79445 JULY 2013 • Number 122 Inclusive Growth Revisited: Measurement and Determinants Rahul Anand, Saurabh Mishra, and Shanaka J. Peiris The call for inclusive growth has been unanimously broadcasted by policy makers across the world. The Arab Spring, the growing divide between Main Street and Wall Street in advanced economies, and the “three-speed� world economy have placed inclusive growth at the forefront of policy debates. However, efforts to measure and assess the determinants of inclusive growth have remained limited. What is inclusive growth? How can the micro and macro dimensions of in- equality and growth be integrated to reflect both the pace and distribution of economic growth? What has driven inclu- sive growth in emerging markets? What Is Inclusive Growth? This note presents a measure of inclusive growth that is in line with the absolute definition of pro-poor growth, Inclusive growth refers to both the pace and distribution of but not the relative definition. Under the absolute defini- economic growth. For growth to be sustainable and effective tion, growth is considered to be pro-poor as long as poor in reducing poverty, it needs to be inclusive (Berg and Ostry people benefit in absolute terms, as reflected in some agreed 2011a; Kraay 2004). The Commission on Growth and Devel- measures of poverty (Ravallion and Chen 2003). In con- opment (2008) notes that inclusiveness—a concept that en- trast, under the relative definition, growth is pro-poor if compasses equity, equality of opportunity, and protection in market and employment transitions—is an essential ingredi- and only if the incomes of poor people grow faster than ent of any successful growth strategy. However, attempts to those of the population as a whole; that is, inequality de- measure inclusive growth have remained limited. Tradition- clines (Dollar and Kraay 2002; IMF 2011). By focusing on ally, poverty (or inequality) and economic growth analyses inequality, the relative definition could lead to suboptimal have been conducted separately. Recent work indicates that outcomes for both poor and nonpoor households. For ex- there may not be a trade-off between equity and efficiency, as ample, a society attempting to achieve pro-poor growth un- suggested by Okun (1975), and “that it would be a big mistake der the relative definition would favor an outcome charac- to separate analyses of growth and income distribution� (Berg terized by average income growth of 2 percent, where the and Ostry 2011b). This paper attempts to integrate the two income of poor households grew by 3 percent over an out- strands of analyses by developing a unified measure of inclu- come where average growth was 6 percent, but the incomes sive growth. Ianchovichina and Gable (2012) describe inclu- of poor households grew by only 4 percent. The dynamic sive growth as raising the pace of growth and enlarging the size measure of inclusive growth proposed here allows an analy- of the economy by providing a level playing field for invest- sis of income distribution that can distinguish between ment and increasing productive employment opportunities. countries where per capita income growth was the same for 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise the top and the bottom of the pyramid by accounting for come growth and (ii) income distribution. Similar to the con- the pace of growth. sumer theory where the indifference curves represent the A recent flurry of media and political attention toward changes over time in aggregate demand, this analysis decom- rising inequality across the globe has generated a tremendous poses the income and substitution effect into growth and dis- amount of interest on its causes and consequences. While the tributional components. The underlying social welfare func- rise in inequality in the Organisation for Economic Co-opera- tion must satisfy two properties to capture these features: (i) tion and Development (OECD) and some emerging markets it is increasing in its argument (to capture growth dimension) is well documented, there is debate on the causes and even and (ii) it satisfies the transfer property—any transfer of in- more controversy on the consequences and what should be come from a poor person to a richer person reduces the value done about it. A number of recent papers have associated the of the function (to capture distributional dimension). rising inequality with technological change, financial deepen- The macro measure of inclusiveness is based on the mi- ing, and certain aspects of globalization (Acemoglu and Autor cro concept of a generalized concentration curve following 2011; Aizeman, Lee, and Park 2012; IMF 2007). Foreign Ali and Son (2007). The