shaving a cumulative 1 percent off real GDP by end- 2017. This would translate into foregone nominal income of about ZAR 1,000 on average per South African. The shock only dissipates three years after the shock, despite being temporary. The model assumes that the authorities will react to the higher debt and ratings downgrade by reducing future expenditures, in an attempt to stabilize the debt. This implies that public debt increases initially due to the shock, but ends up being lower after adjustments to expenditure. GDP could still be lower if the exchange rate elasticity to exports is lower. Box 1.3 figure 1: Shock simulations Real-side response Fiscal reaction and inflation 1,5 2,0 1,0 1,5 0,5 Inflation Debt % deviation from baseline 0,0 1,0 -0,5 0,5 -1,0 % deviation from baseline 0,0 -1,5 -0,5 -2,0 -1,0 -2,5 -1,5 -3,0 GDP Consumption Exports Imports Investment -2,0 -3,5 3 1 3 1 3 1 3 1 3 3 1 3 1 3 1 3 1 3 2016Q 2017Q 2017Q 2018Q 2018Q 2019Q 2019Q 2020Q 2020Q 2016Q 2017Q 2017Q 2018Q 2018Q 2019Q 2019Q 2020Q 2020Q External balance and domestic pressures Elasticity=0.1 Elasticity=0.2 Exchange rate elasticities and exports 1,5 1,4 Elasticity=0.3 Elasticity=0.4 1,0 1,2 Elasticity=0.5 Elasticity=0.6 Elasticity=0.7 Elasticity=0.8 0,5 1,0 Elasticity=0.9 0,0 0,8 % deviation from baseline % deviation from baseline -0,5 0,6 0,4 -1,0 0,2 -1,5 Output gap Current account balance 0,0 -2,0 3 1 3 1 3 1 3 1 3 2016Q4 2017Q2 2017Q4 2018Q2 2018Q4 2019Q2 2019Q4 2016Q 2017Q 2017Q 2018Q 2018Q 2019Q 2019Q 2020Q 2020Q Sources: Hanusch et al. 2016a; Hanusch et al. 2016b; World Bank staff estimates. External Sector high of ZAR 639 billion by December 2015, as investors were surprised by the dismissal The rand recovered from historical weakness of the Finance Minister. The international and the current account deficit narrowed investment position strengthened in rand and Following the end of the commodity super dollar terms. By June 2016, South Africa was cycle, the rand plummeted by 35 percent (in still a net creditor to the world, with ZAR 6.2 trade-weighted inflation-adjusted terms) from trillion held abroad, against ZAR 5.7 trillion in its peak in December 2010, to its low in January liabilities held by nonresidents. 2016, following the commotion around the South African assets held abroad could dismissal of Finance Minister Nene. The be a valuable source for much needed 23 rand has since been recovering. Against the investment at home, should investor sentiment U.S. dollar, the rand registered its strongest continue to improve—including with a view readings in August and November 2016, toward political stability. The latest political at 13.2 rands to the dollar. Several reasons uncertainty shock to the rand occurred in explain the overall strengthening. One reason October, when Finance Minister Gordhan was is the increase in commodity prices for major summoned to court for what many investors South African exports, such as metals, and in perceived as politically motivated corruption The rand particular iron ore (Figure 1.3). charges. Accordingly, within minutes of strengthened over Although still only a third of the high the announcement of the summons by the witnessed in 2011, the increase in commodity National Prosecuting Authority, the rand the course of 2016 prices represents a significant turnaround of depreciated by 3 percent. The charges have but remains weak the falling prices that were still experienced in been lifted since then. 2015. In addition, the U.S. dollar has proven With respect to the current account, there historically weaker than had been expected by many, as is a clear relationship between the depreciation labor market data in the U.S. did not prove of the rand against the currencies of South sufficiently supportive for the Federal Reserve Africa’s major trading partners and the trade Bank to raise interest rates before December balance (Figure 1.13). As the commodity 14, 2016. This situation kept inflation super cycle abated, falling commodity prices differentials favorable for South Africa, and weakened the rand, raising the cost of indeed other emerging markets, for longer imports and resulting in a deteriorating trade than initially expected. Around October 2016, balance. A sustained depreciation typically the rand was also supported by a large forex reduces demand for imports and strengthens transaction, which had been widely anticipated exports due to the change in domestic prices by the markets, when A-B InBev acquired relative to world prices. Indeed, as the rand SABMiller for US$100 billion (~ZAR 1.4 depreciated (in nominal and real terms), the trillion, equivalent to nearly a third of South trade balance became less negative in 2015 and Africa’s GDP and prior to the merger the third improved further in 2016, largely on the back largest company, by market capitalization, of falling import volumes. Exports have been listed on the Johannesburg Stock Exchange), more sluggish in their response, although creating the world’s largest beer company. 5 the spike in 2016 Q2—to a large extent According to the Organisation for driven by the positive production shock to Economic Co-operation and Development manganese and iron ore—resulted in a trade (OECD), the rand-U.S. dollar exchange rate surplus of a magnitude not observed since that would establish purchasing power parity 2011. Thus, although the trade balance has was 5.6 in 2015, a long distance from recent been improving as would be expected from a average spot rates. Although not the only sustained depreciation, exports have yet to pull one, political uncertainty is one important through stronger, especially in sectors where reason for a weak exchange rate, keeping the change in South Africa’s terms of trade foreign capital at bay and providing incentives has enhanced their comparative advantage for locals to hold their assets abroad. South (see Chapter 2). Notably, this observation is Africa’s international investment position not limited to merchandise exports: service strengthened significantly between 2013 exports too have gained competitive position and 2016, swinging from deficit into surplus from the rand’s depreciation, yet the services in the third quarter of 2015, and reaching a balance has yet to improve markedly. SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Overall, South Africa’s current account foreign investors buying South African bonds6 deficit remained wide in 2016, at a seasonally and equities—generating interest receipts and adjusted 5.3 percent of GDP in Q1, 2.9 percent dividend payments that are repatriated. FDI in Q2, and 4.1 percent in Q3. The reason is also results in the repatriation of profits. On mainly the negative income balance, which average, outflows of such income flows exceed has historically been large in South Africa and what South African investors bring back home remains the main reason for the overall deficit by a large margin, resulting in large deficits, on the current account. The current account which were a seasonally adjusted ZAR -128.8 deficit is ultimately a reflection of South Africa’s billion in 2016 Q3. 24 integration in the global financial system, with Figure Main components of the current account and trade weighted exchange rates 1.13 (current account components in seasonally adjusted billion rand; REER and NEER indexes, 2011 Q1 = 100) 100 120 The trade balance improved, but a 100 negative income 0 80 balance kept the current account in -100 60 deficit 40 -200 20 Merchandise Trade Balance Service Balance Income Balance NEER (RHS) REER (RHS) -300 0 2011 Q1 2012 Q1 2013 Q1 2014 Q1 2015 Q1 2016 Q1 Sources: Haver Analytics; World Bank staff. Note: NEER = nominal effective exchange rate; REER = real effective exchange rate. The financial account—financing the inflows, which are less volatile than portfolio current account deficit—continued to be flows, stimulate economic growth, and can supported especially by portfolio investment result in the transfer of critical technology that in the first three quarters of 2016, driven by can help South Africa innovate and create jobs large net outflows in the first quarter, carrying (Chapter 2). over a trend from 2015. South Africa is one Unrecorded transactions, also listed under of the few emerging markets that experienced the financial account, continued to be large, a net outflow in 2016 (Figure 1.14). Net at ZAR 35.5 billion in 2016 Q1 and 15.9 FDI flows turned positive in Q2 and Q3 in billion in Q3, exceeding the total balance on 2016, although inflows fell 28 percent y/y— the financial account by a significant margin. yet outflows fell faster, by 69 percent y/y. These transactions are potentially volatile (and Thus, the improvement in net FDI flows potentially illicit) flows, and a vulnerability generally means that the expansion of South that the authorities are in the process of African firms abroad is slowing—likely as a addressing. Overall, the balance of payments consequence of weak global demand and was in surplus’ in the first three quarters of tougher conditions for South African investors 2016, increasing the country’s international at home—while less FDI is coming to South reserves from US$45.8 billion in December Africa at the same time. It therefore remains 2015 to US$47.8 billion by October 2016. important for policy makers to stimulate FDI Figure Net foreign direct investment to selected countries 1.14  (percent of GDP, first half of 2016) 4 3 2 25 FDI (% of GDP) 1 0 Encouraging -1 foreign investment, especially FDI, would -2 Brazil Mexico India Indonesia Turkey South Africa China Russian Federation support growth Sources: Haver Analytics; World Bank staff. * Only 2016 Q1 data were available at the date of publication. Outlook Private consumption is driving part of this modest recovery, as inflationary pressures The business cycle may have hit its trough and are easing and the credit cycle is expected growth is expected to improve modestly to emerge from the doldrums—potentially The World Bank’s growth estimate for 2016 supported by looser monetary policy as inflation falls back within the SARB’s inflation has been revised down to 0.4 percent (from target band. Little support is expected from 0.8 percent in the last Update); growth for fiscal policy, as the government continues on 2017 and 2018 is projected to be 1.1 and 1.8 its fiscal consolidation path, keeping one eye percent, respectively (Table 1.3). This estimate on the credit rating agencies in 2017 especially. is consistent with the economic momentum Although export growth was somewhat in leading and coincident indicators, which disappointing in 2016, it is expected to pick suggests that South Africa may have reached up steam as the economy—albeit slowly— the trough of the business cycle in 2016 continues making headway in restructuring (Figure 1.15). Risks continue to derive largely and seizing opportunities from the change from global growth, carrying particular weight in the terms of trade: commodity prices and for a small economy like South Africa, with the rand are only expected to recover slowly high dependence on commodity prices. over the next years. As the economy picks up, Domestic risks include natural shocks from imports will rebound too, with consumers climate change (such as the 2015 drought) increasing their appetite for foreign products but also domestic shocks, most notably on the and firms importing intermediate and capital political front. goods to take advantage of new opportunities. SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Table Baseline annual growth forecast 1.3 (percent) Indicator   2013 2014 2015 2016 e 2017 f 2018 f Real GDP growth, at constant market prices 2.3 1.6 1.3 0.4 1.1 1.8   Private consumption 2.0 0.7 1.7 1.3 2.1 2.1   Government consumption 3.8 1.8 0.2 1.6 0.8 1.2   Gross fixed capital investment 7.0 1.5 2.5 -5.9 -2.8 -1.6   Exports, goods and services 3.6 3.3 4.1 1.0 3.7 4.1 26   Imports, goods and services 5.0 -0.5 5.3 -3.6 2.0 2.1 Prices: inflation 5.4 6.4 4.6 6.3 6.0 5.8 Current account balance (percent of GDP) -5.9 -5.3 -4.3 -3.9 -3.4 -3.5 Fiscal balance (percent of GDP)a -3.7 -3.6 -3.7 -3.5 -3.2 -2.8 Gross Debt (percent of GDP)a 45.9 46.8 49.4 51.4 52.5 53.0 Primary balance (percent of GDP)a -0.7 -0.4 -0.4 0.0 0.5 1.0 Poverty rate ($1.9/day 2011 PPP terms)b,c,d 15.7 15.6 15.7 15.9 16.0 15.9 Growth will still be Poverty rate ($3.1/day 2011 PPP terms) b,c,d 33.6 33.6 33.6 34.1 34.1 33.9 Sources: World Bank, Macroeconomics and Fiscal Management Global Practice, and Poverty Global Practice. insufficient to make Note: e = estimate; f = forecast; GDP = gross domestic product; PPP = purchasing power parity. a. Fiscal year.           b. Calculations based on 2010-Income Expenditure Survey.             a marked dent to Projection using neutral distribution (2010)with pass-through = 0.87 based on GDP per capita in constant PPP. c.  d. Projections are from 2013 to 2018.             poverty Fixed investment is expected to continue implementing partners, improving the falling, although at a decelerating pace. Fixed execution of key public investment projects, investment is one area where policy can strengthening cities as South African prove the analysts wrong and decisively turn powerhouses—in spite of the change in around the South African economy. Providing government—and strengthening political an environment conducive for domestic certainty are all areas that can raise gross fixed and foreign investors, by accelerating the investment. This will be crucial to raise the implementation of the National Development growth potential of the South African economy Plan with strong coordination across the and generate sustainable jobs. Figure South African Reserve Bank business cycle indicators 1.15  (2010 = 100) 106 120 104 115 102 100 110 98 105 96 94 100 92 Leading indicator Lagging indicator Coincident indicator (RHS) 90 95 Jan-11 Jul-11 Jan-12 Jul-12 Jan-13 Jul-13 Jan-14 Jul-14 Jan-15 Jul-15 Jan-16 Source: South African Reserve Bank. Given the growth projections, little for policy to address South Africa’s growth progress is expected in reducing poverty and constraints decisively, especially as relatively inequality. Since the majority of the extreme little support can be expected from the poor depend on social grants for their global economy. This strategy will involve income, the growth of these transfers has the transformative interventions, including largest effect on the pace of extreme poverty removing bottlenecks in infrastructure, and alleviation. Employing micro-simulations that improving education and industrial policy model the impact of growth on household (Chapter 2). Yet it is important to bear in consumption, poverty is expected to remain mind the power of localized and smaller 27 roughly constant between 2016 and 2018. Inequality, by contrast, is expected to increase interventions, which can make a large impact by 1.3 percent between 2010/11 and 2017/18, on the ground for poor South Africans. An largely due to the impact of the widening gap example of such initiatives, piloting farming between those with and without jobs. This on rooftops in central Johannesburg to revive situation makes it all the more important urban farming, is featured in Box 1.4. Innovative local Box Ideas in poverty reduction: Farming from rooftops in Johannesburg’s central 1.4 1.4b business district initiatives continue to play an important Cathy Nkambule, age 56, is one of four owners of a flourishing rooftop vegetable garden atop a skyscraper in the inner city of Johannesburg. The garden has 26 tunnels covering bunches of spinach that are harvested daily and sold to role, improving the a market within walking distance. This is the site of an urban farm run by one of the many cooperatives supported by the Johannesburg Municipal