- PRELIMINARY DRAFT - THE ECONOMIC FOUNDATIONS OF GROWTH-ORIENTED DEVELOPMENT POLICIES * Martin Hallet ** 1. Introduction In recent years there has been a remarkably synchronised political shift towards recognising the importance of economic growth, relative to more equity-oriented approaches, for successful development policies. It tends to be based on the observation that policies within or between countries to increase the income of the poor cannot be permanently based on transfers, but must be based on creating long-term opportunities for people to make a decent living on their own. Even if the GDP concept has well-known shortcomings - such as a weak capture of resource use, externalities, the grey economy or income distribution (see e.g. Stiglitz et al. 2010) - it remains the best proxy for welfare or happiness that is currently available and internationally comparable. For these reasons, GDP (per capita) growth is indeed a relevant target for economic policies. The pros and cons of a growth-oriented compared to equity-oriented development policies are not the focus of this paper. Most of the economic literature confirms that an extreme equity orientation of policies will be harmful to growth, while growth-oriented policies may have more open results with a view to distributional effects, depending rather on accompanying policies. There are indeed good reasons why growth has to be an essential component of long-term economic and social development. It implies a more sustained and reliable increase in income through job creation and higher productivity while a reliance on transfers and external aid can be reversed when policies change. Furthermore, growth is essential for providing the tax base for public finance as a primary source for financing social safety nets as well as for public goods, including those which are required for a sustained growth process such as infrastructure or education. Providing the policy framework for unleashing market-based growth forces can also be expected to be a more efficient development strategy rather than trying to lean against such forces by concentrating exclusively on the least favoured regions and people. Success stories of * Keynote speech presented at the IWH/INFER-Workshop on Applied Economics and Economic Policy on “The Empirics of Imbalances and Disequilibria” on 14 February 2011 in Halle (Saale). ** The views expressed in this paper are exclusively those of the author and do not necessarily correspond to those of the European Commission for whose Directorate-General for Economic and Financial Affairs the author is working. 1 persistently high growth rates, such as in China and other emerging markets, are an important reference for many policy makers, while strictly equity-based strategies, such as in Cuba, have shown a poor medium-term performance in improving their citizens’ life. On the other hand, an extreme growth orientation can be accompanied by rising inequalities, even if reducing poverty in absolute terms, which raises concerns of a humanitarian nature. It can also erode the political support for growth policies and eventually trigger a policy reversal towards a stronger equity orientation. The main objective of this paper is to explore what economists know and do not know about economic growth as a basis for sound advice to those designing growth-oriented development strategies. We will first present some evidence to demonstrate that there is indeed a recent shift towards a stronger growth orientation of development policies at various institutional levels (section 2). The paper proceeds by reviewing the theoretical and empirical foundations of such growth policies (section 3). The main hypothesis is that even if there is a good knowledge of the driving factors and determinants of growth, the empirical evidence is fairly weak so as to inform more precisely on the required priorities and the impact of growthoriented policies in specific countries or regions. The main reason is the complexity of growth processes – in particular characterised by non-linearities, circular causalities and dynamic adjustment processes – which are difficult to capture with many of the currently applied empirical methods. Section 4 summarises what this implies for growth-oriented development policies. 2. THE SHIFT TOWARDS A STRONGER GROWTH-ORIENTATION OF DEVELOPMENT POLICIES A stronger orientation at economic growth as the focus of development policies can be observed at international, European, national and regional (sub-national) levels. However, it would be incorrect to understand this as an extreme or sudden shift from the policy objective of income equalisation/poverty reduction towards economic growth. Both objectives have always been important for policy-makers, and indeed no government is likely to remain in power for a long time if it neglected totally one of these objectives. The shift is rather a gradual one with a stronger policy focus on growth, accompanied by policies to ensure that its distribution is not perceived to be grossly unfair. An attempt to reconcile these two policy objectives is frequently made by qualifying that growth needs to be “shared”, "equitable" or “inclusive”. There may be several explanations for this shift. The financial and economic crisis 2008/09 may have reminded everybody of the crucial importance of economic growth without which many years of progress in poverty reduction, job creation or fiscal consolidation can be erased. Another explanation could be the success of emerging market economies in improving their quality of life and reducing