population is arranged in the ascend- trade can exacerbate inequality by rewarding industries and ing order of their income, called the social mobility curve. Let firms that are able to compete in the global marketplace, while yi be the average income of the bottom i percent of the popula- punishing those that cannot. Technological progress has also tion, where i varies from 0 to 100 and y is the mean income. yi been widely put forth as a structural driver of inequality. is plotted for different values of i (curve AB in figure 1). Since Skilled workers are better able to adopt and use new and im- a higher curve implies greater social mobility, growth is inclu- proved technology than other unskilled workers, thereby in- sive if the social mobility curve moves upward at all points. creasing the skill premium and widening the wage gap be- However, there may be degrees of inclusive growth depending tween skilled and unskilled workers. The divide between on: (i) how much the curve moves up (growth) and (ii) how Main Street and Wall Street epitomizes the recent thinking the distribution of income changes (equity), that is, how the on the role of financial deepening in fueling inequity. curvature of the social mobility curve changes. This feature of Welfare considerations of high inequality extend beyond the social mobility curve is the basis of the proposed integrat- the effect on growth and macroeconomic stability, but they ed measure of inclusive growth. Thus, if two generalized con- remain relevant to understanding whether macroeconomic centration curves do not intersect, they could be ranked on fundamentals and structural change (broadly defined) affect social mobility, that is, inclusive growth. inclusive growth. For example, current debate on austerity To illustrate the point made above, figure 1 depicts two and growth, or recent calls to slow the pace of financial deep- social mobility curves with the same average income (y), but ening and globalization, may reduce income inequality, but different degrees of inclusiveness, that is, different income could slow inclusive growth as well. It is vital to assess the dy- distribution. Social mobility curve A1B is more inclusive namics and determinants of inclusive growth, keeping in than the social mobility curve AB, because the average in- mind that the goal of reducing inequality is not to hurt the come of the bottom segment of the society is higher. If both rich at the expense of the poor.1 terms are positive (d y > 0, dω < 0), growth is unambiguously A unified measure of inclusive growth allows researchers inclusive (AB shifting to A1B1 in figure 2); similarly, if both and policy makers to identify growth determinants and pri- terms are negative (d y < 0, dω < 0), growth is unambiguously oritize country-specific constraints to build inclusive growth. noninclusive (AB shifting to A4B4). However, there could To do this, the next section develops a measure of inclusive be a trade-off between y and ω. If the first term is positive but growth using a macro social mobility function following the the second term is negative, higher social mobility is achieved micro literature on income distribution. This note also docu- at the expense of reduction in equity. In figure 1, this case ments the evolution of inclusive growth, focusing on emerg- can be illustrated by the shift of the social mobility curves ing markets and low-income countries. Lastly, this note exam- from AB to A2B2. Similarly, if the first term is negative but ines the sources of inclusive growth in emerging markets and the second term is positive, then higher social mobility is low-income countries. achieved at the cost of contraction in average income—in fig- ure 1, this case can be illustrated by the shift of the social Measuring Inclusive Growth mobility curve from AB to A3B3. To integrate equity and growth in a unified measure, this note To capture the magnitude of the change in income distri- proposes a measure of inclusive growth based on a utilitarian bution, this analysis uses a simple form of the social mobility social welfare function drawn from consumer choice litera- function by calculating an index (or social mobility index) ture, where inclusive growth depends on two factors: (i) in- from the area under the social mobility curve: 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 1. Shifts in Social Mobility Curve Table 1. Inclusiveness Matrix dy>0 dω>0 Unambiguously inclusive dy > 0, dω > 0: AB A1B1 dy>0 dω<0 Higher per capita income at the dy > 0, dω < 0: AB A2B2 B2 dy < 0, dω > 0: AB A3B3 expense of equity (could be dy < 0, dω < 0: AB B1 inclusive if the percentage change income per capita (y) A4B4 B in y > the percent change