Understanding the Middle-Income Trap: Challenges and Policy Implications
The middle-income trap poses challenges for economies stagnating at middle-income levels. This trap is associated with difficulties in capital accumulation, shifting production structures, and increasing productivity. Competing explanations and policy implications are discussed, focusing on South Africa and other countries at risk. Definitions, countries affected, and the need for structural change and innovation are highlighted in this comprehensive analysis.
Download Presentation
Please find below an Image/Link to download the presentation.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. Download presentation by click this link. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.
E N D
Presentation Transcript
David Fryer and Nicolette Cattaneo Rhodes University Presented at the Conference on Manufacturing-led Growth for Employment Equality Johannesburg 20-21 May 2014
Introduction Middle income trap: economies stagnating at middle income levels or risk thereof. Associated with difficulties in a number of dimensions including a) capital accumulation b) shifting the structure of production to more productive sectors (structure) and c) increasing productivity within sectors. Debate and competing explanations on definitions, causes and associated policy implications. Key issues: a) Adverse structural change? Is productivity a useful way of conceptualising development? b) methodology and data: very problematic CIBS countries: characterised as either in the trap (SA and, in most studies, Brazil) or vulnerable to it.
Outline 1. The middle income trap: definitions and countries 2. Competing explanations: structural change, productivity growth and development 3. Productivity and employment growth in South Africa in comparative perspective 4. Conclusion: policy implications and future research
Definitions and countries Middle income trap primarily associated with Latin America and some African countries (including SA). Also some late-industrialising Asian economies such as Malaysia. Concern that China and India are at risk (Aiyar et al, 2013:33) (although debated). Alternative definitions impact on which countries are classified as being in the trap or at risk.
Definitions and countries A common definition is in terms of growth decelerations: middle income countries slow down before convergence with high income countries (Eichengreen et al, 2013). Aiyar et al (2013), for example, look at sustained slowdowns in total factor productivity growth. Others examine the period for which countries have been categorised as middle income (Felipe, 2012, Wade, 2010) Further, the thresholds used for the middle income category are debated and seen as somewhat arbitrary.
Figure 1: Is there a middle income trap? Source: Aiyar et al. (2013: Figure 6)
Competing explanations It is generally held that development requires structural change and increasing innovative effort. The middle income trap is associated with problems in both of these dimensions. UNIDO (2012) defines structural transformation as the evolution of an economy s structure from low productivity traditional activities to higher productivity modern activities . Seen as desirable for growth and development as a source of productivity growth and higher incomes to diversify economic structure and decrease vulnerability to external shocks.
Competing explanations In the process of structural change the role of the manufacturing sector is highlighted due to particular characteristics of manufacturing that facilitate growth and development (UNIDO, 2012; Tregenna, 2012). There is at the same time considerable debate on the pros and cons of alternative growth paths, particularly those associated with the services sector. The problem of whether structural change in the CIBS and other middle-income developing countries is growth-enhancing or growth-reducing is tied to this debate via the different patterns of structural change that are emerging at the sector and sub-sector level.
Competing explanations There are divergent explanations of emerging patterns of structural change and the inability of countries in the middle-income trap to converge or make the transition to high income levels. Orthodox view: based on the standard convergence property of neoclassical (Solow) growth theory. In this view, failure to converge is related to countries resisting market-conforming structural change, weak institutions , populist policies and so forth.
Competing explanations Heterodox Schumpeterian-Keynesian approach: sees the concept of international convergence differently. Depends on lagging countries moving towards increasing returns activities . Instead of treating technology as exogenous, in this view technological diffusion is far from being a passive process. It requires major efforts at catching up and learning (Cimoli et al, 2011: 28). This implies that active industrial and technology policies are critical in order to foster structural change. In this view, a lack of such policies helps to explain development failure.
Productivity and employment growth in South Africa in comparative perspective Remainder of paper explores characteristics and drivers of productivity growth in SA from a comparative perspective, focusing on the role of structural change. Various methodologies exist for decomposing productivity growth and its sources. This section uses data from Penn World Tables (Penn8) Total economy database (TED) and GGDC sectoral databases (GGDC) SA Standardised Industrial Database (Quantec, 2014) Statistics SA (2014) based on QLFS
Figure 2: Different sources for aggregate employment in South Africa
Figure 3a: Value added (Va) per worker (Emp) in the CIBS countries, constant 2005 rand (1000s) Source: Penn8, except: * employment series from TED and capital stock from Penn8; ** SASID.
Figure 3b: Total factor productivity in the CIBS (index, 2005=1) (TED)
Figure 4: Capital labour ratios in the CIBS (R1000/worker, 2005 R)
Productivity and employment growth in South Africa in comparative perspective The shift-share decomposition methodology has been used in recent work on the CIBS countries to explore the role of structural change in particular both at the broad sector level and within manufacturing. However most of the comparative work to date does not include South Africa or does not have up to date South African data. The shift-share methodology decomposes productivity changes as follows:
Productivity and employment growth in South Africa in comparative perspective The left-hand side of the expression reflects the total change in labour productivity reflecting intensity of learning and technological catching-up (Cimoli et al). The right-hand side decomposes this change into different sources of growth: I: that due to changes in employment shares , with labour productivity in each sector held constant II: that due to a rise in employment shares in sectors whose productivity has increased in that period III: that due to productivity growth in each sector, with employment shares held constant.
