Was Growth Pro-poorest in the Arab Region in the 1990s and 2000s?

Dony El Costa, 07 Nov 2017

The visualization of consumption growth premiums,[i] defined as the differences between the growth rate in the consumption of a given percentiles group and the overall survey mean growth, leads to the conclusion that while the Arab Region might have gone through a roughly equal number of pro-poor(est) and anti-poor(est) episodes over the 1990s and 2000s, the poor's consumption appears to have been subject to high volatility, and the very bottom of the distribution might well have been left behind.  This has serious policy-implications and calls for directing poverty reduction efforts in the Arab countries towards sustaining the livelihoods and strengthening the resilience of the poorest of the poor.


The above finding is based on examining more than 30 household surveys[1] that selectively cover Arab countries in the 1990s and 2000s to conduct various poverty and inequality measurements. [ii] Selectively examining two tens of these surveys shows heterogeneous patterns and trends in inequality, and poverty incidence, depth and severity in the Arab countries (as Figure 1 portrays).[iii] Within a general trend of poverty reduction, the lowest levels of poverty are observed in the Maghreb countries (Algeria, Morocco, and Tunisia), followed by the Mashreq countries (Egypt, Jordan, Lebanon, Iraq and Palestine), and finally by the Least Developed Countries (Comoros, Djibouti, Mauritania, Sudan, and Yemen). It is worrying to see the depth and severity of poverty (captured by the size of the bubbles) increase with poverty incidence, meaning that those countries with the highest poverty rates are home to poor populations further down the poverty lines, with higher inequality between them). On the other hand, while various layers of inequality[2] can be identified, there appears to be no specific relation to the countries' development levels. In the 2000s for example, high inequality countries included the Comoros, Djibouti, Mauritania, Morocco and Tunisia; while Jordan, Palestine, and the Sudan showed medium levels of inequality; and Egypt, Iraq, Syria and Yemen appeared to be the most equal societies in the Region.


Figure 1: Poverty and inequality for selected Arab countries, selected years, 1988-2010

Source: Author's calculations based on household surveys sourced in the World Bank's PovcalNet database.


While growth in incomes is necessary for the reduction of monetary poverty, it is important to know which categories of the poor have benefited from that growth between two given survey dates, that is if growth has been pro-poor(est) or anti-poor(est).


In a weak absolute sense, growth is pro-poor if the poor's income or consumption level increases, independently of what happens to the incomes of the non-poor. 


The three main operational definitions of pro-poor growth in the literature can be easily extended to pro-poorest growth analysis.[iv] In a weak absolute sense, growth is pro-poor if the poor's income or consumption level increases, independently of what happens to the incomes of the non-poor. For growth to be pro-poor in a weakly relative sense, incomes of the poor need to grow faster than average growth rates, leading to a decline in inequality (in a relative sense) as the poor's share in total national income or consumption would be increasing. Finally, growth is pro-poor in an absolute strong sense if the gap between the poor and non-poor incomes decreases, which requires growth to be even more biased in favor of the poor, thus leading to a reduction in absolute inequality.


The growth incidence curve (GIC) is a useful intuitive tool to start with that gives important insights on the distributional pattern of growth or de-growth episodes. It summarizes the rate of change in income or consumption between two survey dates (or episodes), for the various percentiles of the distribution. The GICs for Egypt, Jordan, Mauritania, Morocco and Yemen over episodes overlapping the late 1990s and the 2000s illustrate differences within and across country shared prosperity developments,[v] and accordingly in the background, different poverty and inequality outcomes.


In further clarification, a GIC that lies above zero signals that poverty might well have decreased, with higher household consumption growth rates particularly in the bottom of the distribution leading to faster rates of poverty reduction. Growth would in this case be pro-poor in the weak absolute sense. Where this GIC is flat, this further signals equitable growth as this appears to have been the case for Jordan over 1997-2010 for example. When a GIC is decreasing (increasing), this signals pro-poor (anti-poor) growth in the weak relative sense as well as a decrease (increase) in relative inequality.