in ω dy<0 dω>0 Equity objective is achieved at the B4 A1 cost of average income B3 contraction A2 dy<0 dω<0 Unambiguously noninclusive A Source: Authors’ calculations. A3 A4 cumulative share of i = 100 population, 0 ≤ i ≤ 1 (when the entire population is covered) Source: Authors' illustration. Figure 2. Distribution of Emerging Markets on Inclusiveness Matrix average gdppc_ppp region 12 420 AFR AZE 5,000 APD 10,000 EUR 11 14,621 MCD WHD 10 ARM 9 CHN MNE 8 TJK BTN 7 growth in GDP per capita GEO LVA EST 6 VNM MOZ 5 BIH BLR RWA IND BWA GUY LKA KAZ 4 BGD HRV BGR PAN 3 IDN THA NPL BFA ETH MKD ARG JOR MLI 2 TZA BRA SWZ N AM KGZ BOL JAM CMR ECU MRT GMB 1 YEM SEN GTM MEX GIN NER PRY FJI VEN 0 KEN BLZ CIV ZMB MDG CAF -1 UZB UKR MWI TLS DZA -2 GNB BDI -3 2.5 2.0 1.5 1.0 0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 equity growth (omega) Source: Authors’ calculations. Note: The chart measures proportionate average annual change. The period used is the early 1990s to the latest available data. Size represents the initial size of the economy (GDP per capita), that is, the legend represents countries with purchasing power parity (PPP) GDP per capita below 420; 5,000; 10,000; and 14,621, respectively. Different regional codes are denoted by different colors. 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise ture the magnitude of the change in income distribution, this analysis uses a simple f the social mobility function by calculating an index (or social mobility index) from a under the social mobility curve: ��� in others. A decomposition of inclusiveness following table 1 ∗ � �� � � �� �� shows that there is a wide dispersion of possible outcomes. � These dispersions are depicted in figure 2, where most coun- tries fall into the two quadrants that show higher per capita The greater the y,* the greater the measure of inclusive income and a lower/or higher level of inequity, suggesting that eater the�� � ∗ ,growth. the greater If the the measure income of inclusive of everyone in thegrowth. population If the isincome the of everyone there in ∗ is no simple trade-off between growth and equality. In pulation is same (that is, same(that if income is, if income distribution distribution is is completely completely equitable), then � equitable), � will be most cases, economic growth shifted the indifference curves o�� . If �� ∗ is lower than �� , it implies that the distribution then y* will be equal to y. If y* is lower than y, it implies that of income is inequitable. So, upward, and there have been very few intersections of social viation of � � ∗thefrom �� is an indication of inequality in distribution of income is inequitable. So, the deviation of income distribution, which is mobility curves, suggesting unambiguous gains in inclusive- nted by�� (unlike Gini, a higher value of omega represents higher income equality). y* from y is an indication of inequality in income distribu- ness. However, the relative magnitude of shift and curvature ly, � is defined as follows: tion, which is represented by ω (unlike Gini, a higher value of of indifference 4 curves both matter. omega represents higher income equality). Formally, ω is de- 4 inclusive growth is primarily a growth story. Rap- China’s fined as follows: 4 id growth in per capita income has benefitted everyone, but � � ∗ the gains have been much greater for the rich (as depicted by � ��∗ � . � � ∗ � � � . �� negative value on equity [omega] and steepening of the social � � � �. mobility curve). Thus, in China, high growth has eclipsed the � For a completely equitable completely For asociety, ω = 1. Thus,equitable higher society, growing Thus, higher � � ��inequality value to produce of omega a large upward (closer shift of to theone) so- For a completely equitable society, � � �� Thus, higher value of omega (closer to one) value of omega For a completely equitable (closer to represents one) represents society, � higher higher �� Thus, �income income income equal- higherequality. value Rearranging, cialomega of mobility curve and (closer higher inclusive growth. A similar to one) ity. Rearranging, represents higher equality. Rearranging, represents higher income equality. Rearranging, story holds for India, where high growth has benefitted every- ∗ � � � � ∗ �� . one, but equity has gone down. On other hand, the increase in ��∗ � � ∗ � �. inclusiveness in Brazil, Mexico, Malaysia, and Thailand has ��∗ � � ∗ � �. ∗ Inclusive growth requires increasing � ,from come � which could growth be as achieved as well improvementby: (i) increasing in equity (positive ��, that is, Inclusive growth Inclusive growth requires requires increasing y ,* which increasing could be � ∗ , omega), � which but could be