Government as part of an innovative urban agriculture program targeting households living lives of the poor in extreme poverty. “This (garden) has made a huge difference in my life. I never imagined that I would have a steady income let alone my own business. I now have a reason to wake up and to work hard at my own thing. I even have surplus to help others who are less fortunate,” she says. Ms. Nkambule is one of beneficiaries of the Johannesburg Municipal Government’s Food Resilience Program, which is aimed at tackling extreme poverty through urban agriculture. The program seeks to reduce the number of food insecure families by 20 percent, ensure that no one goes to bed hungry, create sustainable markets for subsistence and emerging farmers, and create decent jobs and disposable income. On the one hand, the program aims to reduce food insecurity by enabling beneficiaries to eat what they produce; on the other hand, the program promotes the emergence of farmers who are able to generate income and create jobs. The program was developed in response to a 2011 study by the University of Johannesburg, which showed that 42 percent of the city’s 4.5 million people go without at least one meal between three and 10 days in a month, with food accounting for up to 60 percent of the expenditures of the poor. The program’s objective is to wean beneficiaries from extreme poverty, to a level where they attain food security, create jobs, and generate their own income through a three-stage, staggered approach. Initially, the indigent families are given food parcels. At this stage, they receive training to enable them to transition to the next stage, in which they establish homestead, communal, or rooftop food gardens for which the municipality provides infrastructure support in the form of land, farming implements, electricity, water, seeds, training, and links to markets. The most successful farmers graduate into any of the municipality’s four Food Empowerment Zones (FEZs), which constitutes the third stage. FEZs leverage the municipal-owned large commercial farms and create opportunities for intensive farming by these emerging farmers. Eikenhoff, which is situated in the southern part of Johannesburg, is the biggest FEZ, spanning 270 hectares. It has cost a total investment of around ZAR 12 million since 2012. It benefits at least 160 families directly who are engaged in planting vegetables and livestock farming, including pig raising and poultry production. The other three FEZs have a collective 54 hectares and benefit at least 59 families, costing the municipality a total of about ZAR 10.8 million. Food gardens and FEZs are supported by seven agri-resource centers and satellites across the city, which provide advisory services, including registration of cooperatives. Ultimately, the goal of the Food Resilience Program is to enable beneficiaries to become self-sustainable emerging farmers, with government support expected to reduce as the beneficiaries graduate through the three stages. The program is important in a context where urban farming historically has been repressed, particularly in the townships, and in an environment in which the population of commercial farmers is aging, posing a threat to national food security. In 2010, approximately 3.5 percent of South African urban households were involved in some form of urban agriculture. This innovative program leverages the strengths of multiple partners, including the provincial government, nongovernmental organizations, universities, and the private sector. In Alexandra, one of the city’s poorest communities east of Johannesburg, a communal food garden owned by a group of seven cooperatives of local women has partnered with a commercial farmer. The private sector partner serves as an incubator, helping the group to navigate the agriculture produce value chain from planting to branding, packaging, and marketing the produce. The Alexandra project also sources produce from the Johannesburg fresh produce market to resell to hawkers in the community, to create a steady supply that would help stabilize prices. The aim is to create a fresh produce market within Alexandra through the development of a network SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Box Ideas in poverty reduction: Farming from rooftops in Johannesburg’s central 1.4 1.4b business district of micro farms, which would reduce transport costs to markets. Overall the Food Resilience Program has reduced poverty and food insecurity through creating urban subsistence farmers who are able to contribute to sustained food production at the household level. In addition, some of the farmers have become entrepreneurs who create jobs and generate income. To date, 107,466 indigent households have benefited directly from the program. At least 240 emerging farmers have been linked to markets. More than a 1,000 people have been trained in farming techniques. The program’s ability to attract multiple stakeholders will be key to its sustainability. 28 CHAPTER 2 Private Investment for Job Creation Introduction Africa’s commodity-driven growth model, the authorities have accelerated efforts to Industrial policy is Job creation has been among the most promote industrial development in the past fundamental objectives of the South African decade. Such a policy orientation builds on the a key instrument authorities since democracy. Labor is the expectation that industrial development could to meet South main income source of most South African potentially generate several positive outcomes, citizens, and raising labor remuneration for including higher wages for workers (resulting Africa’s job creation the poorest segments of society is the most from labor productivity due to technology objectives, through effective instrument to reduce poverty and investments); larger employment and growth inequality. Per World Bank staff calculations, multipliers through forward and backward the promotion of each job created in South Africa lifts about one linkages with other sectors; increased domestic private investment person out of poverty (Box 2.1). competition through greater exposure to In 2011, the National Development Plan highly competitive world markets; increased in strategic sectors (NDP) foresaw the need to create 11 million reliance on the larger world demand, as the new jobs between 2012 and 2030, to bring potential for stimulating higher domestic down the unemployment rate to 6 percent and demand through debt financing is being reap the potential economic benefits provided progressively exhausted for households and by the ongoing demographic transition.7 This government; and more stable growth, as goal means creating about 600,000 new jobs underpinned by stable global demand for every year. manufactured goods and a less volatile capital Nonetheless, the pace of job creation in account. Through appropriate investments, the past decade has been too slow to meet industrial development can seek to expand this target, and there is growing evidence production in sectors where comparative that it lost steam in recent years with the end advantage already exists, or alternatively of the commodity super cycle. Between 2005 develop new comparative advantages through and 2015, the private sector created about technological upgrading. 265,000 jobs on average every year (mostly South Africa’s industrial policy in the service sector), and the public sector acknowledges the central role that private created about 50,000 jobs. These numbers are investment should play to promote industrial clearly insufficient to meet NDP’s objectives, development, and the fact that private sector and decelerating growth since 2014 combined investment decisions ultimately drive labor with tighter fiscal space (Chapter 1) further demand and thus job creation.8 The policy thus complicates things. Since democracy, most foresees the government’s role as regulating job creation took place during the commodity private investment decisions (through super cycle, and its recent termination calls for industrial policy instruments of various nature), alternative sources of growth. while providing public goods (infrastructure, Cognizant of the shortcomings of South such as power generation in recent years) and 29 SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS correcting market and coordination failures of productive capacities: only the (special economic zones for instance) that service sector combined capital the private sector cannot handle. Beyond deepening with job creation. support to private investment in key sectors • Such trends in capital allocation across for job creation, the national authorities also sectors have not only been bad for job see the need to boost the acquisition and creation. They have been equally bad provision of the skills needed by the private for GDP growth, generating significant sector (see Box 1.1 on South Africa’s efforts losses in aggregate capital productivity. to reform the education sector); to raise youth Indeed, since 2008, there has been a 30 exposure to jobs and professional experience, significant deterioration in the South to facilitate their insertion into labor markets African economy’s capacity to direct (through employment tax incentives (ETIs) private investment toward sectors with for instance); to combat racial discrimination growing economic potential—the through affirmative action (through the manufacturing sector in particular. Broad-Based Black Economic Empowerment • Although the delayed reactions of program notably); and to facilitate the concentrated industries to changing The contraction of matching of labor supply and demand through opportunities, and the long time industrial sectors better spatial integration in coordination with needed by large infrastructure local authorities. projects to start generating returns since democracy However, the authorities’ efforts to promote may explain this negative trend, the calls for an industrial development in recent years were analysis of the current ITI framework not matched with a significant reallocation of suggests that it may have also strongly assessment of private capital toward industrial sectors, or with contributed to such misallocation of the effectiveness higher industrial employment. Such a contrast capital. Compared with the industrial between policy ambitions and outcomes calls sector, lower marginal tax rates for the and efficiency of for an assessment of the effectiveness and mining and construction sectors make efficiency of industrial policies on private private investment in these sectors industrial policies investment and job creation, and better equally remunerative despite much identification of other factors that may have lower growth and job creation returns generated such disappointing outcomes below for the economy at large. NDP targets. • Notwithstanding, by reducing firms’ This Chapter discusses the role played tax burden compared with general by private investment for job creation provisions, ITIs have encouraged since democracy. The Chapter highlights additional investment in agriculture, recent trends in job creation and the role construction, manufacturing, trade, played by private investment; explores the and other services. Overall, additional determinants of private investment allocation investment generated by tax incentives across economic sectors; and assesses the exceeds the government’s foregone effectiveness, cost, and impact of investment revenue from distributing tax tax incentives (ITIs) granted to the various incentives. Furthermore, the existence economic sectors on additional investment of large employment multipliers and job creation. brings the fiscal cost of job creation to A few important messages emerge from a small fraction of total labor costs, in the discussion: the manufacturing sector in particular. • Job creation since 1994 has taken And ITIs compare favorably with ETIs, place almost exclusively in the service from the fiscal and quality of jobs sector, at a pace much too slow to created perspectives. curb unemployment and poverty • ITIs have thus contained job significantly. destruction in the industrial sector, • Investments have resulted in replacing and explanations for industrial jobs with machinery in the agriculture, contraction since democracy and more mining, and manufacturing sectors, as recently must be found elsewhere, technological upgrading has not been possibly among insufficient skills and accompanied by a sufficient expansion infrastructure, policy uncertainty, a volatile rand, and complicated labor medium term from the new business relations. opportunities from the depreciation Moving • forward, reorienting of the rand, decline of commodity incentives toward the industrial sector prices, and the coming online of large would create additional jobs at no additional power generation capacity. additional fiscal cost, the more so as the industrial sector shall benefit in the Box Job creation and poverty reduction 31 2.1 Relying on the Income and Expenditure Survey conducted in 2010, the World Bank estimates that the proportion of the South African population living below the international poverty lines of purchasing power parity (PPP) $1.9 and PPP$3.1 stood at 15.7 and 33.6 percent, respectively, in 2015 (World Bank 2016d). Admittedly, job creation is expected to be among the main factors to reduce poverty, by potentially boosting the incomes of the large group of unemployed poor. Using statistical matching techniques applied to the National Income Dynamics Study data for 2015, we estimate the impacts of job creation in different economic sectors on poverty. The method identifies unemployed individuals those Job creation strongly whose characteristics (education, age, gender, and so forth) are closest to those of individuals employed in a given sector, and assumes that these individuals would be the direct beneficiaries of job creation in the said sector, having the greatest reduces poverty likelihood of becoming employed. In turn, we estimate the net income impact for these individuals of getting a job in the said sector, estimating the wage they would now receive (using Mincer regressions), deducting unemployment benefits, and estimating their new eligibility for children’s grants, given their new labor income. Box 2.1 figure 1 reports the number of people (including members of the households from which the unemployed would become employed) that would be lifted out of poverty as a result of creating one job in a given sector. Creating one additional job in mining or agriculture lifts about 1.3 people from poverty. Increasing employment in construction and manufacturing also has a significant impact—the effect in these sectors is almost one to one. Employment in trade, financial services, and community services has a smaller impact—only seven in 10 people employed are elevated from poverty. Employment in financial intermediation is already geared toward a relatively better off and educated population; therefore, the impact of employment in this sector on poverty is lower than in other sectors. In some sectors, such as private household employees, the impact on poverty is relatively small, as the wages in those sectors are low and the impact of the loss of transfers on poverty is significant. Box 2.1 figure 1: Changes in poverty due to job creation (number of individuals) 0.4 0.2 0.0 Change in poverty (number of individuals) -0.2 -0.4 -0.6 -0.8 -1.0 -1.2 -1.4 -1.6 es re t n es de g ity ng r e io rin po vic ltu nc vic a ini ric ct Tr tu ns a icu r r tru M t Fin se se ac ec a r Tr ns er uf Ag El ld Co an o th eh O M us Ho employment effect grants/UIF effect Total Source: World Bank staff calculations. Note: UIF = Unemployment Insurance Fund. SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Job