poverty on the basis of sustained and high growth, even if at times at the expense of increased income inequality. Increased globalisation may also imply that isolated 2 policy strategies that put an excessive emphasis on income distribution might have become more difficult to pursue as these tend to go along with losses in external competiveness. Finally, several decades of external transfers in the form of development aid or regional policies, with their often not very visible results, might also have created perceptions of the limits of such policies and the risks of aid dependency. At a global level, the Millennium Development Goals (MDGs) have been the primary orientation of development policies. In 2000/2001 the international community had agreed on the MDGs regarding poverty, hunger, health and education, expressed in quantitative global targets to be achieved by 2015. This was accompanied by the 'Monterrey consensus' in 2002 to specify the actions by which policies by developing countries themselves and by donors should contribute to achieving these MDGs. One of the first signs for a reconsideration of this strategy can be seen in the creation of the “Commission on Growth and Development” in 2007. It worked under the chairmanship of Nobel Laureate Michael Spence and with Robert Solow and various policy-makers as members; the 'Spence Commission' was supported by World Bank staff. Its key output called "The Growth Report - Strategies for Sustained Growth and Inclusive Development" was published in May 2008. The report takes the view "that growth is a necessary, if not sufficient, condition for broader development, enlarging the scope for individuals to be productive and creative." The objective of the report was to identify some of the distinctive characteristics of high-growth economies and to offer a policy framework that could help political leaders with a serious longterm commitment create a growth strategy for their country. The shift from an exclusive orientation towards achieving the (MDGs) towards a stronger growth orientation is also taking place at UN level. The Outcome Document of the UN HighLevel Plenary Meeting on the MDGs of June 2010 states in paragraph 43: "We stress that promoting sustained, inclusive and equitable economic growth is necessary for accelerating progress towards achieving the Millennium Development Goals as well as promoting sustainable development, but it is not sufficient: growth should enable everyone, in particular the poor, to participate in and benefit from economic opportunities and should lead to job creation and income opportunities and be complemented by effective social policies". A similar Koreaninitiative resolution on "Sustained, inclusive and equitable economic growth for poverty eradication and achievements of the MDGs" was adopted by the UN General Assembly in November 2011. The resolution pursues four concrete aims: (i) to promote the exchange of best practices and lessons learned on "sustained, inclusive and equitable growth"; (ii) to encourage regional cooperation in this respect; (iii) to ask ECOSOC to hold a panel discussion on sustained, inclusive and equitable growth during its July 2011 substantive session; and (iv) to ask the Secretary General to include analysis and recommendations on growth in its annual MDG progress report until 2015. 3 In November 2010 G20 Leaders agreed on the "Seoul Development Consensus for Shared Growth". It has the "overarching objective of helping low-income countries improve and maintain the levels and quality of growth, thereby reducing poverty, improving human rights and creating decent jobs". G20 Leaders also agreed on a Multi-Year Action Plan based on six principles of which the first is a "focus on economic growth". The Action Plan is grouped under nine 'pillars' which are infrastructure, private investment and job creation, human resource development, trade, financial inclusion, growth with resilience, food security, domestic resource mobilization, and knowledge sharing. In the EU, the Lisbon strategy and the Europe 2020 strategy also imply a strong growth orientation of EU economic policies. Based on a proposal by the European Commission in spring 2010, the European Council agreed on the Europe 2020 strategy for jobs and smart, sustainable and inclusive growth. This should be implemented by boosting competitiveness, productivity, growth potential, social cohesion and economic convergence. Among others, Member States should identify the main bottlenecks to growth and indicate, in their National Reform Programmes, how they intend to tackle them. In January 2011, the European Commission published the first ‘Annual Growth Survey’ as a starting point for the “European Semester”. Its objective is to provide ex ante coordination between EU governments and to improve the prospects for growth and stability in the EU and the euro area. Furthermore, Member States receiving support from EU/IMF programmes to restore debt sustainability have to implement conditions not only on fiscal policies but also on a number of structural reforms so as to enhance the medium-term growth potential. The Europe 2020 strategy is projected into various EU policies, including in EU cohesion policy as guidance to the action of Member States and the Union. The European Commission's Fifth Cohesion Report proposes to establish an explicit linkage of Cohesion Policy and Europe 2020 in order "to continue helping the poorer regions of the EU catch up, to facilitate coordination between EU policies, and to develop the EU cohesion policy into a leading enabler of growth, also in qualitative terms, for the whole of the EU, while addressing societal challenges such as ageing and climate change." While the Objective 1 and Convergence Objective always had a strong growth orientation by focussing on infrastructure, human capital and the productive sector, Europe 2020 might give a reinforced guidance to prioritise the creation of growth and jobs. The European Commission further proposes to give its external development policies a stronger growth orientation. It published in November 2010 a consultative Green Paper on "EU development policy in support of inclusive growth and sustainable development" which asks stakeholders questions around four main objectives to be pursued collaboratively by the EU and its Member States: (i) how to ensure high EU impact development policy so that every euro spent provides the best value added and value for money; (ii) how to facilitate more, and more inclusive, growth in developing countries, as a means of reducing poverty and provide a chance for all to have a decent living as it is increasingly obvious that MDGs will not be achieved 4 without it. Each percentage of growth can significantly improve countries' capacity to achieve poverty reduction and have a multiplier effect through employment creation and social protection; (iii) how to promote sustainable development as a driver for progress; and (iv) how to achieve durable results in the area of agriculture and food security. The tension between an orientation at equity/distribution objectives as compared to an orientation at efficiency/growth objectives is also a long-standing debate in regional development policies within countries. Most EU Member States in a process of catching-up are facing such choices in defining their regional development strategies (see e.g. European Commission 2002, 2004). For example, following the German unification there was widespread financial support across the whole East German territory, based on the constitutional requirement for policies to establish similar living conditions throughout the German territory. However, towards the end of the 1990s there were increasingly doubts on the efficiency and sustainability of such an approach. Subsequently, policies were often implicitly or explicitly adjusted towards a stronger concentration of support on growth poles, while keeping some support to rural development in lagging regions. 3. THE THEORETICAL AND EMPIRICAL FOUNDATIONS OF GROWTH POLICIES Traditional growth theory, by identifying some of the key growth ingredients, provides a basic orientation for the design of growth policies. The main production function variables of physical capital, labour and, and the ‘Solow residual’ of technology and innovation are analysed in order to identify the steady state rate of balanced growth. An important implication of this theory was that the higher the savings and investments rate in a country, the higher will be its growth rate. Furthermore, capital should flow into economies with lower income so that these countries' growth rates should be higher and their income would converge to the level of highincome economies. As some of the implications were hardly supported by the facts, growth theory evolved further during the last decades. There was in particular criticism regarding the ‘Solow residual’ in that the key factor of long-term productivity growth remained an unexplained black box. It was also observed that capital was flowing mainly between high-income economies or even “uphill” from less to more developed economies. Even if the traditional growth theory does not predict an automatism of absolute convergence of per capita incomes, there is a still a need to identify the conditions under which convergence can occur (Barro/Sala-i-Martin . New or endogenous growth theory tried to account for these observations by introducing variables such as human capital or R+D into this model which have externalities that can lead to equilibria which result in permanent growth and income differences (Romer). A lot of literature in the last two decades aimed at identifying the main determinants of long-term productivity growth. There is fairly wide agreement on the importance of trade, geography, and the quality of economic 5 institutions, governance or policies (Dollar/Kraay 2003, Jeffrey Sachs 2003, and Danni Rodrik et al. 2002 respectively). Furthermore, in open economies there can be an – at least temporary – equilibrium of some countries having an export-led growth with savings exceeding investments to support a debt-financed growth in countries where investments exceed savings. This constellation has become most visible in the form of global imbalances. The policy relevance of growth models also suffers from the absence of a monetary and financial sector. There is ample evidence about the importance of a well-developed financial sector for economic development (Beck et al. 2007). By intermediating between savings and investments, the financial sector provides important functions, notably the accumulation of capital and the allocation to its most efficient uses, as well as a smoother adjustment to external shocks. The financial crisis also taught us that markets are not always efficient or in equilibrium, that momentum effects can lead to deviation of markets from fundamentals, and that such bubbles will burst at some turning point triggered by a “Black Swan”. The financial crisis also made economists realise that there is a significant gap between macro-financial phenomena in the real world and in their models. This has led to a