achieved growth has not by: been (i)enough fast increasing to �, that � benefit the is, Inclusive growth requires increasing increasing � �average ∗ , which income could through be achieved growth; by: (ii) (i) increasing increasing � � equity , that through �; or (iii) a is, achieved by: (i) increasing increasing y, that average income is, increasing through average in- growth; (ii) entire increasing population equity as much through as in China. �; or (iii) a increasing average income combination through growth; of (i) (ii)and (ii). Differentiating increasing equitythe throughthe �above equation: ; or (iii) a as a whole, the heterogeneity in come through growth;combination (ii) increasing of (i) and equity (ii). through Differentiating ω; or above Across equation: emerging markets combination of (i) and (iii) a combination of(ii). Differentiating (i) and (ii). Differentiating the abovethe ∗ above equation: economic growth performance and income distribution out- �� ∗ � � � ∗ �� � � �� ∗ � � equation: �� � � � ∗ �� � � �� ∗ comes � � provide insight to the growth-equity trade-off. Conven- �� � ∗ � � ∗ �� � � �� ∗� � tional wisdom suggests that ii growth comes at the price of ris- where ∗ �� � ∗ is the change in the degree of inclusive growth. ii Growth in is more inclusive if where �� � is the change in the degree ∗ of inclusive growth. Growth ing inequality, but regions is more differ inclusive their if growth-equity where ��� ∗ is the change ∗ �� in the � �. of inclusive � degree growth. ii Growth is trade-off. more In some inclusive instances, if strong growth has been achieved �� � � �. �� �∗ � �. d y is the change in the degree of inclusive growth.2 where without compromising equity. Growth is more inclusive if d y* > 0. Sources of Inclusive Growth Tracking the Evolution of Inclusive Growth While there is broad agreement on the basic policies impor- Tachieved RACKING THE EVOLUTION tant I OF NCLUSIVE for GROWTH growth and reducing poverty, little is known about Relatively few countries TRACKING have strong EVOLUTION THE inclusive growth. OF INCLUSIVE what may foster Ginclusive ROWTH growth. Rapid growth is unques- TRACKING THE E Previous studies VOLUTION have focused on the INCLUSIVE OFconvergence (or lackGROWTH Relatively few countries have achieved strong tionably inclusive necessary growth. Previous for substantial studies have poverty reduction (see Relatively few countries have achieved strong thereof) of the distribution of income across the world (see inclusive growth. Previous studies have Relatively Dollar few countries focused have achieved on the convergence strong [2006])inclusive (or lack growth. thereof) Kraay [2004] Previous of the and distribution Lopez and Servén of income [2004]), but studies haveof income across the world across the for inclu- world and Kraay focused [2003] on the convergence and Sala-i-Martin or the(or lack ris- thereof) of theto distribution focused on the convergence (see Dollar (or[2007]). lack andnote thereof) Kraay of the [2006] and Sala-i-Martin distribution sive growth of income [2006]) be sustainable across theor the rising level of inequality (see world in the long run, it should be ing level of inequality (see Dollar (see IMF and Kraay This [2006] sheds and Sala-i-Martin light [2006]) or the rising level (see of inequality (see on bothand (see Dollar Kraay [2006] IMF [2007]). and Sala-i-Martin This note sheds [2006]) light on both broad-based of those across aspects sectors and by mapping equitable out Berg the andchange Ostry in of those aspects IMF by[2007]). mapping out This the change note sheds lightor in inclu- on the rising both of those [2011a]). level aspects This of inequality is even by more (see mapping important out the because change some of in the IMF [2007]). siveness orThis note shedsinclusiveness light on or social mobility across countries over the last few decades. While cross- social mobility across inclusiveness or both countries social of over those the last mobility aspects few across by mapping countries key generally out the over determinants the change oflast few growth in decades. (for example, While cross- education, open- inclusiveness or social decades. While cross-country country mobility comparisons across comparisons countries of inequalityof over inequality the are last are few decades. plagued While by cross- problems of poor data reliability, country comparisons of inequality are generally ness, and plagued by problems financial depth) established ofinpoor data reliability, the literature (Barro country comparisons of lack inequality ofof coverage, are generally andplagued inconsistent lack of by problems methodology, this analysis relies on income distribution