Creation and Investment since an additional 470,000 jobs.9 From 1994 to 1994 2004, net job creation in the private sector was cumulatively almost nil, as periods of Net job creation has accelerated significantly net job creation were offset by periods of net since 2006, but has been insufficient to job destruction. From 2005, the pace of job accommodate the growing labor supply creation steadily accelerated, and was only Between 1993 and 2015, South Africa’s reversed in 2009–10 as the economy adjusted private sector created 2.65 million formal to the consequences of the global financial 32 and informal jobs in net terms, see Figure crisis. Since 2010, 1.76 million jobs were 2.1. Meanwhile, government services created created in the private sector (and another 0.2 Figure Cumulative private sector job creation, 1994–2015 2.1 (million jobs) 3 Job creation accelerated since 2.5 2004 but was 2 insufficient to 1.5 absorb the growing Jobs (millions) pool of people 1 seeking work 0.5 0 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 -0.5 -1 Source: World Bank staff calculations based on Quantec 2016. million in government services). Between 2011 and 2015, an average of Although job creation accelerated in the 589,000 workers entered the labor force every past decade, it was nonetheless insufficient year. Of these, on average, only 424,000 (72 to absorb the growing pool of people seeking percent) found employment, while 165,000 work. Lack of comparability in labor surveys (28 percent) became unemployed (narrow over time prevents strict monitoring and definition). Meanwhile, the number of reporting of unemployment rates since 1994.10 discouraged workers grew by 20,000 every But the fact that by 2015 the unemployment year. Of the total employed, 22.1 percent were rate (narrow definition) stood at 24.5 percent skilled, 47.5 percent semi-skilled, and 30.4 of the active population suggests that job percent low-skilled; these proportions have creation in the preceding years was largely remained relatively stable since 2008. Of the insufficient to absorb the growing labor supply. total unemployed, 4.2 percent were skilled, The problem is even larger when discouraged 23.2 percent semi-skilled, and 17.3 percent workers are considered as well, bringing the low-skilled, and for 55.3 percent the skill set unemployment rate (expanded definition) to is unknown, but probably the vast majority are 33.8 percent of the active population by the among the unskilled. end 2015.11 Jobs were primarily created in services, created altogether about 4.3 million formal and mainly lost in agriculture, mining, and and informal jobs over 1994–2015. As manufacturing further discussed in this Chapter, while not Since democracy, the agriculture, mining, entailing the same multiplier magnitudes (for and manufacturing sectors have lost jobs. employment and GDP through backward and This observation holds for the entire period forward linkages), the service sector cannot be from 1994 to 2015, and since 2008; it is also seen as just “consumption” sectors benefiting valid for subsectors, as reflected in Figure from investment and productivity growth 2.2, which reports extrema of job creation/ from tradable sectors. For example, while 8.6 33 destruction per subsectors.12 In contrast, percent of South Africa’s nonmineral output except for construction over 1994–2008, was exported in 2012, private services alone all service sectors created jobs over the two exported 3.9 percent of output. Likewise, periods, and actually created almost all private services represented 20 percent of jobs since 1994. Trade/accommodation, nonmineral exports in 2012, and business finance/business, and private social services services alone 8 percent. Since democracy Figure Job creation by sector, 1994–2015 2.2 almost all job (million jobs) creation took 1.5 place in services. 1.25 Jobs created 1993-2015 Agriculture, mining Jobs created 2009-2015 1 and manufacturing Maximum jobs created within sector 0.75 Minimum jobs created within sector shed jobs Jobs (millions) 0.5 Construction 0.25 Manufacturing 0 Electricity Other Trade Transport Finance services -0.25 Mining -0.5 Agriculture -0.75 Source: World Bank staff calculations based on Quantec 2016. Over time, the South African economy has in aggregate capital intensity was partially become more capital intensive offset through a composition effect, with The number of private sector jobs grew faster investment in labor intensive sectors. on average by 0.9 percent annually between Between 1994 and 2015, the share of other 1994 and 2015, and the private capital services’ capital in total capital grew from 1 stock (summing up past investment minus to 4 percent (Figure 2.6), and this increase depreciation; source: Quantec 2016) grew alone explains most of the composition effect. by 2.2 percent over the same period. Thus, Other services—comprising medical services, the economy became more capital intensive, nonmedical services, and other producer and the aggregate capital-labor ratio rose by subsectors—accounted for 18 percent of 25 percent (Figure 2.3). Nonetheless, while private sector jobs in 2015. Other producers, the vast majority (35 of the 46) of sectors estimated to be mostly constituted of informal became more capital intensive,13 the increase jobs, accounted for 15 percent of private SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS sector jobs in 2015, and was the second least and Rodrik 2011). The pattern underlines remunerative for workers of all sectors after the slow structural transformation of the agriculture. This pattern of overall capital South African economy, characterized by the deepening combined with job reallocation increased concentration of workers in low- to services in not uncommon, as observed in productivity jobs. most African countries since 1990 (McMillan Figure Evolution of the capital-labor ratio within and between sectors, 1994–2015 34 2.3 (cumulative growth, %) 50 40 30 Production became Cumulative growth (%) more capital- 20 intensive over time 10 0 -10 -20 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 Between Within Aggregate Source: World Bank staff calculations based on Quantec 2016. The impact of investment on job creation intermediation, insurance, real estate, in the various sector clusters is shown in and business services (business services in Figure 2.4. The blue bars report the number particular) clusters combined increased of jobs that would have been created or lost investment with a declining capital-labor in each cluster, given the observed investment ratio, thereby leading to massive job creation. levels, had the sector’s technology (capital- All the other clusters recorded increased labor ratio) stayed unchanged since 1993. In capital intensity. In the cases of other services; contrast, the orange bars report the variation electricity, gas, and water; and transport, in jobs resulting from a change in technology, storage, and communications, the overall keeping capital stock unchanged at the 1993 increase in capital stock was large enough level. The red dots in the Figure 2.4 indicate to entail net job creation. In the cases of the number of jobs created or lost over the agriculture, fishing, and forestry; mining and period, and are equal to the sum of the blue quarrying; manufacturing; and construction, and orange bars. increased capital intensity was not matched As shown in Figure 2.4, the trade, catering, by sufficient expansion in the overall capacity and accommodation services (wholesale and to create jobs. In these sectors, capital retail trade in particular) and the financial progressively replaced jobs. Figure Contribution of expansionary and technology investment to job creation, 1994–2015 2.4 (million jobs) 10 8 6 4 35 Jobs (millions) 2 0 Electricity Trade Finance Agriculture Mining Manufacturing Transport -2 Construction Capital progressively -4 replaced jobs in -6 agriculture, mining, Other services -8 Expansionary investment Technological investment Net jobs created manufacturing and construction Source: World Bank staff calculations based on Quantec 2016. Allocation of Private Capital across new capital is likely to be more mobile across Sectors sectors than already installed capital,14 it would be expected that periods of high investment The determinants of private investment levels (and corresponding rapid growth in allocation across sectors need to be capital stocks) would coincide with more understood to influence them through rapid capital reallocation across sectors. industrial policy. As further discussed in Thus, the macroeconomic and doing business this Chapter, the South African authorities environment may impact the speed at which consider that private investment allocation capital is reallocated across sectors to match across sectors should be influenced by a set of changes in relative profitability. industrial policy instruments—tax incentives, notably—with a view to correct market failures South Africa has recorded significant and generate positive externalities (especially restructuring since 2008, as evidenced by the through sustainable job creation). Through sustained reallocation of private capital across their impact on profitability, tax incentives are sectors expected to make private investment more It is not clear, however, that general attractive in the sectors to be favored. macroeconomic conditions have strongly The effective allocation of private capital influenced the pace of capital reallocation across sectors is theoretically determined by across sectors in South Africa since democracy. country- and sector-specific considerations. Figure 2.5 plots the annual growth in total Country-specific considerations include capital stock against the pace of reallocation the macroeconomic and doing business of capital across sectors in the same year.15 environment, which influences overall Three distinct periods can be observed: a first financing costs in all sectors, expected period (1994–2002) during which capital is demand for goods and services produced reallocated across sectors despite slow capital in the country, relative prices vis-à-vis the accumulation; a second period (2002–08) of rest of the world, provision of public goods, rapid capital accumulation contrasted with and the degree of stability needed to make lower capital reallocation; and a third period informed investment decisions. And because (2009–15) during which capital reallocation SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS records unprecedented levels since democracy resulted in clear shifts in capital allocation despite slower capital accumulation than (for example, portfolio investments16 versus before the global financial crisis. The factors real economy investments), as well as by the that influenced capital accumulation and commodity super cycle, which also led to reallocation during the first period include sector shifts. Influential factors during the dramatic changes in trade policy (tariff third period include the greater involvement protection, international and regional of public utilities in infrastructure investment commitments, and so forth). Influential (energy and transport in particular; see Figure factors during the second period include 2.6).17 36 the “financialization” of the economy, which Figure Capital stock growth and reallocation across sectors, 1994–2015 2.5 (percent) 5.0 The restructuring 4.5 of the South African 4.0 economy accelerated 3.5 since 2008 3.0 2.5 2.0 1.5 1.0 0.5 0.0 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 Sum of absolute changes in sectors’ share of capital Change in total capital Source: World Bank staff calculations based on Quantec 2016. In the past decade, investments moved away services), by far the largest destination of from the manufacturing sector private investment, is more difficult to analyze From a sector perspective, the accelerated from a productive capacity and employment reallocation of capital across sectors perspective, as reflected in the very deep and observed since 2009 is reflected in the diverse intermediation activities of South significant relative decline in investment in Africa’s financial markets. South Africa’s manufacturing, contrasting with a symmetric markets finance gross fixed capital formation increase in investment in electricity and water. in South Africa and elsewhere, as well as private Over the longer 1994–2015 period, transport and other services saw their shares in total and public consumption. Although the share capital stock grow steadily, while the opposite of the financial sector has declined steadily can be observed for agriculture. Finance since 1994, the share of business services grew (including the financial sector and business until 2008 before dropping sharply. Figure Sector shares in total capital stock, 1994–2015 2.6 (shares) 18 44 16 43 14 42 All sectors except Finance (%) 12 41 37 10 Finance (%) 40 8 39 6 38 4 2 37 Since 2008, 0 36 investments in 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 20 9 10 11 12 13 14 15 0 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 manufacturing Agriculture Mining Manufacturing Electricity Construction Trade Transport Other services Finances decelerated, and accelerated in the Source: World Bank staff calculations based on Quantec 2016. electricity sector Economic gains from capital reallocation have can be decomposed between (i) the sum of slowed since 2008; declining investment in the sectoral capital productivity growth and (ii) manufacturing sector largely contributed to the reallocation of capital toward the sectors this poor outcome that exhibit the highest capital productivity The impact of tax incentives depends to a growth. This computation suggests that large extent on the ability of financial markets capital reallocation gains were significant to move capital to the sectors where it is most in 1994–2008, generating on average 0.2 profitable. Such allocative efficiency can percentage point of annual GDP growth.20 be quantified in different ways. First, under However, such gains decelerated starting in the assumption of diminishing returns to 2008, and turned negative in 2012, suggesting capital accumulation, the allocative efficiency reversed reallocation of capital toward sectors of financial markets can be measured by with lower productivity growth. Indeed, the dispersion in sector rates of return. A sectors such as mining, electricity, transport, significant difference in the rates of return finance, and other services saw their share in should prompt capital to be reallocated total capital grow, while experiencing at the toward the most profitable sectors until the same time a decline in capital productivity. marginal rates of return are equalized across Agriculture, manufacturing, construction, and sectors. Figure 2.7 plots the dispersion of trade experienced inverse trends. Declining average sector rates of return over 1994– investment in the manufacturing sector largely 2015.18 Although dispersion decreased overall contributed to these overall losses due to until 2008, reflecting improvements in capital reallocation. allocation, after 2008, it strongly deteriorated. The evolution in the allocative efficiency of the capital market since 1994 is confirmed by a second metric, which measures the contribution of capital reallocation to aggregate capital productivity growth. Following Syrquin (1986), aggregate capital productivity growth19 SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Figure Economic gains of capital reallocation across sectors, 1994–2015 2.7 (% and index) 0.8 2.25 0.6 2.2 2.15 0.4 38 2.1 0.2 2.05 0.0 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2 -0.2 Since 2008, the 1.95 economy recorded -0.4 1.9 losses in