recent surge in behavioural models which aim to capture more adequately such developments on financial markets. Empirical research, facilitated by progress in information technology and an improved availability of data, confirmed the relevance of most of the above mentioned growth determinants. The predominant empirical approach applied to gain more insights into the importance of the different growth determinants is to use linear regressions (OLS) on crosscountry, time series or pooled data. This became possible because of progress on the availability of data on long-term income developments, notably by the Penn World Table by Heston and Summers, and of the technology to process them. Data allowing international comparisons across countries on the growth determinants are becoming increasingly available, for example on doing business, on governance, or on corruption. In the EU, significant progress was also made on the availability of regional statistics, based on an agreed classification of regions (“NUTS regions”), which allow more empirical work on regional growth in Europe. Nevertheless, finding adequate data for the different variables across a sample of countries or regions and for sufficiently long time periods remains a major challenge. Linear regressions cannot fully reflect the complexity of growth which is the result of a multidimensional process of many decentralised decisions by economic agents. Theoretical approaches already suggest that an impressively long list of variables needs to be included in regressions in order to take into account all possible effects. These variables often interact among them. Even where more complex equilibrium models are used, their parameters usually have to rely on the results from regressions. Furthermore, any empirical approach to model growth has to effectively deal with at least three important phenomena: (i) non-linearities, (ii) circular causality (or endogeneity), and (iii) the dynamics of adjustment. 6 • Non-linearities: As the growth of production and income tends to happen in the form of clusters because of Marshallian externalities, it can hardly be expected to be a linear process across space, time, economic sectors or population groups. There is ample literature with hypotheses about the non-linearity of income, growth and development in a hump-shaped or U-shaped relationship with other variables. Among these variables are income inequality (Simon Kuznets), regional disparities (Jeffrey G. Williamson, Albert O. Hirschman, François Perroux), as well as economic integration and trade costs (Paul Krugman and others on New Economic Geography). Other observations like lock-in or threshold effects can also be underlying explanations for non-linearities across countries. • Circular causality and endogeneity: The concept of circularity is well established in economics since François Quesnay, John Maynard Keynes and others. In the context of development policies the concept was applied by Gunnar Myrdal who illustrated the existence of poverty traps arising from a vicious circle of education and poverty. In practice, it is indeed hard to find any growth determinant for which one could safely say that causality is only going one way. For example, is a country poor because its administration is corrupt or is there widespread corruption because the government cannot afford to give its civil servants a decent salary? Is growth constrained by a lack of infrastructure or is growth and the tax base too weak to afford higher spending on infrastructure? Aid selectivity is another example of the difficulty in establishing whether development aid worked better in countries with good policy frameworks or whether these countries received more development aid because of their better policies. • The dynamics of adjustment: The fact that growth is observed in the form of business cycles can also be seen as an expression of continuous adjustment to imbalances. In addition, changes in institutions and policies often take considerable time until their economic impact fully unfolds. This is in particular the case for supply-side measures, such as improvements in infrastructure or education, for which the effects are to be measured in decades rather than years. The ‘Lucas critique’ points to the additional difficulty of evaluating the impact of such policy measures which may have changed the historical parameters derived in regression estimates. For other measures, such as fiscal consolidation or the liberalisation of state monopolies which are usually painful in terms of immediate job losses, things may get worse before they get better. International commitments on green growth and low-carbon development strategies may imply that in the future growth will increasingly need to take place under an emissions or resource constraint. Economists are of course aware of these problems of their empirical approaches, and indeed a number of econometric techniques have been developed to deal with them. Notably lagged, interacted and instrumental variables as well as fixed effects are frequently applied and bring some improvement in this respect. By using average or trend growth rates one can address to some extent business cycles and other adjustment processes. However, many authors claim 7 that these techniques can fully address these underlying problems and that their results can be taken for hard evidence beyond any doubt. This seems to underestimate the nature of the problems and to overestimate the power of their empirical methods to deal with them. In particular, many of the above problems occur simultaneously and interact with each other, which makes it difficult to fully control them. 4. IMPLICATIONS FOR GROWTH-ORIENTED DEVELOPMENT POLICIES Growth theory provides relevant guidance on what are the most important drivers of growth. Capital accumulation and the available labour force are essential ingredients. Their productivity is determined by how efficiently capital is allocated (financial sector), how labour is able and forced to adapt, to innovate and to use knowledge (education, R&D, management, trade openness), and how efficiently markets are working (infrastructure, institutions and economic policies). These elements are, for example, mostly reflected in the pillars of the G20 Development Action Plan. Empirical research tends to confirm the relevance of these growth determinants, but has limits in determining their relative importance. Due to a number of inherent problems of empirical methods described above – i.e. non-linearities, circular causality, and the dynamics of adjustment – it is difficult for economists to provide clear advice on the prioritisation and sequencing of policy reforms and spending to support growth. This is even more so when country-specific development strategies need to be defined. Limits of data availability, notably in developing countries, further complicate such endeavours. It is similarly difficult for economists to evaluate the success or failure of growth-oriented policy measures. Also because of the above mentioned problems of empirical methods, evaluating programmes or projects regarding their effects is extremely difficult since the counterfactual situation needs to be established in order to assess what the effects would have been without the measures taken. For example, an infrastructure project might take decades rather than years to unfold its full supply-side effects which will require rather bold assumption in any evaluation approach. Econometric techniques and models have to be used to demonstrate the ‘mechanics’ of growth policies in a given country, and the results have to be interpreted with some degree of caution. The difficulties in providing evidence on the positive effects of aid to poorer countries or regions often make it difficult for policy-makers in donor countries or regions to stabilise their taxpayers’ support for development aid and regional transfers. Altogether, economists know quite something about growth policies, but they do not know everything. Policy-makers should take this into account when seeking advice from economists instead of falling into an opportunistic approach of ‘whatever works’ if economists admit that they do not have all the answers to their questions. At the same time, economists should not pretend to know everything when asked questions by policy-makers. Sooner or later this will 8 create political frustrations if results do not turn out as predicted. There is still considerable progress possible for strengthening the economic foundations of growth-oriented development policies. REFERENCES Barro, Robert J. and Xavier Sala-i-Martin (1991): Convergence across states and regions; Brookings Papers on Economic Activity, No. 1, pp.107-182. Beck, Thorsten, Asli Demigurc-Kunt, and Ross Levine (2007): Finance, inequality and the poor; Washington D.C. Commission on Growth and Development (2008): The Growth Report Strategies for Sustained Growth and Inclusive Development; Washington D.C. Dollar, David and Aart Kraay (2003), Institutions, Trade, and Growth; Journal of Monetary Economics 50, pp.133-162. European Commission (2002): EU Economy Review - Chapter 5: Real and nominal convergence in the transition accession countries - Selected issues; Brussels/Luxembourg. European Commission (2004): EU Economy Review - Chapter 2: Catching-up, growth and convergence of the new Member States; Brussels/Luxembourg. European Commission (2010): Investing in Europe’s future November 2010 - Fifth report on economic, social and territorial cohesion; Brussels/Luxembourg Fujita, Masahisa, Paul Krugman, and Anthony Venables (2001): The Spatial Economy: Cities, Regions, and International Trade; Cambrige/Massachusetts. Hirschman, Albert O. (1958): The strategy of economic development; New Haven. Kuznets, Simon (1955): Economic Growth and Income Inequality; American Economic Review, Vol. 45, No.1. Lucas, Robert E. jr. (1976): Econometric Policy Evaluation: A Critique; in: Brunner, K.; Meltzer, A.: The Phillips Curve and Labor Markets, Carnegie-Rochester Conference Series on Public Policy, New York, pp.19–46. Myrdal, Gunnar (1957): Economic theory and underdeveloped regions; London. 9 Perroux, François (1955): Note sur la Notion de 'Pole de Croissance'; Economie Appliquée, Vol.7, pp.307-320. Rodrik, Dani, Arvind Subramanian, and Francesco Trebbi (2002): Institutions Rule: The Primacy of Institutions over Geography and Integration in Economic Development; IMF Working Paper No.189. Romer, Paul M. (1990): Endogenous technological change; Journal of Political Economy, Vol.98, pp. S70-S102. Sachs, Jeffrey D. (2003): Institutions don't rule: direct effects of geography on per capita income; NBER Working Paper 9490. Solow, Robert M. (1956): A contribution to the theory of economic growth; Quarterly Journal of Economics, Vol.70, pp.65-94. Stiglitz, Joseph E., Amartya Sen, and Jean-Paul Fitoussi (2010): Report by the Commission on the Measurement of Economic Performance and Social Progress; Paris Williamson, Jeffrey G. (1965): Regional Inequality and the process of national development: A description of the patterns; Economic and Cultural Change, Vol.13, pp.1-84. 10