and Lee of poor data reliability, generally plagued by problems poor data reliability, lack of coverage, and inconsistent methodology, this 2000; analysis Dollar relies and Kraay on income 2003; Levinedistribution 2005) have coverage, and inconsistent lack of coverage, and inconsistent data from methodology, the thislatest analysis methodology, World relies this Bank on analysis Povcal relies database on income constructed distribution by Chen and Ravallion (2004) data fromthe the latest World Bank Povcal database constructed been associated by Chen with higher and Ravallion inequality (Barro 2000; (2004) IMF income data from thedistribution latest World for a large data from Bank latest Povcal number World database of emerging Bank Povcal constructed markets by Chen anda using more the Ravallion rigorous (2004) approach to filtering for a large number of emerging markets using 2007), a more thus rigorous begging approach question, to what filtering proximate factors database constructed by Ravallion for a large number of emerging individualand Chen markets income (2003) using and afor more consumption a large rigorous data for approach differences to filtering in quality than other commonly used number of emergingindividual income and consumption for differences datasupport in quality than other commonly used inclusive growth? individual income anddatabases. marketsdatabases. using a more rigorous approach consumption data for differences in quality Panel than other commonly regressions used measure of inclusive of the unique to filtering individual income and consumption data for dif- databases. growth on a broad sample of emerging markets provide in- ferences in quality than other The limitedused commonly gains databases. in inclusiveness are explained by relatively low growth in some The limited gains in inclusiveness sights are explained into bythe proximate relatively determinants low growth in of inclusive some growth. The limited gains in inclusiveness countries are and explained widening by rela- The limited tively lowgains growthinin inclusiveness countries some countries and are widening and explained widening byinequality inequality inequalityrelatively in low The in others. others. growth measure decomposition A in of A decomposition some growth, inclusive of ∂y,* is explainedfollowing orinclusiveness of inclusiveness by a set following table 1 shows that there is a countries and widening inequality in others. A decomposition of inclusiveness following wide dispersion of standard of possible control outcomes. variables used in These dispersions cross-country growthare table 1 shows that there is a wide dispersion of possible outcomes. These dispersions are table 1 shows that there is adepicted depictedwide in figureof in dispersion figure 2, where where most 2,possible countries outcomes. most countries These fall fall dispersions into the twoare into the quadrants that show higher per two quadrants that show higher per depicted in figure 2, where capita most income countries and falla lower/or into the twohigher level quadrants of inequity, that show suggesting higher per that there is no simple 4 POVERTY REDUCTION capita income AND ECONOMIC and a lower/or MANAGEMENT (PREM) higher NETWORK    level of inequity, suggesting that there is no simple www.worldbank.org/economicpremise capita income and a lower/or trade-off higher between level growth and of inequity, equality. suggesting In most that therecases, is economic growth shifted the no simple trade-off between growth and equality. In most cases, economic growth shifted the trade-off between growth indifference and equality. curves In most upward, cases, and economic there have growth been very shifted few the intersections of social mobility indifference curves upward, and there have been very few intersections of social mobility indifference curves upward, curves, suggesting and there have been unambiguous gains in inclusiveness. very few intersections However, the relative magnitude of of social mobility Table 2. Panel Regression: Emerging Markets Dependent variable: growth in inclusive growth 1 2 3 4 5 6 7 -0.211** -0.203* -0.300*** -0.468*** -0.605*** -0.528** -0.558*** Lag GDP per capita (logs) (0.0904) (0.107) (0.101) (0.139) (0.184) (0.198) (0.145) 0.397** 0.309* 0.120 0.261* 0.783*** 0.173 0.560** Education (0.149) (0.180) (0.205) (0.151) (0.280) (0.220) (0.235) 0.246** 0.194* -0.0442 0.418*** 0.223 0.0130 -0.00118 Trade openness (0.1000) (0.114) (0.120) (0.133) (0.240) (0.0970) (0.0943) -0.160 -0.164 -0.0390 -0.0176 -0.0822 0.112 -0.137 Credit to GDP (0.144) (0.174) (0.171) (0.186) (0.0946) (0.146) (0.184) -0.718 -0.340 -0.394 0.367 -2.849*** -0.00748 -1.250 Government