For example, growth in Egypt over 1990-1995, and somewhat less so over 2004-2008, was pro-poor, even though it did not benefit the very bottom of the distribution in that latter case; while over 1995-1999, growth was anti-poor in a relative sense (yet pro-poor in the weakly absolute sense with the exception of the bottom percentiles), or pro-non poor as some would put in the sense that the top of the distribution benefited relatively more from that growth. Yemen witnessed decreases in welfare from 1998 through 2005, while in Mauritania only the very bottom and very top of the distribution appear to have benefited from growth over 1995-2008.


Figure 2: Growth incidence curves, Egypt, selected years, 1990-2008

 Source: Author's calculations based on household surveys sourced in the World Bank's PovcalNet database.


Figure 3: Growth incidence curves for selected Arab countries, selected years, 1995-2010

Source: Author's calculations based on household surveys sourced in the World Bank's PovcalNet database.


As a final illustration (Figure 4), Morocco's GIC over 1998-2007 shows that as a result of a 2.5% average annual growth in mean consumption, and a distributional pattern of growth that appears to have been slightly pro-poorest (with the exception of the very lowest centiles), extreme poverty decreased from 7% to 3%, and poverty at the national poverty line decreased from 16.2% to 8.9%.


Figure 4: Growth incidence curve and selected poverty rates for Morocco, 1998-2007

Source: Author's calculations.


In a comprehensive analysis where I have examined growth pro-poorness for 32 episodes covering those Arab countries for which more than one survey is available over 1988-2010,[vi] I found that growth have been pro-poor (est) for approximately half of those episodes. However, the poor's consumption appeared to have been subject to high volatility, and the very bottom of the distribution, the bottom 5% population, might well have been left behind.


[1] There are concerns regarding the quality of the surveys, but these equally apply to all countries.

[2] Note that when we look at changes in inequality, this observation holds as far as "relative inequality" measures are used. When we use "absolute inequality", the picture is somewhat different.


[i] Also known as "shared prosperity premiums".

[ii] I have sourced the data forming the basis of these calculations, as well as all the below analysis, in the World Bank's PovcalNet database of standardized household surveys.

[iii] The Headcount poverty ratio measures the proportion of the population that is poor. The Palma ratio, is an index of income (or consumption) concentration that has been gaining popularity in recent years. It is defined as the ratio of the share of national income accruing to the richest 10% over the share accruing to bottom 40% of the distribution. In an abuse of language, I use it as an indicator of inequality here. For more on inequality in the Arab region, see the blog: "It is Absolute not Relative Inequality that Matters: a Fresh Look at Inequality in the Arab Region pre-Arab Spring". Finally, the Sen-Shorrocks-Thon index combines poverty incidence, poverty depth, and inequality among the poor.

[iv] See the pro-poor growth toolbox in Klasen, S. (2008). Economic growth and poverty reduction: Measurement issues using income and non-income indicators. World Development, 36, 420–445.

[v] On the shared prosperity concept, see the blog:" On the Likelihood of Arab Countries Achieving the Sustainable Development Goals Monetary Poverty Targets: Some Shared Prosperity Simulations".

[vi] Namely Algeria, Egypt, Jordan, Mauritania, Morocco, Palestine, Tunisia, and Yemen.


Dony El Costa is a poverty, energy and environmental economist with an 18-year professional experience in the consulting industry, academia and research centers, and the United Nations System since 2011, holding various positions and assignments notably with the United Nations Economic and Social Commission for Western Asia, the United Nations Convention to Combat Desertification, and the United Nations Development Program.


The views expressed here are solely those of the author in his/her private capacity and do not in any way represent the views of neither the Arab Development Portal nor the United Nations Development Programme. 

Dony El Costa Dony El Costa
  • You need to be logged in to post comments
No votes yet

Popular posts

Recent comments

  • No Comment was found.