Table 1: Shift share components, South Africa and Brazil 1972-2005 (three main sectors) a) South Africa Pio/Po 0.655 1.280 1.121 II III Annual Pi/Pio Si -0.190 0.001 0.189 I (shift) -0.124 0.001 0.212 0.089 primary secondary tertiary b) Brazil -0.095 0.001 0.073 -0.022 0.163 0.112 0.186 0.461 0.528 1.29% Pio/Po 0.218 2.073 1.470 II III Annual Pi/Pio Si -0.271 0.009 0.262 I (shift) -0.059 0.020 0.385 0.345 primary secondary tertiary -0.151 0.002 -0.074 -0.223 0.257 0.035 -0.101 0.191 0.313 0.83% Sources: SASID (South Africa); GGDC (Brazil)
South Africa Brazil a) employment share b) value added per worker Source: own calculations from SASID (South Africa) and GGDC (Brazil)
Table 2: Shift-share decomposition for Brazil (9 sectors) P/Po (Unido) annual P/Po annual 0.80% 0.10% 0.50% I - II - III - I II III 05-09 00-05 95-00 90-95 95-09 0.2904 -0.008 -0.0208 -0.114 - -0.1288 0.5184 -0.1919 0.089 - 0.3664 0.0126 -0.00272 -0.0139 -0.08% 0.3127 0.0145 -0.00859 0.0377 0.86% -0.0294 -0.01272 0.0703 0.56% - - 0.525 - 0.5624 - 0.80% 72-95 0.309 -0.208 0.2127 0.008 Source: Left panel: Own calculations based on GGDC; Right panel adapted from Unido (2012: Table 3.2)
Table 3: Shift-share decomposition for South Africa a) GGDC data P/Po Annualised 2.13% 1.26% 2.24% -1.85% 2.56% 4.18% 9 sectors 1994-2010 1994-2000 2001-10 1982-93 1973-81 1960-72 I (shift) 0.107 0.033 0.017 0.127 0.162 0.263 II III (within) 0.430 0.111 0.216 -0.246 0.067 0.306 overall 0.400 0.078 0.221 -0.186 0.224 0.634 -0.137 -0.066 -0.012 -0.067 -0.005 0.066 3 sectors 1994-2010 1982-93 1973-81 1960-72 0.051 0.071 0.097 0.267 0.000 -0.030 0.005 0.066 0.349 -0.227 0.121 0.302
b) SASID data (shift) I (dynamic) (within) II Pi/iPo total Pi/iPo total III annual I II III annual 3 sectors '94-13 '82-93 '73-81 '73-13 9 sectors 94-13 82-93 73-81 '73-13 46 sectors '94-13 '82-93 '73-81 '73-13 manufacturing only (28 sectors) '94-13 0.1143 -0.0910 '82-93 0.0106 '73-81 0.0214 -0.0115 '73-13 0.1785 -0.0839 tertiary only (10 sectors) '94-13 0.0730 -0.0823 '82-93 0.2805 -0.1548 -0.1136 0.0121 '73-81 0.1549 -0.0254 -0.0578 0.0716 '73-13 1.0047 -0.7467 0.0711 0.0168 0.0236 0.1067 -0.0666 0.0006 0.0042 -0.0700 0.4982 -0.0093 0.0693 0.6250 0.5027 0.0080 0.0971 0.6617 2.17% 0.07% 1.17% 1.28% 0.7481 0.7714 3.06% 0.07% 1.88% 1.70% 0.0249 -0.0279 0.0076 0.1510 0.1609 0.8712 0.9659 0.1657 0.1225 0.0928 0.7954 -0.0598 -0.1067 -0.0101 -0.6270 0.3968 -0.0078 0.0144 0.4933 0.5027 0.0080 0.0971 0.6617 2.17% 0.07% 1.17% 1.28% 0.2791 0.2698 1.26% 0.11% 0.87% 0.84% 0.1378 0.3958 0.2176 0.1767 0.1430 1.0306 -0.1185 -0.1146 -0.0205 -0.7242 0.4036 -0.0541 -0.0253 0.3554 0.5027 0.0080 0.0971 0.6617 2.17% 0.07% 1.17% 1.28% Source: SASID
Policy implications and research An important policy lesson is that sectors important for growth and those important for employment may not be the same (see Tregenna, 2012). Need to study the impact of structural change and productivity growth on employment, poverty AND inequality In addition, need to consider value added growth and investment rates in decompositions; sectoral dist of investment and the investment-productivity relationship Group MEC and non-MEC sectors to do decompositions It is necessary to study sectoral and sub-sectoral patterns of change alongside the broader macro/ development/ comparative analyses to facilitate the development of industrial and technology policy in context'. Is a need to interrogate the usefulness of the productivity concept as an indicator of progress (especiallyTFP).
Policy implications and research The paper emphasises at the outset that there are alternative perspectives on what drives growth and development The policy prescriptions that flow from conventional and heterodox approaches differ accordingly. Trade liberalisation and associated policies of the late 80s and the 1990s had important consequences for economy-wide productivity in Latin America and some African countries, leading to productivity- reducing structural change and, in some BRICS countries, premature deindustrialisation.
Policy implications and research The alternative heterodox approach emphasises a coherent nexus of trade, industrial and technology policies to facilitate learning by doing and appropriate structural change. In addition, labour market policies that move towards a decent work agenda and a macro framework that addresses stability concerns arising from volatile financial flows are required.