consumption (0.866) (0.956) (0.731) (0.616) (0.571) (1.003) (0.905) 0.949** 1.030 0.945 0.786 -0.141 0.439 1.018** Investment (0.438) (0.646) (0.582) (0.650) (1.102) (0.781) (0.485) -0.0275* -0.0280* -0.0227* -0.0830*** -0.0524*** -0.00349 -0.00129 Inflation (0.0143) (0.0143) (0.0129) (0.0110) (0.00326) (0.00270) (0.00313) -2.126** -2.175** -0.991 0.223 -1.604 -0.788 -1.235 GDP volatility (1.065) (1.076) (0.875) (1.781) (2.065) (1.066) (1.042) 0.000547** Financial openness (0.000274) 0.0101*** FDI (0.00248) -0.718 ICT (0.432) -0.00245*** REER deviations (0.000779) 0.131*** Infrastructure quality (0.0385) 0.500*** Service export sophistication (logs) (0.165) 0.390* Goods export sophistication (logs) (0.216) 5.123** 4.899** 7.453*** 12.06*** 15.43*** 12.46** 5.816** Constant (2.167) (2.587) (2.443) (3.579) (4.370) (4.902) (2.573) Observations 261 234 234 111 98 139 146 R-squared 0.263 0.284 0.376 0.285 0.514 0.150 0.288 Number of countries 99 89 89 36 63 49 58 Source: Authors' compilation. Notes: Both country and time effects are included. *, **, and *** denote significance at the 10 percent, 5 percent, and 1 percent level, respectively. Robust t-statistics are in parentheses. Openness is trade openness; investments are fixed investments (percent of GDP), followed by government consumption (percent of GDP). Financial deepening is private sector credit to domestic sector (percent of GDP); education is Barro-Lee years of schooling above age 15; ICT is the total stock of ICT software- and hardware-related investments as a share of total capital stock; FDI is total FDI (liabilities) capital stock; infrastructure quality is the database developed by World Bank; inflation is consumer price index annual percentage change. Export sophistication uses United Nations Conference on Trade and Development Comtrade data for manufactured goods and balance of payments, IMF for service exports. and inequality literature in a non-over-lapping, unbalanced moderate inflation and output volatility, and a better-educat- five-year panel of 143 countries from 1970–2010. ed workforce have helped countries achieve more inclusive Consistent with results in Barro and Lee (2000) and Dol- growth. Foreign direct investment (FDI) has a significantly lar and Kraay (2003), table 2 shows that lower initial incomes positive impact on inclusive growth, as in IMF (2007), while (conditional convergence), trade openness, fixed investment, information and communication technologies (ICTs) in the 5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise total capital stock do not have a discernible impact; the latter fixed investment are important, while the role of technologi- could also reflect the lack of data on ICT investment in many cal change emphasized in the literature has a less discernible emerging markets and low-income countries. Financial open- impact. The weak data on technology and research and devel- ness more generally also shows a positive association with in- opment spending in emerging markets may explain the indis- clusive growth. Interestingly, financial deepening, measured cernible impact; however, these transitions in production ca- by the credit-to–gross domestic product (GDP) ratio, has a pabilities are mirrored through product and service negative impact, as in IMF (2007), but is not statistically sig- sophistication measures that foster inclusive growth. Moving nificant. This could be because inclusive growth encompasses up the value chain in both goods and services exports helps both the pace and distribution of growth, while previous find- foster inclusive growth, so the focus should not only be on ex- ings, such as Levine (2005), positively linked financial devel- port promotion, but the quality of exports, including servic- opment to growth, while IMF (2007) associated it with great- es. This could be even more important in the future, as tech- er inequity. nological changes are increasingly making service activities Structural transformation and moving up the value more productive, digitally tradable, and integrated across chain in both goods and services have also attracted lot of at- global supply chains. tention in terms of driving economic growth (Anand, Mishra, In terms of structural change and globalization, trade and Spatafora 2012; Hausmann, Hwang, and Rodrik 2007). openness and FDI foster inclusive growth, with a potentially In addition to modernizing manufacturing, the globalization positive role played by financial openness, which warrants of services is increasingly a driver of economic growth in further analysis. However, financial deepening could have a emerging markets (Mishra, Lundstrom, and Anand 2011). negative impact, as in IMF (2007), possibly related to its asso- The results in this note illustrate that countries that upgraded ciation with financial crises, although the impact is not statis- either manufacturing or service sophistication had higher in- tically significant. Macroeconomic stability is reinforced as a clusive growth. Sophistication in service exports, driven by key ingredient for inclusive growth. Drivers of connectivity, forces of globalization in computing and information net- business creation, and job growth measured as quality of in- works, seems to have a greater impact on inclusive growth. frastructure and competitiveness (REER deviations from The deviation of the Real Effective Exchange Rate PPP) are also important for inclusive growth. (REER) from its purchasing power parity (PPP)–implied lev- Looking forward, there are a number of unresolved is- el is negatively associated with inclusive growth, suggesting a sues and areas for future research. Many countries responded role for competitiveness. Infrastructure quality, as measured to the global financial crisis through large fiscal stimuli and/or by Calderon and Servén (2004) and Seneviratne and Sun bank bailouts, which are being withdrawn or met with auster- (2013), also plays a positive role in fostering inclusive growth, ity. The relationship between fiscal consolidation and inclu- possibly by reducing the cost of doing business and creating sive growth is an area worthy of study. The availability of employment.3 The importance of competitiveness through more granular data will be important to analyze the evolution indicators such as the deviation of the REER from its PPP- of inclusive growth at the national and subnational levels by implied level and infrastructure quality should be interpreted providing a local lens to view inclusive growth. Secondly, re- with caution given the limited observations. garding job creation, it will be important to understand the links between unemployment and labor market institutions What Should Policy Makers Take Away that foster inclusive growth; for example, the design of collec- and Do? tive bargaining programs and rights for workers might play This note quantifies and integrates two strands of literature to crucial roles in reaching inclusive growth goals in both ad- define inclusive growth. This approach is in line with the ab- vanced and emerging markets. Lastly, the speed of technologi- solute definition of pro-poor growth and goes beyond just fo- cal advancement, its reach and access, and the channels cusing on distribution issues. The integrated measure devel- through which it can foster or hinder inclusive growth, are oped in this paper is useful to delve deeper into the patterns additional areas for future research. Policy makers at national of and study the sources of inclusive growth. The methodol- and regional agencies, as well as global strategies such as pro- ogy here directly links the micro and macro dimensions of gram design for post-2015 Millennium Development Goals, inequality and growth to reflect both the pace and distribu- must target and track the evolution of inclusive growth. Lat- tion of income growth. est developments in technology, open data, and open govern- Macroeconomic stability, human capital, and structural ment initiatives may offer greater government transparency, changes are key determinants of inclusive growth in emerging economic capabilities, and civic participation. The aspirant markets. The standard economic growth drivers in the litera- for inclusion must keep growth as a primer, since inclusive ture, such as conditional convergence, education levels, and growth is about both the pace and distribution of growth. 6 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise About the Authors International Development Working Paper No. 042, Cam- bridge, Massachusetts. Rahul Anand is Senior Economist working on India at the Inter- Berg, Andrew, and Jonathan D. Ostry. 2011a. “Inequality and national Monetary Fund’s (IMF) Asia Pacific Department Unsustainable Growth: Two Sides of the Same Coin?� IMF Staff (APD). Saurabh Mishra is with the Research Department of the Discussion Note 11/08, Washington, DC. ———. 2011b. “Equality and Efficiency.� Finance & Development 48 (3). IMF working on jobs and growth, and previously also worked Birdsall, Nancy. 2010. “The (Indispensable) Middle Class in Devel- at the World Bank on economic policy, public expenditures, oping Countries; or, The Rich and the Rest, Not the Poor and and open governments. Shanaka J. Peiris is the IMF’s Resident the Rest.� Center for Global Development Working Paper 207. Representative to the Philippines and previously also worked in Calderon, C., and L. Servén. 