capital -0.6 1.85 allocation and Reallocation effect (%) Returns dispersion (RHS, index) increased dispersion Source: World Bank staff calculations based on Quantec 2016. in profit rates Firm-level data analysis over 2006–12 suggests the same time (such as macroeconomic and that investment response to changes in sectoral shocks). Thus, over the long run, profitability differs widely across sectors capital demand k from firm i in period t (as With a view to measure the impact of tax measured by the value of the firm’s assets)21 incentives on investment, World Bank (2016f) would depend negatively on the relative cost econometrically estimates the response of of capital with respect to the output price (the investment to changes in the return to capital at user cost of capital, ucc), positively on output the firm level. The analysis relies on tax-related (measured by sales, s), and time and firm and accounting data (source: South African specific effects, as expressed in the equation Revenue Service (SARS)) for all corporations below. The detailed computation of the user that filed tax returns for 2006–12, for a total cost of capital, which combines the price of of more than one million observations from investment goods (including financial costs) more than 250,000 firms (among which and taxes (including tax incentives), is further 63,000 firms have observations for all seven discussed in the next section. years). The specification that was empirically tested assumes that firms are cost minimizers kit= α- σuccit+ μsit+ αt+αi+ εit and output price takers, and therefore that demand for factors (including capital) Regressions are run for different subsets and output are simultaneously determined of the entire data set: large and small business through the equalization of factors’ marginal corporations, together and by sector of activity, returns with costs. The specification also and for the period 2006–12. Table 2.1 reports assumes that firms progressively reach their the estimated long-term responses of demand optimal demand for factors, through an error for capital to changes in the user cost of capital correction model. Finally, firm-specific and (UCC), expressed as elasticities. For instance, time-specific fixed effects are introduced a 10 percent increase in the user UCC for large to control for differences in productivity business corporations is associated with a 2.8 levels and time shocks affecting all firms at percent decline in the demand for capital. Table Long-term capital demand response to a change in the relative cost of capital, 2.1 2006–12 Large business corporations Small business corporations Total -0.28 ** Total -0.43 ** Agriculture -0.30 ** Agriculture -0.30 ** Mining -0.36 Mining -0.15 Manufacturing -0.30 ** Manufacturing -0.21 Electricity -0.09 Electricity -0.43 Construction -0.49 ** Electricity -0.64 ** 39 Trade -0.36 ** Trade -0.16 Transport 0.13 Transport -0.24 Financea -0.36 Financea -0.63 ** Other services -0.28 ** Other services -0.48 ** Source: World Bank 2016f. ** indicates statistical significance at the 95% confidence level. a. Includes only real estate and business services, excluding financial and insurance services. Investment response to change in profitability differs The results suggest that, overall, large responsive to the average sector output price. and small business corporations’ demand Market segmentation (for instance, spatial across sectors for capital responds positively to profitability segmentation) for small firms operating (or equivalently, negatively to increased in small markets could also be a candidate UCC). Nonetheless, this overall response for explaining such differences. Structural conceals large differences across sectors. differences in technology among firms in Among large business corporations, the the same sector could also explain the large responses of mining, utilities, transport, and variance in the estimated responses, as these de real estate cannot be considered statistically facto measure the technological substitution different from zero. Among small business elasticities between the various factors of corporations, the responses of the mining, production. Firms adopting rigid technological manufacturing, utilities, trade, and transport choices through long-term investment sectors to changes in the cost of capital cannot decisions (say, firms in mining, transport, and be considered statistically significant. This utilities) may not be reactive to short-term does not necessarily mean that corporations variations in the cost of capital. Finally, the long in these sectors do not respond to profit time needed by large infrastructure projects to considerations, but rather that there exist too start generating returns may also explain this large differences between corporations within negative trend. The energy sector is a case in said sectors to consider that a common pattern point: investments regularly cumulated since can be established. 2008 (Figure 2.6) only resulted in a significant Several explanations can be tentatively increase in power generation capacity since advanced to explain such differences across 2015.23 sectors. Among large business corporations, These differences may in turn contribute the presence of dominant noncompetitive to explaining short-term variations in capital positions (the result of regulatory barriers to allocative efficiency,24 as well as the impact of entry and/or economies of scale)22 could lead tax incentives on investment, as discussed in to firms that have market power to influence the next section. output prices, and therefore the firms are not SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Costs and Potential Benefits of resources in support of these objectives, and Investment Tax Incentives the existence of positive externalities for the society from higher employment and balanced Since 2007, South Africa has accelerated policy spatial development, which markets cannot, efforts to promote industrial development by definition, internalize in their decision through various policy instruments making. Building on past failures, the NPIF In the past decade, South Africa’s policy acknowledges the need to eschew picking efforts to promote industrial development winners to promote horizontal policies that accelerated, mirroring the resurgence build on potential comparative advantages.26 40 of industrial policies observed in many From a practical perspective, several middle-income countries, and driven in industrial policy tools have been deployed in part by the successful experience of East South Africa in support of these objectives Asia’s manufacturing-driven structural and principles, at the national and local levels. transformation.25 The National Industrial (Box 2.2 provides a discussion of the spatial Policy Framework (NIPF), which was adopted dimension of South Africa’s new industrial in 2007, identifies microeconomic constraints policy.) Such tools have taken the form of tax Since 2007, to growth and employment, and provides incentives to various economic sectors, public policy efforts the overarching coordinating framework for procurement differentiation, provision of the development of 13 strategic programs in infrastructure in support of specific sectors, accelerated to the following domains (PBO 2016): sector lending to and equity participation in strategic strategies, industrial financing, trade policy, sectors (through the Industrial Development promote industrial skills, competition, public procurement, Corporation (IDC)), competition policy (to development, upgrading, innovation, infrastructure, reduce induced markups and input costs small enterprises, and empowerment. The for strategic sectors), and softer “persuasive through the framework has since been implemented initiatives to secure greater private sector deployment of through a series of annual Industrial Policy support for local manufacturing” (DTI 2016). Action Plans. Although it is difficult to assess comprehensively tax incentives in The objectives of the NIPF comprise (given the multiplicity of interventions and particular the diversification of the economy away the absence of a counterfactual), PBO (2016) from excessive reliance on traditional reports that ZAR 476.5 billion (at 2015 commodities and non-tradable services, constant prices) was cumulatively devoted to industrialization and movement toward a national industrial development programs knowledge economy, higher employment and initiatives between 1994 and 2014 creation through labor-intensive industrial (excluding IDC financing27 and the impacts development, greater economic participation of procurement and competition policies). of historically disadvantaged people From this total, 71 percent took the form of and marginalized regions, and regional tax incentives (“tax expenditures”), and the integration and industrial development remainder was devoted to the provision of of the African continent. Policy-induced public services in support of industrial sectors, reallocation of factors (private investment spatial development, export promotion, in particular) is justified on the grounds of small business, transformation, and skills and market and coordination failures to allocate development. Box Promises and challenges of spatialized industrial policies 2.2 South Africa’s industrial policies have a strong spatial dimension. First, they retain the objectives of balanced spatial development, and the greater participation of marginalized regions, especially in the mainstream of the industrial economy. Second, the provinces and cities have a direct mandate and responsibility in the implementation of various aspects of such policies, most notably the special economic zones (SEZs). The SEZ Act was gazetted in 2014. Under the framework of the act, the South African authorities envisage the development of 15 SEZs across the country, supported by various tax incentives (reduced corporate rates, capital allowances, and employment subsidies) and direct infrastructure financing from the central government (through the dedicated SEZ fund). (As of November 2016, seven locations had been designated.) International evidence suggests that SEZs can be responsible Box Promises and challenges of spatialized industrial policies (continued) 2.2 for substantial job creation if they can generate employment spillovers (indirect job creation outside the SEZ) through backward supply chain linkages that arise from the newly established investments. In line with international evidence, projections for South Africa’s SEZs suggest that indirect jobs vastly outnumber the direct jobs created in the zones, see Box 2.2 figure 1. Indirect jobs include construction jobs, but over the medium term, supply chain linkages will be the critical channel if SEZs are to deliver on their promise as large-scale job creators. Box 2.2 figure 1: Jobs created by special economic zones 41 (number of jobs) 45000 40000 35000 30000 The ability of Special 25000 Economic Zones to 20000 create job depends 15000 on their integration 10000 in the local 5000 economic fabric Richards Bay East London COEGA (2015) Dube Tradeport Saldanha Atlantis Industrial Industrial (2020 target) (Steady State) (Steady State) Development Zone Development Zone (2015) (2015) Direct jobs Indirect and construction jobs Sources: World Bank staff calculations from DTI SEZ Bulletin July 2015; various IDZ/SEZ reports. Note: IDZ = industrial development zone; SEZ = special economic zone. Making this happen thus requires strong policy coordination provision between the government, provinces, and cities, with a view to maximize the integration potential of SEZs in local economies. Nowadays, this level of coordination constitutes a challenge in South Africa, in the absence of clear guidelines and overlapping responsibilities in the provision of complementary infrastructure and services to the SEZs, licensing and regulation of investment in the SEZs, and sharing of financial risks across public partners related to SEZ management.a Neither the SEZ Act nor its regulations make any reference to cities, urban areas, or metropolitan areas, and as such there exist no provisions that support or preclude a formal relationship between the SEZs and the municipalities where they are located. Thus, the lack of clear guidelines is reflected in the different fortunes of coordination observed in different zones. For example, development of the SEZ at Dube Tradeport was initially hampered by the project’s location being beyond the “edge” defined in the eThekwini municipality’s spatial development plan. More recently, examples of successful planning collaboration in Cape Town and Ekurhuleni highlight the importance of coordinating efforts to facilitate access to land (in alignment with province and city spatial development plans), streamlining regulatory approval processes, improving transport access for workers, and supporting training and skills development. Coordination also needs to address explicitly the burden sharing of the financial risks and contingent liabilities that may eventually arise and fall disproportionally on local governments in the long run. These risks and responsibilities stem from local government responsibility, which often assigns license holders to manage the SEZs (which may become financially unsustainable for various reasons), and from the capital investments linked to SEZs that municipalities will need to sustain in the medium term after initial funding from the national government has disappeared.28 Locating SEZs with a view to foster balanced development could also undermine the possibility to meet other industrial development objectives, such as job creation in particular. Indeed, locating SEZs in metropolitan areas, where most growth and job creation will likely continue to occur, provides the best chance of unlocking the agglomeration forces that will make SEZs economically successful and better integrated in local economies. These agglomeration forces may arise in three broad areas. First, a larger market allows for a more efficient sharing of indivisible facilities (such as infrastructure), risks, and the gains from variety and specialization. For instance, a larger cluster of firms will make it easier to construct a dedicated facility or, for specialized input providers, to pay a fixed cost and enter the (larger) local market. Second, a larger market allows for better matching between employers and employees, buyers and suppliers, partners in joint projects, or entrepreneurs and financiers. Third, a larger market can facilitate learning about new technologies, market evolutions, or SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Box Promises and challenges of spatialized industrial policies (continued) 2.2 new forms of organization. Direct interactions between economic agents in a cluster can therefore facilitate the creation, diffusion, and accumulation of knowledge. Hence, although there are indisputable gains to be had from more equitable geographical distribution of economic activity, there is also clear evidence that increasing urbanization has strong, positive externalities through deeper and more extensive forward and backward linkages. In this context, the approach to spatial industrial policy in general, and to SEZs in particular, may benefit from being more attentive to exploiting the potential of urban agglomerations to deliver the large-scale job creation that is urgently needed in South Africa. This may be all the more critical in a constrained fiscal 42 environment (see Chapter 1), where the spread of public investment may seriously dilute its impact. Likewise, South Africa’s three-tier system of investment promotion—national, regional, and city—is cumbersome, a.  uncoordinated, and potentially confusing for incoming investors, who are often unclear about entry points and which agencies to contact for what services and information. World Bank (2015a) also underlines the significant variation in local implementation of national regulations. b. The SEZ fund is designed to finance basic infrastructure development costs for 3 to 5 years. Assessing the impact of industrial policies sectors; the second estimates the impact of It is difficult to remains challenging, given attribution issues investment taxes (as one component of firms’ ascertain whether and the multiplicity of instruments and user UCC) on firms’ investment behaviors objectives and, from there, deducts the impact of tax investments would Assessing the impacts of these various incentives on additional investment and not have been initiatives on their stated objectives nonetheless employment and related fiscal costs. Thus, remains a daunting methodological challenge. the approach followed cannot be strictly made without the This is because, first, multiple objectives considered as one comprehensively assessing incentive are simultaneously sought, and second, the industrial policy, first because it does not beneficiaries can benefit at the same time consider all industrial policy instrument tools from various initiatives, whose performance (such as nontax instruments), and second can also be influenced by various other factors. because it considers some tax instruments that Attributing the causality of an outcome to a may not necessarily be justified on industrial specific initiative is thus complicated. And policy grounds (such as tax provisions specific last but not least, assessment is challenging to the mining sector). because the counterfactual situation of not providing support cannot be observed: as Tax provisions introduce large differences in stressed by PBO (2016), it is critical, but post-tax returns to investments across sectors, inherently challenging, to ensure that the and may have contributed to the capital support results in an increase in investment, allocation losses observed since 2008 exports, research, or employment (depending To measure the impact of tax incentives on the objective) and is not redundant. That on investment, World Bank (2015b) uses is, support should not be provided under the concept and metric of marginal effective circumstances wherein the firm would have tax rates (METRs). In a nutshell, METRs undertaken investment, exports, research, or measure the difference in investment rates employment regardless of whether support of return before and after taxes are paid by was provided. Indeed, redundancy risks the firms. The larger the METR, the lower have been found to be widespread in many the economic incentive to invest in a given developing countries. In 14 surveys conducted sector. METR computations consider several in 14 countries, redundancy rates exceeded 70 tax instruments, some of which are applied percent in 10 of them (James 2014).28 to the use of various investment assets,29 and To clarify some of these methodological some of which are sector specific. Table 2.2 issues, the Davis Tax Committee invited the reports the principal tax parameters used in World Bank to conduct research on the the METR computations in 2014. Noteworthy effectiveness and efficiency of ITIs granted to is the fact that de jure (rather than de facto) South African firms. The exercise led to the tax parameters are considered in the METR production of two reports (World Bank 2015b; computations, ignoring the reality and impact 2016f). The first report measures the extent of tax administration efficiency, tax evasion, and dispersion of investment taxes across and the informal economy. Table Tax parameters used in marginal effective tax rate computations 2.2 Type of tax Rates Remarks Income taxes     Corporate income tax 28% x% = taxable income from gold mining/total Corporate income tax for gold 34 – (170/x) % revenue (turnover) from gold mining. Taxed at progressive rates from 18% to 40% Income was eligible for a primary rebate of 43 Personal income tax in 2014 ZAR 12,080 on the tax calculated in 2014. Eligible for an interest exemption of ZAR Treatment of interest income Taxed at the rates for PIT 23,800. Treatment of dividend income 15% Withheld on distribution. Only 33.3% of the capital gains is included Treatment of capital gains in the taxable income and calculated at the marginal PIT tax rate Indirect taxes     Value-added tax 14% METRs measure the Property tax (immovable property) Various rates 0.15% for farming to 1.7% for commercial difference in the and business property. Sectors such as manufacturing receive rebates Customs duty Various rates on customs duty. investment rate of Other taxes     return before and Mining royalty (unrefined ores) 0.5 + {EBIT/(gross sales x 9)} x 100 after taxes are paid Mining royalty (refined ores) 0.5 + {EBIT/(gross sales x 12.5)} x 100 For generation of electricity from polluting Electricity levy 35 cents/kwh sources. Taken as 1% of the turnover. Sector-specific tax allowances     Depreciation of plant and machinery of 40%, Additional depreciation benefits for invest- 20%, 20%, 20% ments in preferred sectors and IDZs. Manufacturing Additional investment allowance of 100%, This allowance is over and above those who 75%, 55%, or 35%, depending on whether qualify for the accelerated 40%, 20%, 20%, the investment is in the IDZ or a preferred 20% depreciation schedule. sector Depreciation of plant and machinery of Agriculture 50%, 30%, 20% 100% depreciation of plant and machinery Mining Employee housing allowed to be depreciated at 10% straight line as compared with 5% straight line for other sectors 100% depreciation of plant and machinery used in manufacturing Small business corporations Depreciation of plant and machinery of 50%, 30%, 20% for nonmanufacturing activities Source: World Bank 2015b. Note: EBIT = earnings before interest and taxes; IDZ = industrial development zone; PIT = personal income tax. METRs show significant variation across given sector. Thus, even when the depreciation sectors and asset types (Table 2.3). This allowances for separate asset classes are the variation reflects the differences in tax same in different sectors, the fact that sectors rates across some sectors as well, and the use different mixes of assets causes the METRs accelerated depreciation allowances. The to vary. Lastly, the METR for inventory is the accelerated depreciation allowances generate result of the first in, first out accounting for a “tax advantage” that depends on how the tax inventory in South Africa, whereby the assets rate of depreciation compares with the actual rate of economic depreciation for different that are bought first are treated as sold first. asset classes (for example, the depreciation This means that any changes in the value of of buildings is far slower than that for heavy inventory due to inflation results in higher machinery), as well as the actual asset mix of a taxation and higher METRs. SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Table Marginal effective tax rate, by sector and class of investment assets 2.3 (percent) METR Manufac- for invest- turing Mining Agriculture Construction Tourism Services* Transport Electricity ment in asset Land 11.4 15.1 11.4 11.4 11.4 11.4 11.4 12.3 44 Residential 5.8 -3.3 3.5 3.5 3.5 3.5 3.5 7.0 buildings Non-res- idential 23.8 28.1 12.3 9.5 3.5 3.5 3.5 7.0 buildings Construc- tion works, roads, and 18.8 20.9 49.9 47.2 18.8 18.8 18.8 22.4 parking areas The mining and Land im- provements 11.4 -17.6 -24.6 -24.6 -24.6 11.4 11.4 12.3 Network tourism sectors equipment 13.6 31.8 25.6 25.6 25.6 25.6 21.0 33.2 Computers enjoy very large and other 8.6 23.7 37.6 18.2 37.6 37.6 37.6 45.7 IT equip- ment tax advantages Motor vehicles compared with other and other 9.2 24.6 29.1 11.0 19.4 19.4 10.2 29.7 transport equipment sectors Plant, ma- chinery, and -3.1 -30.5 6.3 3.7 2.2 10.1 8.3 -7.9 other office equipment Capital work in -5.9 -30.3 35.8 29.3 34.1 34.9 30.9 37.4 progress Other property, -5.9 -17.0 35.8 29.3 34.1 34.9 30.9 37.4 plant, and equipment Computer 11.9 29.7 11.9 11.9 11.9 11.9 11.9 25.2 software Inventory 35.5 30.8 35.5 35.5 35.5 35.5 35.5 36.4 Overall 19.6 -1.2 17.0 19.5 6.1 14.0 18.8 23.0 METR Source: World Bank 2015b. Note: IT = information technology; METR = marginal effective tax rate. * Includes only real estate and business services, excluding financial and insurance services. Although all sectors enjoy an METR below By contrast, the tourism and mining the corporate income tax (CIT) rate of 28 sectors exhibit the lowest METRs, reflecting percent, manufacturing stands out as the for tourism the negligible weight of inventory second most taxed sector after electricity.30 and the high weight of buildings in its asset This ranking is primarily driven by the high structure. As for mining, and notwithstanding weight of inventory (40 percent) in the asset important variations between minerals,32 the structure of the manufacturing sector, as well low METR is mainly driven by the provision as the comparatively high rate of inflation in that mining companies can immediately and South Africa (5.9 percent when computed in fully write off their capital investment in the September 2014; 6.1 percent in September year it is incurred. The CIT formula for gold 2016).31 In the manufacturing sector, small also contributes to the low METR.33 and medium enterprises enjoy a negative Thus, to generate a post-tax rate of return METR of -4.7 percent, benefiting from a 100 of 10 percent on investment, a pre-tax rate percent capital allowance (as does the mining of return of 8.8 percent is needed in mining, sector). against a pre-tax rate of return of 29.6 percent in the manufacturing sector. These differences Bank (2016f) computes firms’ UCC for 2006– are further exacerbated when the actual debt- 12 using SARS data. 34 The UCC combines, for to-assets ratios are used to compute the METR each asset, the price of the asset, price of the (World Bank 2015b). output of the sector, cost of financial capital, Therefore, although strictly speaking capital allowance rate, present value of capital it is difficult to attribute the recent capital allowances, and corporate tax rate. Thus, allocation losses (Figure 2.7) to these tax estimating the impact of a change in the UCC provisions, given the absence of a common of firms’ investment decision, as described investment response to profitability indicator in the previous section, allows assessing the across mining companies (Table 2.1), such impact (or benefit) and cost of tax incentives. 45 tax provisions are nonetheless very plausible By controlling for the factors that explain candidates. At the extreme, the negative METR permanent differences across firms and over in the mining sector means that companies are the years, such analysis allows addressing the encouraged to invest in projects that provide risk of erroneously attributing investment hardly any economic returns, just to lower outcomes to policy interventions. their tax burden. Table 2.4 shows the average changes in percent in the UCCs for the sectors as a Using firm level Nevertheless, capital allowances have generated result of the various tax incentives compared data allows for a additional investment and jobs in a few selected with regular tax treatment. There are about sectors 290,000 firms and 86 percent of them are small granular assessment All in all, the wide variation in METRs across business corporations (SBCs). For the purpose sectors stems principally from differences in of the analysis, the SBCs are grouped together of the impact of tax capital depreciation allowances. Considering and the results for large businesses are shown incentives such differences by asset type and sector, World by sector. Table Changes in the user cost of capital as a result of investment tax incentives, by sector 2.4 (percent) Sector 2006 2007 2008 2009 2010 2011 2012 Large businesses Agriculture -4.7 -5.0 -4.8 -4.6 -4.4 -4.3 -4.2 Construction -3.9 -4.0 -4.0 -3.8 -3.6 -3.5 -3.5 Manufacturing -4.5 -4.9 -4.9 -4.5 -4.2 -4.0 -4.0 Mining -5.7 -6.5 -6.4 -5.4 -4.1 -4.4 -3.9 Finance -3.3 -3.3 -3.4 -3.2 -3.2 -3.2 -3.2 Services -3.6 -4.0 -3.8 -3.6 -3.4 -3.3 -3.3 Trade -3.6 -3.8 -3.7 -3.5 -3.3 -3.2 -3.1 Transport -3.7 -3.8 -4.0 -3.7 -3.3 -3.2 -3.1 Electricity -3.5 -3.6 -3.5 -3.5 -3.2 -3.2 -3.2 Small business corporations -1.0 -1.3 -1.4 -1.3 -1.2 -1.1 -0.9 Source: World Bank 2016f. The biggest percentage reduction in the is modest for the SBCs. This is because the UCC for 2006–12 is in the mining sector, impact of the tax incentives on SBCs has two with an average reduction in the UCC of -5.2 counteracting effects. Although the higher percent, followed by agriculture (reduction of depreciation allowances are beneficial to 4.6 percent) and manufacturing (reduction the firm, the tax impact of the depreciation of 4.4 percent). The reduction in the UCC allowances is higher when the tax rate is SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS higher, because the resulting deduction is reduced cost of capital from capital allowances higher. However, SBCs benefit from a lower would allow firms to expand their production tax rate and hence the net effect on the firm to satisfy an unchanged demand in value is ambiguous. for their goods and services. Large business The estimated elasticities of the investment corporations are distinguished by sector, and to the UCC (Table 2.1) and changes in the SBCs are aggregated together, given their UCC (Table 2.4) allow measuring the impact relatively smaller size. Sectors with estimated of capital allowances on additional investment elasticities not significantly different from and job creation (Table 2.5). It is assumed zero are assumed not to respond to capital 46 that, in a partial equilibrium setting,35 the allowances. Table Additional capital and labor demand due to investment tax incentives, by sector 2.5 (percent) Sector 2006 2007 2008 2009 2010 2011 2012 Although tax Additional capital demand Agriculture 4.1 4.4 4.2 4.0 3.8 3.7 3.6 incentives Mining significantly reduce Manufacturing 2.5 2.7 2.7 2.5 2.3 2.2 2.2 the cost of investing Electricity Construction 3.1 3.2 3.2 3.0 2.9 2.8 2.8 in mining, there is Trade 2.8 3.0 2.9 2.7 2.6 2.5 2.4 little evidence that Transport Finance they systematically Services 1.5 1.7 1.6 1.5 1.4 1.4 1.4 encourage Small business corporations 0.7 0.9 1.0 0.9 0.9 0.8 0.4 investment Additional labor demand Agriculture 2.0 2.1 2.0 1.9 1.8 1.8 1.7 Mining Manufacturing 1.1 1.2 1.2 1.1 1.0 0.9 0.9 Electricity Construction 1.1 1.2 1.2 1.1 1.0 1.0 1.0 Trade 1.5 1.5 1.5 1.4 1.3 1.3 1.2 Transport Finance Services 0.5 0.6 0.5 0.5 0.5 0.5 0.5 Small business corporations 0.3 0.4 0.4 0.4 0.3 0.3 0.2 Source: World Bank staff calculations. The results suggest that capital allowances in 2012, up from ZAR 2.8 billion in 2006. This increased capital stocks up to 4.4 percent in increase compares with estimated foregone agriculture, against a maximum of 1.0 percent revenue of ZAR 3.9 billion in 2012, 36 up from for SBCs, the latter enjoying smaller changes ZAR 3.0 billion in 2006.37 Thus, foregone in UCC, and 0 percent for sectors in which revenue was more than offset by additional capital demand does not significantly respond investment, and one can consider that capital to changes in user costs. Sectors with initially allowances create some additionality, even higher capital-output ratios tend to benefit if, to reduce their overall fiscal cost, the more from reduced UCCs, given the larger capital allowances could be better targeted to impact on their overall cost structure. All in all, responsive sectors. tax incentives are estimated to have generated The impact on additional demand for jobs an increase in investment of ZAR 4.5 billion stems from two opposite factors: a reduction Figure Impact of investment tax incentives on investment and job creation, 2012 2.8 (percent) Small Business Corporations Services Finance 47 Transport Trade Construction Electricity Manufacturing Tax incentives Mining encouraged significant Agriculture investment in 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Additional labor demand Additional capital demand Reduction in user cost of capital agriculture, construction, trade, Source: World Bank staff calculations. manufacturing and services, also in labor demand due to increased capital Better targeting of capital allowances across resulting in higher intensity (as the price of capital declines sectors would allow generating the same in relation to the price of labor assumed amount of additional investment at a much labor demand in unchanged), and an increase in labor demand lower fiscal cost these sectors due to increased production. The impact is From this calculation, it could be inferred particularly pronounced in the agriculture that the fiscal cost of creating an additional job and trade sectors (ranging between 1.7 and 2.0 through the capital allowance was on average percent and 1.2 and 1.5 percent, respectively, in 2012 close to ZAR 188,377 per year (ZAR against only 0.2 to 0.4 percent in SBCs in 3.934 billion divided by 20,883). Comparing 2012), see Figure 2.8. All in all, tax incentives this estimate with the average gross labor are estimated to have directly created 20,883 remuneration per worker in 2012 of ZAR additional jobs. 107,509 unfavorably suggests that directly Thus, this counterfactual analysis creating one job through capital allowances suggests that absent ITIs, the contraction would exceed the cost of fully subsidizing the of economic activity and job destruction in cost of remunerating a worker. Nonetheless, the manufacturing sector would have been this seemingly excessive cost stems to a large larger. Explanations for industrial contraction extent from the fact that allowances granted since democracy and more recently shall be to some sectors do not generate any additional found elsewhere, possibly among insufficient investment or jobs. Indeed, only retaining skills and infrastructure, policy uncertainty, a allowances granted to sectors creating volatile rand, and complicated labor relations. additional jobs would bring the average As discussed in Box 2.3, the commodity super cost to ZAR 89,026, as more than half of the cycle may also have diverted private investment allowances go to other sectors (Table 2.6). from manufacturing toward the mining sector, and the end of the super cycle could generate an inverse reallocation. SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Table Impact and cost of investment tax incentives on the creation of jobs, 2012 2.6 (jobs and rands) Cost of direct and Direct job creation Foregone revenue Cost of direct job Average remunera- Sector indirect job creation (jobs) (ZAR millions) creation (ZAR) tion (ZAR) (ZAR) Agriculture 2,041 124 60,497 26,427 41,658 Mining - 279 - 192,999 - Manufacturing 5,192 376 72,388 165,616 15,079 Electricity - 617 - 432,400 - 48 Construction 1,961 69 35,194 58,998 18,284 Trade 2,012 78 38,924 53,306 29,501 Transport - 1,038 - 126,759 - Finance - 141 - 116,815 - Services 5,566 157 28,153 133,850 18,443 Small business corporations 4,111 1,056 256,802 107,509 a - Reorienting tax incentives to Total 20,883 3,934 188,377 107,509 - responsive sectors Source: World Bank staff calculations. a. Average national labor remuneration in the absence of sufficient information on small business corporations. such as agriculture, Indirect job creation strongly reduces the unit agriculture, trade, services, construction, manufacturing, fiscal cost of creating a job through capital and manufacturing. Among these sectors, construction, trade allowances manufacturing has by far the largest indirect Furthermore, these cost estimates do not job creation component. For every job that and other services account for indirect job creation through is directly created in manufacturing, another multiplier effects. Following Tregenna (2016), 3.8 jobs are indirectly created, bringing down would increase their we compute import-adjusted employment the fiscal cost of creating a job by 4.8, from effectiveness and multipliers,38 using the social accounting ZAR 72,388 to ZAR 15,079. Accounting for matrix for 2012 computed by Chitiga-Mabugu indirect job creation also brings the fiscal cost-efficiency (2016), at the sector disaggregation level cost of creating an additional job down to used in World Bank (2016f). Although the ZAR 18,284 for construction, ZAR 18,443 for estimates of the fiscal cost to create a job services, ZAR 29,501 for trade, and ZAR 41,658 cannot be strictly compared with employment for agriculture. Among these five sectors, multipliers (as the former only consider large support to services would create the largest businesses), it is nonetheless reassuring to see proportion (53 percent) of skilled workers, that the five sectors for which a significant followed by trade (46 percent), manufacturing effect of tax incentives can be observed are also (41 percent), construction (25 percent), and those with the largest employment multipliers: agriculture (18 percent), see Figure 2.9. Figure Employment multipliers, 2012 2.9 a. Direct and indirect job creation b. Unskilled and skilled job creation Agriculture Agriculture Trade Trade Other services Other services Construction Construction Manufacturing Manufacturing Finance Finance Transport Transport Mining Mining Electricity Electricity 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 Direct Indirect Unskilled Skilled Source: World Bank staff calculations based on Chitiga-Mabugu 2016. The cost of directly creating a job through 2014 and March 2016. Larger firms were the Investment tax incentives is broadly comparable largest claimants by rand value. Financial to that of directly creating a job through services, trade, manufacturing, and agriculture employment tax incentives, when controlling were the largest claimants of the incentive, as for remuneration these sectors had the highest proportion of Finally, the impact and cost of capital eligible workers in the overall workforce. allowances can be compared with those The National Treasury uses administrative obtained from direct employment incentives.39 data to estimate econometrically the impact In 2014, South Africa launched the ETI, to of such incentives on job creation. The results encourage firms to employ workers ages 18 to 49 suggest that on average the incentive increased 29 years. The incentive acknowledges the lower probability of being employed without prior labor demand (number of workers ages 18-29 professional experience (Anand, Kothari, and earning less than ZAR 78,000 annually) by Kumar 2016). Since its inception, the take- 10.6 percentage points, with greater relative up of the incentive has been strong, and the impacts for smaller firms, see Figure 2.10. related, cumulative foregone revenue was Such job creation implied a unit fiscal cost per estimated at ZAR 6.1 billion between January job created of about ZAR 37,334 per year. Large employment multipliers in the Figure Impact of employment tax incentives on job growth, 2014/15 manufacturing 2.10 (percentage points) sectors strongly 30 amplify the impact 20 of tax incentives on job creation 10 10001 to >50000 50000 Percentage points 0 1 to 5 6 to 10 11 to 50 51 to 101to 201to 401to 601to 801to 1000 to 5001 to 100 200 400 600 800 1000 5000 10000 -10 -20 -30 -40 Firm size by employment (full year equivalent) Source: National Treasury 2016b. As these computations exclude potential in percentage of labor remuneration. By indirect job creation through multiplier benefiting only workers earning less than ZAR effects, they need to be compared with the costs 78,000 annually, creating a job through ETI of direct job creation through ITIs (Table 2.6). represents a subsidy equal to at least 48 percent At ZAR 37,334 per year, the cost of direct job of the labor remuneration. In comparison, creation through ETI compares favorably with creating a job through ITI is equivalent to a that of ZAR 188,377 through ITI. The cost also subsidy representing about 80 percent of the compares favorably with the cost of direct job average labor remuneration in the responsive creation through ITI in responsive sectors only, sectors, and to a subsidy representing about which is ZAR 89,026 on average. Nonetheless, 44 percent of the average labor remuneration the difference in costs narrows when expressed in the manufacturing sector. In other words, SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS compared with ETIs, ITIs have the potential suggest that the economy has been suffering to create better remunerated jobs, although at at the aggregate level from lower commodity a higher cost. This finding is not surprising, as prices (inducing a loss equivalent to 4 percent ETIs encourage job posting of unexperienced of GDP in 2012). But the results also suggest workers, while ITIs encourage capital that investing in the manufacturing sector 50 deepening and related demand for skilled and should become much more attractive under experienced labor. this new external environment, and increase Recent external developments could favorably overall demand for labor. Thus, the results amplify the positive impact of investment tax of the simulations exercise further justify the incentives use of effective industrial policy instruments South Africa’s industrial policy aims to to overcome structural rigidities and capture reveal and amplify the country’s comparative positive externalities. Employment advantages and build on them to boost growth To date, the reallocation of factors toward tax incentives and job creation. Since the outset of the sectors enjoying new comparative advantages current decade, the South African economy has nonetheless been modest, although complement has experienced two major external shocks: a reallocation has started to pick up, as noted in investment tax sharp depreciation of the domestic currency Chapter 1. Between 2012 and 2015, the mining against that of most of the country’s trade sector saw its actual share in total capital incentives as they partners, and a significant decline in the world stock increase by 0.3 percentage point, while target different price of South Africa’s main export commodity, the simulation results suggest a significant mineral products. Although such changes in contraction in the sector’s capital demand types of jobs relative prices over a relatively short period over the long term. Conversely, the business of time may certainly not be the only factors services sector saw its share decline since 2012 influencing the determination of comparative (see Figure 2.6) while the model suggests it advantages and related investment decisions, would gain from terms of trade changes. they cannot be ignored, given their large Long reaction time from investors may magnitude and likely persistence.40 Indeed, explain these delays. But other factors may based on trade patterns (by product and have also contributed to prevent a rapid partner) recorded in 2012, between 2012 and reallocation of factors (and the avoidance of 2015, import prices (expressed in rand) grew allocation losses recorded between 2012 and by 35.8 percent; exports prices, excluding 2015; see Figure 2.7). The other factors could minerals, grew by 39.6 percent; and mineral include policy uncertainty; increased risk export prices decreased by 18.7 percent. aversion to investing in sectors presenting new, As such, these evolutions could contribute but maybe fragile, comparative advantages; and to reverse some of the negative factors that the influence of nonmarket forces in capital are often cited to explain South Africa’s allocation. International Monetary Fund slow growth and modest employment gains: (2014) indeed emphasizes the negative impact an overvalued currency and an excessive of electricity bottlenecks, limited market concentration of factors of production in competition, and labor market constraints on the extractive industry, non-tradable, and the export supply response to the depreciation consumption sectors (Zalk 2014; Bhorat, Tian, of the rand. The continued preference given and Ellyne 2014). through ITI to invest in poorly remunerative Box 2.3 reports the results of a simulation mining projects is also a factor in explaining exercise meant to capture the long-term the delayed response to the changing terms of impacts of these external developments on the trade. allocation of capital across sectors. The results Box Terms of trade, exchange rate movements, and new comparative advantages 2.3 In recent years, the South African economy was strongly shaken by several external shocks, namely, the end of the commodity super cycle, capital flight from the BRICs (Brazil, the Russian Federation, India, China, and South Africa) to high- income economies, and China’s rebalancing of its production toward domestic markets. Combined, these shocks had a very significant impact on South Africa’s terms of trade. Box 2.3 figure 1 reports variations in exchange rates with trade partner groups, and variations in mining export prices, expressed in nominal rand. The cumulative variation in consumer prices and factor remuneration (gross domestic product (GDP) deflator) is also reported. 51 Box 2.3 figure 1: Cumulative price variations, in rand, 2012–15 (percent) } Consumer price index 17% GDP deflator index 17% Aggregate Indexes } -35% Other mining With the end of the Gold & uranium ore mining 9% Export prices of.. Coal mining commodity super ROW 2% cycle, South Africa Oceania 13% Western Asia 54% recorded massive South Eastern Asia 41% South Central Asia 29% changes in its terms Eastern Asia 56% European Union 34% Export prices to.. and import prices of trade from Europe ex EU 2% South & Central America NAFTA 55% SADC 17% Africa ex SADC 36% -60% -40% -20% 0% 20% 40% 60% 80% 100% Source: World Bank staff calculations based on World Bank 2016b.  AFTA = North American Free Trade Area; ROW = rest of the world; SADC = Southern African Development Note: N Community. In all likelihood, such ample variations in relative prices are expected to influence investment decisions, given their impact on the relative profitability of capital across sectors. For instance, the 38 percent drop in the export price of other mining combined with the 17 percent increase in the remuneration of capital and labor observed across sectors should have significantly reduced the profitability of the sector. Nonetheless, the impact of these shocks is difficult to estimate for a couple reasons: first, because other factors may have influenced investment decisions over the same period (one of them being “policy uncertainty” reflected in short-term exchange rate volatility; another one being insufficient electricity and inland transport capacity); and second, because the period might be too short to observe the long-term consequences of shocks. For these reasons, we employ a computable general equilibrium model to assess ex ante the long-term impact of the shocks. The model is static, in the sense that the quantity of factors of production (labor types by skills, capital) is fixed and fully used. The model also assumes perfect factor mobility across sectors, to identify the impact of shocks on long-term investment decisions (capital reallocation) under perfect factor market competition. Obviously, the reality is different, as the potential market failures that could justify the industrial policy discussed in this Chapter are not considered in this modeling exercise. But these assumptions are precisely useful to identify potential comparative advantages that could be mobilized through the correction of market failures. In other words, the assumptions are useful for assessing the scalability of some policy interventions as helping a given sector to realize its comparative advantage and meet its effective business opportunities. The model is built using a social accounting matrix for 2012 (Chitiga-Mabugu 2016). The model comprises 54 sectors of activities, four labor types, capital, 12 trade partners, and one representative household. It assumes imperfect substitutability between imports and domestic products (on the demand side), as well as between exports and domestic products (on the supply side). The model assumes imperfect substitution elasticities (equal to 1.0, equivalent to Cobb-Douglas functions of production) between factors of production, and fixed consumption of intermediate inputs in proportion of total output (and thus value added) at the sector level (Leontief structure). Traditional closure rules apply, including (i) savings-driven SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS Box Terms of trade, exchange rate movements, and new comparative advantages (cont.) 2.3 investments, (ii) fixed trade balance, and (iii) fixed real government expenditures and policy tax rates (thus, an endogenous fiscal balance affecting the savings available for investment). We use the model to simulate the long-term impact of the observed variations between 2012 and 2015 in the terms of trade, through exchange rate variations and changes in commodity prices (as reported in box 2.3 figure 2; induced variations in the Consumer Price Index and GDP deflator being endogenously determined by the model). The simulation results suggest that the terms of trade observed between 2012 and 2015 induced a negative transfer from the rest of the 52 word, equivalent in real terms to 4.0 percent of GDP in 2012. This estimate, which is to be compared with the actual 5.3 percent cumulative real GDP growth over 2012–15 (1.7 percent a year) could be interpreted as a lower bound estimate, as market failures could have inhibited efficient factor reallocation across sectors. In other words, cumulative real GDP growth could have at least reached 9.3 percent in the absence of such shocks (3.0 percent a year). The results also suggest that the induced decline in the real remuneration of factors was felt differently across sectors. On the one hand, the remuneration of unskilled labor and capital declined by 6.0 and 1.8 percent, respectively, with respect to the GDP deflator. On the other hand, remuneration increased for informal labor by 9.2 percent (highly concentrated in retail trade, one of the main beneficiary sectors of the shocks), for highly skilled labor by 3.2 percent, and for skilled labor by 2.6 percent. The recent change In other words, since the factor supply is fixed at the aggregate level, terms of trade changes could have induced higher demand for skilled and informal labor, and lower demand for capital and unskilled labor. in the terms of Box 2.3 figure 2: Capital reallocation across sectors trade has induced (percentage point change in sector capital share) a significant redistribution Other services of comparative Finance advantages across Transport sectors, in favor of Trade manufacturing and Construction trade sectors Electricity Manufacturing Mining Agriculture -5 -4 -3 -2 -1 0 1 2 Source: World Bank staff calculations. From a sector perspective, the results suggest that capital would need to be reallocated toward manufacturing (based on nonferrous metals), trade and accommodation (wholesale and retail trade). and financial services (business services) to meet new comparative advantages. Capital would need to be reallocated away from the mining sector (other mining), which would see its share in total capital stock drop from 11.8 to 7.5 percent. (A second best option, more likely to materialize in the short run, is a strong decline in mining profitability, given capital’s relative inertia across sectors.) Note: World Bank (2016e) provides a detailed discussion of the nature and impact of recent shocks. Conclusion thus greater employment demand. Since 1994, South Africa’s capital-labor ratio has increased Three consecutive years of negative per steadily. Although the capital-labor ratio rose capita economic growth highlight the structural in most sectors, capital went disproportionally challenges South Africa faces in creating jobs to low-paid, labor intensive sectors (services in and achieving significant reductions in poverty particular, most of them being shielded from and inequality. international competition). This composition Although job creation was almost nil in the effect reduced the substitution of labor by first decade of the democratic era, it suddenly capital and protected jobs; but it conversely accelerated from 2005, was temporarily slowed the reallocation of labor toward the 53 halted during the peak of the global financial most productive sectors, which is also called crisis in 2008, and reversed from 2015. The structural transformation. period of job creation coincided with the Although several policy instruments commodity super cycle, and likely illustrates have been deployed to promote industrial the continued high dependency of the South development, fundamentally they have all African economy on mineral commodities. consisted of encouraging the redeployment of We estimate that the change in the terms of private investment toward the industrial sector. trade recorded since 2012 may have cost 4 PBO (2016) reports that ZAR 476 billion (in percentage points of GDP. 2015 constant prices) was cumulatively devoted The authorities are fully cognizant of the to national industrial development programs shortcomings of South Africa’s economic and initiatives between 1994 and 2014, to structure, which is characterized by low which can be added the direct financing of domestic competition and dependency on industrial projects from IDC (ZAR 178 billion natural resources. The authorities have thus between 1994 and 2013, in 2013 prices), the accelerated efforts to promote industrial impact of procurement and competition development, with the expectation that policies, and local government initiatives. the acceleration could potentially generate Nonetheless, such initiatives translated into several positive outcomes, including higher neither a significant reallocation of private wages for workers (benefiting from potential capital toward industrial sectors, nor higher labor productivity gains from technological industrial employment. Thus, the share of the investments); larger employment and growth manufacturing sector in South Africa’s total multipliers through forward and backward capital stock has decreased since 1994, by 3 linkages with other sectors; increased domestic percentage points (with most of the decline competition through greater exposure to occurring since 2008), and 335,000 jobs were highly competitive world markets; increased lost over the same period (including 207,000 reliance on the larger world demand, as the jobs since 2008). potential for stimulating higher domestic Such a contrast between industrial policy demand through debt financing is being ambitions and industrial development calls progressively exhausted for households and for an assessment of the impact of industrial government; and more stable growth, as policies on private investment. Thus, in underpinned by stable global demand for response to a request from the Davis Tax manufactured goods and a less volatile capital Committee, the World Bank undertook to account. Through appropriate investments, measure the extent of ITIs across sectors, and industrial development can seek to expand their impact on demand for capital. production in sectors where comparative As a first step, marginal effective tax rates advantage already exists, or alternatively (METRs) were computed for the various develop new comparative advantages through sectors of activity. In a nutshell, the METR technological upgrading. measures the difference in investment rates In the process of industrialization, of return before and after taxes are paid by investment is expected to foster technological firms. The larger the METR is, the lower is the upgrading. Thus, capital-labor ratios are economic incentive to invest in a given sector. expected to rise, with the expectation that METR computations consider several tax related improved competitiveness will entail as instruments (including in particular capital well the expansion of volumes produced, and allowances), some of which are applied to the SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS use of various investment assets, and some of in the beneficiary sectors would have been which are sector specific. The METRs show 30,000 jobs less. At first sight, this looks like significant variation across sectors: although a very modest and expensive outcome, given all sectors enjoy an METR below the corporate the implicit high fiscal cost per job created income tax rate of 28 percent, manufacturing (ZAR 188,377). This cost exceeds the average stands out as the second most taxed sector. annual remuneration of workers in 2012, In contrast, the tourism and mining sectors and, with a view to encourage job creation, exhibit the lowest METRs, reflecting for directly subsidizing the full employment cost tourism the negligible weight of inventory would have thus been cheaper for the same 54 and the high weight of buildings in its asset result. However, when only considering the structure. As for mining, and notwithstanding sectors that are responsive to incentives, the important variations between minerals, the fiscal cost of creating jobs through capital low METR is mainly driven by the provision allowances is more than halved, as sectors such that mining companies can immediately and as mining, electricity, transport, and finance fully write off their capital investment in the benefit from more than half the total capital year it is incurred. The corporate income tax allowances in value. Moreover, accounting for formula for gold also contributes to the low the indirect job creation generated by higher METR. production further reduces the fiscal cost. In As a second step, the impact of capital the manufacturing sector in particular, the allowances—which are the principal cost was brought down to about ZAR 15,079, contributors to the observed differences in which is comparable to that estimated for the METRs across sectors—on private investment service sector, which is much labor intensive was tested econometrically using the South and where remunerations are lower. African Revenue Service tax data for more Therefore, the results suggest that than a million of firms over 2006–12. The the investment incentives granted to the methodology used allows for controlling manufacturing sector actually contained, at a for other factors that could have affected modest cost, the destruction of jobs recorded private investment, and is thus able to provide in this sector since 1994. estimates of the counterfactual situation In light of these results, it is likely that without capital allowances. The results suggest several other factors may have affected first that several sectors are not systematically investment in the manufacturing sector. responsive to changes in the cost of capital, Although it is beyond the scope of this report which is influenced by capital allowances: to explain this phenomenon, it is striking to see mining, electricity, transport, and finance. that capital allocation has deteriorated since Second, the overall implicit cost for the fiscus 2008 and generated economic losses since is significantly larger than that reported in the 2012, indicating that capital moved away from Budget Reviews, although it was still modest the most lucrative sectors. This finding could (at ZAR 3.9 billion) in 2012 compared with the suggest that investors and capital markets have full panoply of industrial policy instruments. become more risk averse, or less attracted by Third, without such capital allowances, potentially higher remuneration in sectors private investment would have been lower by where profitability may be questionable over a quantum (ZAR 5.1 billion in 2012) larger time, as policy and macroeconomic uncertainty than the subsidy. Thus, capital allowances can remains. High capital rents in some sectors, be considered effective and efficient overall in the result of insufficient competition, may their attempt to encourage private investment. also deter the government’s efforts to alter In addition, the methodology was also used relative prices at the margin. Other factors, to derive the implicit job creation resulting such as the insufficient supply of skilled labor, from the additional investment encouraged by reflecting the poor quality of the education capital allowances. The results suggest that the system, or the poor reliability of the electricity induced substitution of labor by capital, which supply, may have prevented the development was encouraged by the capital allowances, of new industrial projects. But the continued was more than offset by larger volumes of preference given through tax incentives to production. All in all, it was estimated that invest in poorly remunerative mining projects without capital allowances, labor demand is also a plausible factor to explain this capital misallocation and delayed response to the the various sectors. In particular, the question massive changes in the terms of trade recorded arises whether the capital investment in mining since 2012. should be given preferential treatment, and Some successes have already been if investment allowances in mining should recorded in the energy domain, for instance, be brought closer in line with other sectors, where supply now meets demand, given such as manufacturing. The greater potential accelerated investment in capacity in the for job creation in manufacturing has been recent past. While continuing to address the reinforced in recent years by the end of the macroeconomic, competition, education, and commodity super cycle and the protracted infrastructure challenges, it may be warranted devaluation of the rand. 55 to revisit some of the tax provisions granted to SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS 56 Endnotes 1 See Maveé, Perrelli, and Schimmelpfennig is not a given. Extra working-age people (2016). These findings confirm those of does not equal more output if the extra Clark et al. (2016), who underline the people are not working. Bruni, Rigolini, prime influence of commodity prices and Troiano (2016) suggest that, given the on the volatility of emerging markets’ current joblessness, demographic change external capital accounts (as opposed to will exert a drag on the economy as the U.S. monetary policy). population keeps aging. On the contrary, 2 Due to the data categories that were if South Africa were to converge to the available, Chapter 1 uses Statistics South current Organisation for Economic Co- Africa’s Labor Force Surveys for labor operation and Development employment figures, and Chapter 2 uses Quantec’s ratio by 2050, gross domestic product per employment figures, which are derived capita would be 41 percent higher than from Statistics South Africa’s Quarterly under current policies. Employment Surveys. 8 In 2015, the private sector, including 3 The Talent Shortage Survey is conducted agriculture, accounted for about nine in annually across 43 countries by Manpower 10 jobs, formal and informal, in South Group. Africa. 4 See Ethics Institute (2016). 9 Throughout this Chapter, we define 5 Given SAB Miller’s cross-listing at the the private sector as all sectors minus London Stock Exchange, only a part of government services, that is, including the transaction will have impacted the parastatals that are commercially run exchange rate and/or can attract private capital. 6 Over a third of South African government 10 Before 2000, labor force surveys bonds is held by nonresidents. conducted by StatsSA were annual and 7 Demographic transition, reflected in the did not consider informal employment. increasing share of working age population From 2000, they started to consider in total population (the result of significant informal employment. From 2000 to 2007, gains made over 1950–90 in increased they became bi-annual, before being life expectancy and reduced fertility), conducted every quarter since 2008. could allow workers to build massive 11 StatsSA defines unemployed workers as: financial savings (instead of supporting “persons aged 15–64 years who: a) Were dependents’ daily consumption) to be not employed in the reference week; and invested in education, health, technology b) Actively looked for work or tried to start and innovation, infrastructure, and a business in the four weeks preceding the productive capacity, thereby significantly survey interview; and c) Were available for raising labor productivity and real wages. work, i.e., would have been able to start Such a “demographic dividend,” however, work or a business in the reference week; or d) Had not actively looked for work 17 The authors thank Jorge Maia for his in the past four weeks but had a job or characterization of the three distinct business to start at a definite date in the periods. The accelerated growth of capital future and were available.” StatsSA defines stock since the late 1990s is generally a discouraged job seeker as a person who attributed to the adoption of the inflation- was not employed during the reference targeting monetary policy regime and the period, wanted to work, and was available related decline in real interest rates. to work or start a business, but did not 18 Sector rates of return to capital are take active steps to find work during the measured by the ratio of sector gross past four weeks, provided that the main operating surplus over sector capital 57 reason given for not seeking work was any stock (source: Quantec 2016). Dispersion of the following: no jobs available in the is measured by the standard deviation area, unable to find work requiring his/ over the average of sector rates of return her skills, or lost hope of finding any kind (weighted by the share of the sector in of work. total GDP) for a given year. The results 12 Agriculture, forestry, and fishing are robust to the level of disaggregation is not comprised of any subsectors. used. Figure 2.7 reports computations Mining and quarrying is comprised of using a 45-sector disaggregation level. A three subsectors; manufacturing, 28 similar computation using a nine-sector subsectors; electricity, gas, and water, two disaggregation level also suggests growing subsectors; construction, two subsectors; dispersion in rates of return to capital trade, catering, and accommodation, since 2008. two subsectors; transport, storage, and 19 Capital productivity is measured by the communications, two subsectors; financial ratio of GDP to aggregate capital stock. intermediation, insurance, real estate, 20 This relatively high contribution of and business services, two subsectors; and capital reallocation to GDP growth, by community, social, and personal services, international standards, could be reflective four subsectors. of South Africa’s high untapped potential 13 Sectors that recorded a decline in capital for capital reallocation across sectors and intensity include beverages, tobacco, firms. Using South Africa’s firm-level textiles, wearing apparel, leather and data, including services, manufacturing, leather products, coke and refined and nonagricultural primary activities petroleum products, basic iron and steel, for 2007, Lashitew (2012) estimates metal products excluding machinery, that the complete elimination of capital other transport equipment, wholesale and misallocation across firms could induce retail trade, and business services. a gain of 84 percent in total factor 14 As an illustration, it is likely to be simpler productivity (or equivalently GDP) levels. to purchase a new machine that would This compares with an average of 38 respond to the needs of a given sector percent for the 77 low- and middle-income than to change the functions of an already countries considered in the study, ranking existing machine. For instance, a tractor South Africa 74th (from best to worst) in used in agriculture cannot be converted capital allocation. into a computer for services. 21 Firms’ capital stock is here measured by 15 Total capital stock is the sum of sectoral the sum of the following assets: Property, capital stocks used throughout this Plant and machinery and Other Assets. analysis (source: Quantec 2016). The pace One methodological caveat with this of reallocation of capital across sectors is approach is the possible re-evaluation of measured by the sum of absolute changes assets from one year to another, which in sectoral capital shares with respect to would induce a change in the capital the previous year. stock not resulting from depreciation or 16 The computations of sectoral capital investment. While there is no obvious way reallocation and capital stock growth to control for assets re-evaluation (as it is in Figure 2.5 do not include portfolio not reported in tax files), it is likely that re- investments. evaluation mostly concern firms’ financial SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS assets, which are not considered in this METR measure, because the basis for the analysis. assessment of property values is subjective, 22 See World Bank (2016e) for a discussion and thus unlikely to represent market on competition. values uniformly. 23 Between end-2008 and July 2016, an 31 Comparison of METRs across countries additional capacity of 33,400 gigawatt for manufacturing and services sectors hours was put online, including 19,769 nonetheless suggests that South gigawatt hours from July 2015. The Africa corporate income tax regime is increase resulted from very regular annual internationally competitive, in particular 58 investments throughout the period. with respect to G7 countries (World Bank 24 Although it is beyond the scope of this 2015b). report, a review of the allocative efficiency 32 When the inflation rate is changed, say of South African capital markets may also to 2 percent, the METR decreases to 12.1 be warranted to explain the allocation percent, showing the sensitivity of the losses recorded since 2012. METR on inventory to inflation. 25 This acceleration is reflected in the 33 The METRs for minerals range from a share of the national budget devoted to high of 31.9 percent for iron ore to a low national (as opposed to local) industrial of -19.7 percent for chrome (World Bank development initiatives. From 2008 to 2015b). 2013, an annual average of 3.4 percent of 34 Profit-based incentives and general tax the public budget was devoted to national incentives are generally not recommended industrial development initiatives, against for activities that generate location-specific 2.6 percent from 1994 to 2007 (source: rents, such as mining. staff calculations based on PBO (2016)). 35 Compared with World Bank (2015b), 26 Costly past failures in African and Latin World Bank (2016f) only retains three American countries (Robinson 2009) have types of investment assets (as recorded underlined the risks of political capture in the SARS data set): building, plant and that industrial policy has a smaller and equipment, and other fixed assets. chance of success when the selection The formula for calculating the UCC of industrial projects and locations is also assumes (due to lack of information driven by political considerations as on firms’ financing variables) that all opposed to economic ones (PBO 2016). investment is financed by retained Close monitoring and evaluation and earnings. As such, the UCC computations sunset clauses can help reduce the risks do not capture the advantage (in terms of of political capture; and reliance on METR) given to firms relying on debt to comparative advantages determined by fund their investments. endowment structures can help reduce 36 Consistent with World Bank (2016f), the risks of economic failure (Lin 2012). it is assumed that firms minimize costs 27 Over 1994–2013, IDC financing of projects through factor demand derived from in South Africa cumulatively amounted to a constant elasticity of substitution ZAR 178 billion (2013 constant prices). production function, and that the price of 28 The SEZ fund is designed to finance basic output is fully determined by factor costs. infrastructure development costs for 3 to 5 Demand for the products is assumed to be years. exogenous, and is thus not affected by the 29 Redundancy rates are measured by the introduction (and related financing) of percentage of investors that claim they ITIs. would have invested even without tax 37 These estimates differ from the tax incentives. expenditures reported in the Budget 30 Payroll taxes and excise taxes on fuel are Reviews (National Treasury 2016a), also omitted from consideration, since, because the former consider not only arguably, their amounts are unaffected the impact of capital allowances provided by marginal increases in the capital under sections 12G and 12L, but also stock. Property taxes affect the returns that of capital allowances provided under to capital but are excluded from our sections 12B, 12C, 12D, and 13. Further, tax expenditures are not additive and may created in other sectors through derived interact with each other. See World Bank demand for domestically produced inputs (2016f) for more details. (hence, the computations do not include 38 The inclusion of additional corporate imported inputs). taxes from additional investment only 40 These paragraphs draw from the marginally affects the total fiscal cost (an preliminary analysis conducted by the estimated increase in corporate taxes of National Treasury (2016b). ZAR 36 million only in 2012). 39 Employment multipliers measure the 41 World Bank (2016a) foresees that 59 number of jobs created directly and world nominal prices of mineral goods indirectly from additional production in a (expressed in US$) will stay below their given sector. Indirect effects capture jobs 2012 levels until 2025. SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS 60 References Anand, R., S. Kothari, and N. Kumar. 2016. Rating Downgrade: What Happens to “South Africa: Labor Market Dynamics Borrowing Costs When a Government and Inequality.” Working Paper Loses Its Investment Grade Credit 16/137, International Monetary Fund, Rating?” Macroeconomics and Fiscal Washington, DC. Management Discussion Paper 13, World Bhorat, H., N. Tian, and M. Ellyne. Bank, Washington, DC. 2014. “The Real Exchange Rate and ———. 2016b. “South Africa and the Ghost Sectoral Employment in South Africa.” of a Rating Downgrade to Sub-Investment Development Policy Research Unit Grade.” Macroeconomics and Fiscal Working Paper 201404, University of Cape Management Practice Note 15, World Town, South Africa. Bank, Washington, DC. Bloom, N. 2009. “The Impact of Uncertainty Haver Analytics. 2016. “Middle East and Shocks.” Econometrica, 77: 623–685 African Emerging Markets – South Africa.” Bruni, L., J. Rigolini, and S. Troiano. 2016. Haver Analytics, London. “Forever Young? Social Policies for a International Monetary Fund. 2014. “South Changing Population in Southern Africa.” Africa’s Exports Performance: Any Role World Bank, Washington, DC. for Structural Factors?” Washington, DC. Chitiga-Mabugu, M. 2016. “SAM for South James, S. 2014. “Tax and Non-Tax Incentives Africa 2012.” Mimeo, University of and Investments: Evidence and Policy Pretoria, South Africa. Implications.” Investment Climate Clark, J., N. Converse, B. Coulibaly, and Advisory Services, World Bank, S. Kamin. 2016. “Emerging Market Washington, DC. Capital Flows and U.S. Monetary Policy.” Lashitew, A. 2012. “Misallocation, Aggregate International Finance Discussion Paper Productivity and Policy Constraints: Cross- Note, Board of Governors of the Federal Country Evidence in Manufacturing.” Reserve System, Washington, DC. University of Groningen, Netherlands. DTI (Department of Trade and Industry). Lin, J. 2012. “New Structural Economics: A 2016. “Industrial Policy Action Plan: Framework for Rethinking Development Economic Sectors and Employment and Policy.” World Bank, Washington, DC. Cluster, IPAP 2014/15 – 2016/17.” DTI, Maveé, N., R. Perrelli, and A. Schimmelpfennig. Pretoria, South Africa. 2016. “Surprise, Surprise: What Drives Ethics Institute. 2016. “South African Citizens’ the Rand/U.S. Dollar Exchange Rate Bribery Survey 2016.” http://www.tei.org. Volatility?” Working Paper 16/205, za/images/pdf/Citizens-Bribery-Survey- International Monetary Fund, 2016-FINAL.pdf. Washington, DC. Hanusch, M., S. Hassan, Y. Algu, L. Soobyah, McMillan, M., and D. Rodrik. 2011 and A. Kranz. 2016a. “The Ghost of a “Globalization, Structural Change and Productivity Growth.” Working Paper ———. 2015b. “Sector Study of Effective Tax 17143, National Bureau of Economic Burden & Effectiveness of Investment Research, Cambridge, MA. Incentives in South Africa: Part I.” World National Treasury. 2016a. “Budget Review Bank, Washington, DC. 2016.” Pretoria, South Africa. ———. 2016a. “Africa’s Pulse.” No. 14, ———. 2016b. “Employment Tax Incentives October. World Bank, Washington, DC. Assessment.” Pretoria, South Africa. ———. 2016b. “Commodity Markets Outlook.” PBO (Parliament Budgetary Office). 2016 “The Cost and Outcomes of Industrial World Bank, Washington, DC. Development Initiatives, 1994/95– ———. 2016c. “El Niño Drought in SADC: 61 2014/2015.” Cape Town, South Africa. Economic Impacts and Potential Quantec. 2016. “SA Standardised Industry Mitigation Solutions.” World Bank, Data.” Pretoria, South Africa. Washington, DC. Robinson, J. 2009. “Industrial Policy and ———. 2016d. “Macro Poverty Outlook for Development: A Political Economy South Africa.” World Bank, Washington, Perspective.” World Bank ABCDE DC. Conference. Seoul: Harvard University. ———. 2016e. “South Africa Economic Syrquin, M. 1986. “Productivity Growth and Update: Promoting Faster Growth and Factor Reallocation.” In Industrialization Poverty Alleviation through Competition.” and Growth: A Comparative Study, edited by H. Chenery, S. Robinson, and M. World Bank, Washington, DC. Syrquin. New York: Oxford University ———. 2016f. “’South Africa: Sector Study of Press. Effective Tax Burden and Effectiveness Tregenna, F. 2016. “Sectoral Dimensions of of Investment Incentives in South Africa Employment Targeting.” In Employment – Firm Level Analysis.” World Bank, Targeting: Macroeconomic and Sectoral Washington, DC. Approaches, edited by K. Chatani, I. Islam, Zalk, N. 2014. “Industrial Policy in a Harsh and F. Lapeyre. Geneva: International Climate: The Case of South Africa.” Labour Organization. In Transforming Economies: Making World Bank. 2015a. “Doing Business in Industrial Policy Work for Growth, Jobs South Africa 2015: Comparing Business Regulations for Domestic Firms in 9 Urban and Development, edited by Manuel Areas and 4 Major Ports with 188 Other Salazar-Xirinachs, J. I. Nübler, and R. Economies.” World Bank, Washington, Kozul-Wright. International Labour DC. Organization, Geneva. SOUTH AFRICA ECONOMIC UPDATE—PRIVATE INVESTMENT FOR JOBS The World Bank 442 Rodericks Road, Lynnwood, Pretoria 0081, Republic of South Africa Tel: +27 (0) 12 742 3100 Fax: +27 (0) 12 742 3135 Twitter: #SouthAfricaEU www.worldbank.org/za www.facebook.com/WorldBankSouthAfrica