2004. “The Effects of Infrastructure APD and Monetary and Capital Markets Department of the Development on Growth and Income Distribution.� World IMF, covering the Association of Southeast Asian Nations (ASE- Bank Policy Research Working Paper 3400, Washington, DC. Commission on Growth and Development. 2008. Growth Report: AN) and South Asia, respectively. Strategies for Sustained Growth and Inclusive Development. Wash- Notes ington, DC: World Bank. Dollar, D., and A. Kraay. 2002. “Growth Is Good for the Poor.� 1. Inclusive growth can also conceptually go beyond tradition- Journal of Economic Growth 7 (3): 195−225. al lines of poverty change and should also reflect changes in ———. 2003. “Institutions, Trade, and Growth.� Journal of Monetary Economics 50 (1): 3–39. the size and distribution of the middle class (Birdsall 2010). IMF (International Monetary Fund). 2007. “Globalization and 2. Inclusive growth is defined as the change in the social mo- Inequality.� In World Economic Outlook, chapter 4. Washington, bility index d y,* which is used here interchangeably. DC. 3. The lack of consistent historical (un)employment data in ———. 2011. “Regional Economic Outlook: Asia and Pacific.� Octo- emerging markets precluded the estimation of a link between ber, Washington, DC. Hausmann, Ricardo, Jason Hwang, and Dani Rodrik. 2007. “What employment and inclusive growth, as stressed by Ianchovi- You Export Matters.� Journal of Economic Growth 12 (1): 1–25. china and Gable (2012). Ianchovichina, Elena, and S. Lundstrom Gable. 2012. “What Is Inclusive Growth?� In Commodity Prices and Inclusive Growth References in Low-Income Countries, ed. by Rabah Arezki, Catherine Pat- Acemoglu, Daron, and David Autor. 2011. “Skills, Tasks and Tech- tillo, Marc Quintyn, and Min Zhu. International Monetary nologies: Implications for Employment and Earnings.� NBER Fund. Working Paper No. 16082, Cambridge, Massachusetts. Kakwani, N. 1980. Income Inequality and Poverty: Methods of Estima- ADB (Asian Development Bank). 2002. “Growth and Poverty: tion and Policy Applications. New York: Oxford University Press. Lessons from the East Asian Miracle Revisited.� ADB Institute Kanbur, Ravi, and Michael Spence (eds.). 2010. Equity in a Globaliz- Research Paper No. 33, Manila. ing World. World Bank. Aizenman, Joshua, Minsoo Lee, and Donghyun Park. 2012. “The Kraay, A. 2004. “When Is Growth Pro-Poor? Cross-Country Evi- dence.� IMF Working Paper No. 04/47, Washington, DC. Relationship between Structural Change and Inequality: A Levine, Ross. 2005. “Finance and Growth: Theory and Evidence� Conceptual Overview with Special Reference to Developing (chapter 12). In Handbook of Economic Growth, First Edition, Asia.� ADBI Working Paper No. 2012. ed. Philippe Aghion and Steven Durlauf, 865–934. North Ali, Ifzal, and H. Hwa Son. 2007. “Measuring Inclusive Holland. Growth.�Asian Development Review 24 (1): 11–31. Lopez, H., and L. Servén. 2004. “The Mechanics of Growth-Pover- Anand, Rahul, Saurabh Mishra, and Shanaka J. Peiris. 2013. “Inclu- ty-Inequality Relationship.� Mimeo, World Bank, Washington, sive Growth: Measurement and Determinants.� IMF Working DC. Paper No. 135, Washington, DC. Mishra, Saurabh, Susanna Lundstrom, and Rahul Anand. 2011. Anand, Rahul, Saurabh Mishra, and Nikola Spatafora. 2012. “Struc- “Service Export Sophistication and Economic Growth.� World tural Transformation and the Sophistication of Production.� Bank Policy Working Paper No. 5606, Washington, DC. IMF Working Paper No. 59, Washington, DC. Okun, Arthur. 1975. Equality and Efficiency: The Big Tradeoff. Wash- ———. 2013. “Inclusive Growth: Measurement and Determinants� ington, DC: Brookings Institution Press. IMF Working Paper No. 135, Washington, DC. Ravallion, M., and S. Chen. 2003. “Measuring Pro-Poor Growth.� Barro, Robert J. 1996. “Determinants of Economic Growth: A Economics Letters 78: 93−99. Cross-Country Empirical Study.� NBER Working Paper No. Sala-i-Martin, Xavier. 2006. “The World Distribution of Income: 5698. Falling Poverty and Convergence Period.� The Quarterly Journal ———. 2000. “Inequality and Growth in a Panel of Countries.� Jour- of Economics CXXI (2). nal of Economic Growth 5 (1): 5–32. Seneviratne, D., and Yan Sun. 2013. “Infrastructure and Income Barro, Robert J., and Jong-Wha Lee. 2000. “International Data on Distribution in ASEAN-5: What Are the Links?� IMF Working Educational Attainment: Updates and Implications.� Center for Paper No. 41, Washington, DC. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. They are produced by the Poverty Reduction and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at: www.worldbank.org/economicpremise. 7 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise