Can Bangladesh continue to grow without ‘good governance’?

Selim Raihan

If we look at the growth pattern of Bangladesh from 1990, we discover two specific characteristics: first, the growth rate has been on the rise, and second, it is less volatile compared to those of many other countries (for example, India, Vietnam, Cambodia, China, Malaysia, Thailand and Ghana) which are known as ‘high growth performing countries’. Bangladesh’s growth experience has often been termed as ‘Bangladesh paradox’ given that the country has been able to perform well despite ‘weak governance’. Now, the big question is: can Bangladesh continue to grow without ‘good governance’? If we look over the last three decades, obviously, Bangladesh had been growing without the so-called ‘good governance’. Then what does this ‘good governance’ mean?

Four contemporary analytical approaches can be linked to the discussion on ‘good governance’. The new institutional economics (contemporary lead presenters are Daron Acemoglu and James Robinson), representing a variant of the neo-liberal orthodoxy, argue for specific and well-defined rules and property right systems (‘good governance’) for economic growth. There are three alternative approaches to this new institutional economics. The approach by Douglass North, Joseph Wallis and Barry Weingast argues for ‘limited access order’ in a large number of developing countries in contrast to ‘open access order’ in the advanced economies. In ‘limited access orders’, political elites divide up control of the economy, each getting some share of the rents; and since outbreaks of ‘violence’ (conflicts among the elites) reduce the rents, the elite groups have incentives to reduce conflicts among them. The approach by Mushtaq Khan stresses on ‘political settlement’, which highlights on the relative holding of the power of different groups and organizations contesting the distribution of resources, and a ‘political settlement’ emerges when the distribution of benefits supported by its institutions is consistent with the distribution of power in society. Mushtaq Khan also emphasizes on ‘growth-enhancing governance’ (un-orthodox institutional arrangements) in contrast to ‘market-enhancing governance’ (orthodox institutional arrangements, as signified by new institutional economics). Finally, the approach by Lant Pritchett, Kunal Sen and Eric Werker emphasizes on ‘deals space’, ‘rents space’ and ‘political settlements’ for growth acceleration and growth maintenance in developing countries. The rents space is characterized by private sector firms who can be rentiers (securing rent from the export of natural resources), powerbrokers (securing rent from the regulated domestic market), magicians (firms participate in competitive export markets), and workhorses (firms participate in unregulated domestic markets). Deals, in contrast to rules, among the political and economic elites, can be open (access is open to all) or closed (access is restricted); and also they can be ordered (deals are respected) or disordered (deals are not respected). The countries are likely to exhibit high growth when deals are open and ordered.

Can we explain the growth experience of Bangladesh through these four approaches? The approach by new institutional economics cannot explain the growth of Bangladesh, since Bangladesh never had the so-called ‘good governance’ but the economy continued to grow. Furthermore, all these approaches have three major problems. First, approaches of ‘limited access order’ and ‘political settlement’ emphasize more on the ‘elite agreement’ at the macro level, thus ignore the perspectives at the sectoral level. However, the ‘deals-rent space’ approach has a better holding on the sectoral level analysis. Second, all these approaches emphasize on the process of ‘elite agreement’ rather than on the outcome, which does not convincingly show how such process affects economic growth. Third and most importantly, all these approaches emphasize on ‘elite agreement’, and overlook the critical nexus between elites and non-elites within the society. Only in ‘limited access order’ approach, such nexus is shown through the ‘power of violence’ of non-elites.

Empirical research suggests that there are four major drivers of growth in Bangladesh: exports of readymade garments (RMG), remittances, growth in agriculture, and microfinance.  Now, it is clear that we cannot explain these growth drivers of Bangladesh with the usual definition of governance or politics by the aforementioned four approaches.

From a political economy perspective, in my view, there must be some substances by which these growth drivers are fueled; and I want to name these substances as ‘political capital’. The usual meaning of ‘political capital’ is a feeling of trust that politicians build among the common people through which they exert their influence in the society. But, according to my opinion, ‘political capital’ is an outcome of agreements among the political elites and support from the non-elites on such agreements over certain growth drivers in the economy. In order to source such support, elites ensure some critical benefits for non-elites. Over the last three decades, Bangladesh has been able to generate crucial stock and flow of ‘political capital’ in favor of the aforementioned growth drivers. Bangladesh is not rich in natural resources, which did not help to generate substantial rents for the political elites. Elites, thus, found the RMG sector as a source of generation of rents, and they were able to draw support from the non-elite through the creation of large-scale employment opportunities in the RMG sector. In the case of remittances, international migration of a large number of people helped alleviation of poverty, and thus gathered support from the non-elites. For the agricultural sector, this ‘political capital’ is generated from the experience of the 1974 famine, as the political elites realized that the country like Bangladesh cannot afford anything like this in the future. Therefore, subsequent governments, focused on the development of the agricultural sector to ensure food security. Finally, as microfinance, another example of elite and non-elite nexus, played important roles in generating growth and alleviating poverty in Bangladesh, there had been a construction of significant stock of ‘political capital’ around microfinance over the last three decades.

Therefore, Bangladesh can continue to grow until the ‘political capital’ provides returns over the existing drivers of growth. Given the fact that there are growing challenges for these existing drivers, political elites in Bangladesh also need to find new drivers for growth acceleration. There are two new prospective drivers, for which critical ‘political capital’ is yet to be formed. The first one relates to the comprehensive economic and trade integration with neighboring countries, and the second one is government’s initiative of setting up 100 special economic zones (SEZs) by 2030 for rapid industrialization of the country through large-scale domestic and foreign investments. It is a high time that political elites in Bangladesh come out from their comfort zone of old drivers towards the journey of building ‘political capital’ for new drivers.

Dr. Selim Raihan. Executive Director, SANEM. Email:


Cross-country differences in income inequality: Where do South Asian countries stand?


In recent years, there has been a growing interest among general people, researchers and policy makers in income inequality, its causes, and its effects. The most popular index of income inequality is the ‘Gini index’ which measures the inequality among levels of income of the people of any country. A Gini coefficient of zero means perfect equality, where everyone has the same income, and a Gini coefficient of 1 (or 100%) expresses maximum inequality.

For meaningful comparisons among different countries with respect to their levels and trends in income inequality we need comparable data. National surveys on households’ incomes and expenditures in different countries provide data on the Gini index of these countries for some years. However, we are not in a position to use these data for cross-country comparisons due to various reasons. In those surveys there are differences in the population covered, differences in coverage on geography, age and employment status, differences in the definition on welfare (whether to use market income or consumption data), differences in the use of equivalence scale (whether to use household per capita or household adult equivalence), and differences in the treatment of various other items, such as non-monetary income and imputed rents. The Standardized World Income Inequality Database (SWIID), introduced in 2008, provides a dataset on income inequality that facilitates comparability for the largest possible sample of countries and years. A custom missing-data algorithm is used to standardize data on cross-country income inequality using the data from national surveys (Solt, 2016). Using the SWIID database, the World Economy Database (WED) version 9.1 has generated a time series database on the “Gini index” for 207 countries over the period between 1970 and 2015 by filling missing observations with the help of different estimation methods.

Using the WED 9.1, we have produced a scatter plot diagram with data on Gini indices for 207 countries in 1980 in the horizontal axis and data on Gini indices of the same countries in 2015 in the vertical axis. In the scatter plot, dots around the 45 degree line are the countries with ‘no or very small’ changes in Gini indices during 1980-2015; dots above the 45 degree line are the countries which experienced an increase in the Gini index; and finally, dots below the 45 degree line are the countries which experienced a decline in the Gini index. Out of those 207 countries, 18 experienced ‘no or very small’ changes in Gini indices, 109 experienced increases and 80 experienced declines. Among the 8 south Asian countries, 5 countries (Afghanistan, Bangladesh, India, Pakistan and Sri Lanka) observed rises while the rest 3 countries (Bhutan, Maldives and Nepal) experienced declines. We also brought China and South Korea into the picture, and it appears that the Gini index in China increased quite astonishingly, whereas that of South Korea declined.

We have also categorized the values of Gini index as follows: a Gini index value lower than 30 is considered low; an index value between 30 and less than 40 is considered medium; an index value between 40 and less than 50 is considered high; and an index value above 50 is considered very high. Depending on these classifications, we can observe some interesting movements of the South Asian countries during 1980 and 2015. Afghanistan moved from a status of low inequality to medium inequality; Bangladesh moved from medium inequality to high inequality; though Nepal, Pakistan and Sri Lanka remained within the medium inequality range, Sri Lanka was at the border of high inequality; India moved from high inequality to very high inequality; and both Bhutan and Maldives moved from very high inequality to medium inequality. In comparison, China moved from low inequality to very high inequality, whereas South Korea moved from medium inequality to very close to low inequality.

We also explored the factors affecting inequality in the cross-country and over time contexts. Results from a fixed effect panel regression suggest that while rise in the real GDP per capita tends to have a small negative association with the Gini index, an increase in both life expectancy at birth and net secondary school enrollment are strongly associated with the decline in the Gini index. These suggest that, an increase in per capita real GDP is not a guarantee for the reduction in income inequality, whereas investment in social infrastructure with the aim of raising the life expectancy at birth and a rise in secondary school enrollment can be very instrumental in reducing income inequality.

Reference: Solt, F. (2016). “The Standardized World Income Inequality Database”. Social Science Quarterly.

First published at the Thinking Aloud on 1 September 2016

Published at The Daily Star on 1 September 2016

Dynamics of economic growth in Bangladesh

Selim Raihan and Wahid Ferdous Ibon

Rapid and sustained economic growth is very critical for Bangladesh economy in its way towards a middle income country. In this article, we have investigated the major determinants of economic growth in Bangladesh using time series data for 44 years (1972-2015). We start with a production function approach, which incorporates the features of neo classical and new-growth theories. Subsequently, we have investigated the impacts of trade policies, fiscal policies, FDI, interest rate, inflation, infant mortality rate, enrolment in secondary education, infrastructure and institution on growth in Bangladesh’s real GDP (gross domestic product). A new database (World Economy Database, version 9.1) has been used, which is complemented by data from the Peen World Table (PWT8.1) and World Bank’s World Development Indicators (WDI). Most of the variables under consideration are found to be non-stationary (integrated of order one). Two non-stationary time series may lead to a spurious relationship between them if they are not co-integrated. Therefore, we checked for the possibility of co-integrating relationship, using the Johansen co-integration test, and found at least one co-integrating relationship in all the regressions, which confirms that the long run estimates show causal relationships. We ignore bi-directional causality in the regression model, as this is not what we want to explore in this analysis.

The basic production function

With the aim of identifying the determinants of economic growth in Bangladesh, we start with a Cobb-Douglas production function. Along with employment and physical capital stock, we have incorporated human capital into the production function. We multiply the data on human capital with employment data to create the human capital adjusted employment variable. The regression results suggests that, in the long run, on average, one percent increase in the human capital adjusted employment leads to 0.25% increase in the real GDP. Furthermore, one percent increase in the physical capital stock leads to 0.12% increase in the real GDP. As the variables of the production function are co-integrated, there must be an Error Correction representation which shows the short run adjustments of the variables under consideration if there is any deviation from the long run equilibrium relationship. Error Correction term is -0.0197 which is statistically significant, negative and less than unity, as expected. About 1.97% error is thus being corrected each year following any deviation from the long run equilibrium.

Secondary school enrolment helps

There are both theoretical and empirical literature which provide evidence that the educational level and its quality are important causal determinant of income, both at the individual and national levels. A highly educated labor is more productive relative to his/her less educated counterpart, and this increased labor productivity helps a nation grow faster. Education is a key component of human capital. In terms of the net secondary school enrolment, though Bangladesh made a progress during 1972 and 2015 from around 16% to 52%, still there is a need for substantial further improvement. Here, we have investigated the effect of the net enrolment in secondary school on real GDP and have found positive effect, as expected. One percentage point rise in the net secondary school enrolment ratio leads to, on average, 0.013% increase in the real GDP.

Reduction in the infant mortality rate helps

Bangladesh has shown its capacity to reduce infant mortality rate rapidly over the past four decades. Among 1000 live births, the rate came down from 148 in 1972 to 30.7 in 2015. In the regression, the infant mortality rate appears with a negative and significant coefficient. On average, one point reduction in the infant mortality rate contributes to the rise in real GDP by 0.01%.

Greater trade-orientation promotes growth

Theoretically, trade liberalization results in productivity gains through increased competition, efficiency, innovation and acquisition of new technology. Trade policy works by inducing substitution effects in the production and consumption of goods and services through changes in prices. These effects, in turn, change the level and composition of exports and imports. In particular, the changing relative prices induced by trade liberalization cause a re-allocation of resources from less efficient to more efficient uses. Trade liberalization is also thought to expand the set of economic opportunities by enlarging the market size and increasing the effects of knowledge spill over.

Since its independence, Bangladesh underwent a variety of trade policy reforms, which resulted in the rise in trade-GDP ratio, import-GDP ratio and export-GDP ratio from 10.6%, 6.5% and 4.1% respectively in 1972 to 41.7%, 23.3% and 18.4% respectively in 2015. To identify the growth effects of these three trade-orientation variables, we incorporated them into the production function through three separate regressions. The regression results indicate that, these variables are statistically significant with positive signs. One percentage point increase in trade-GDP ratio, import-GDP ratio, and export-GDP ratio account for, on average, 0.014%, 0.023% and 0.029% increase in the real GDP respectively.

Larger FDI-orientation propels growth

Foreign direct investment (FDI) is another driver of economic growth, particularly for a least developed country (LDC) like Bangladesh. FDI contributes to transfer the technical knowhow from advanced countries to the less developed countries. In 2015, the FDI inflow in Bangladesh was only US$ 2.2 billion which was about 1% the GDP, whereas government, as stated in the 7th five year plan, aims to achieve a level of FDI inflow of US$ 9.6 billion by 2020. In the regression, the coefficient of the FDI-GDP ratio is found to be statistically significant and positive, as expected. One percentage point increase in the FDI-GDP ratio leads to the rise in real GDP, on average, by 0.12%. In order to attract more FDI, there is a need to maintain political stability, improvement in infrastructure and reduction in the cost of doing business. The planned 100 special economic zones, if they are implemented successfully, can be helpful in attracting FDI.

Positive effect of government transfer payments

The regression result confirms a positive significant impact of government transfer (social security payments, safety net programs, pension payments etc.) on the rise in real GDP in Bangladesh economy. On average, one percentage point rise in the ratio of government transfer to GDP leads to a rise in real GDP by 0.05%.

Reduction in lending interest rate helps

Interest rate is the price of fund that private investors lend from the banks. Therefore, more private investment takes place following a reduction in lending rate, which in turn promotes economic growth. This is evident from our regression analysis that one percentage point reduction in the lending rate, on average, increases real GDP by 0.03%.

Inflation hurts growth

Rise in the general price level hurts Bangladesh’s growth. An increase in the price level decreases the real wage earned by the laborers. This lower real wage is followed by a lower aggregate private consumption demand, which in turn affects national income badly. Our regression analysis suggests, one point increase in consumer price index accounts for, on average, 0.001% reduction in real GDP.

Infrastructure promotes growth

Infrastructure is a key ingredient for high and sustained economic growth. Better infrastructure helps total factor productivity to rise by lowering transaction cost and a more efficient use of inputs of production. Due to the lack of time-series data on different dimensions of infrastructure, here we consider total number of mobile and fixed line telephone subscriptions as a proxy for infrastructure. In the regression analysis, we find that one percent increase in total telephone subscription results in, on average, 0.12% rise in real GDP.

Quality of institution matters

We have considered an index of institution in the regression. We have constructed the index of institution using the data of six major ICRG ( variables, namely bureaucracy quality, control of corruption, investment profile, democratic accountability, government stability, and law and order. As values of these six ICRG variables have different scales, we have rescaled them between 0 and 10. The aggregate institution index is the average of these six indicators with the range between 0 and 10, where 0 and 10 respectively indicate the lowest and highest levels of quality of institution. In 1980, the index value was 2.15, which increased to 5.5 by 2015. The regression suggests a positive significant role of institution on real GDP in Bangladesh. On average, one point rise in the institution index leads to the rise in real GDP by 0.05%.

What do we learn?

The analysis in this article suggests that, for further economic growth acceleration in Bangladesh, there is a need for reforms in economic policies and institutions, investment in infrastructure, and making most of the demographic dividend through investment in public health, education, and human capital development. All these will require increased domestic private investment and FDI targeting broader economic and export diversification. Reform of economic and political institutions for efficiency gains is critically important.

Transitions between growth episodes: Do institutions matter and do some institutions matter more?

Selim Raihan, Sabyasachi Kar and Kunal Sen

A large literature has examined the role of institutions in explaining economic growth. While the earlier literature has examined the role of institutions in determining long-run per capita income, a new literature examines the determinants of growth accelerations and deceleration episodes – which are large discrete changes in medium term growth rates that are common in developing countries. Some of these studies examine the onset of growth accelerations while others examine the onset of growth decelerations. However, these studies look at only the timing of the shift in the growth rate (either as an acceleration or a deceleration), and the econometric methodology they use are probit models (where the year of the break is taken as one, with other years as zero) to study the likelihood a growth break occurring in a given year, for a set of correlates. An important limitation of these studies is that they do not differentiate between the different growth episodes that a country is transitioning from or to. For example, when a country moves from a growth collapse to rapid growth, it is a different growth transition qualitatively than when it moves to an episode with slightly positive but slow growth rates.

In this paper, we investigate the role of economic and political institutions in determining the likelihood of a country transitioning from one growth episode to another. In contrast to the previous literature, in this paper, we provide a richer characterisation of the growth process where a country may move between six different types of growth episodes, ranging from growth collapses to rapid growth episodes. By doing so, we are better able to capture the episodic nature of growth and that many countries tend to switch frequently between growth collapses to slow growth episodes to rapid growth episodes.

We differentiate between six types of growth episodes – from growth collapses (where the episode specific per capita real GDP growth rate, g, is -2 per year), to negative growth (g between -2 and 0), stagnation (g between 0 and +2), stable growth (g between +2 and +4), moderate growth (g between +4 and +6), and rapid growth (g over +6). Using multinomial logit models, in the context of a panel dataset of 125 countries from 1984 to 2010, we examine the likelihood of switching from one growth episode to another growth episode. We examine the role of contract viability (as a measure of the quality of economic institutions) and the role of democracy and bureaucratic quality (as measures of political institutions) in explaining the switches that countries experience between different types of growth episodes. The data on contract viability, democracy and bureaucratic quality are derived from the ICRG database (

We find that though bureaucracy quality has a positive effect while switching from negative growth episode to positive growth episodes, it doesn’t matter in most of the cases while switching from lower order growth episodes to higher order growth episodes. Both contract viability and democratization can explain the switching from negative growth episode to positive growth episodes. Contract viability and democracy can also explain the movements from lower positive growth episodes to higher positive growth episodes. However, while contract viability is important for moving from stable growth episode to rapid growth episodes, democracy is not important in explaining this switch. This suggests that while better economic and political institutions matter in taking a country from growth collapses to stable growth, economic institutions matter more than the political institutions for the switching from stable growth to rapid growth.

Our results suggest that, democratic episodes do not necessarily witness transitions to rapid growth episodes from moderately positive growth episodes. However, democratic episodes do witness a transition from negative to positive growth episodes, indicating that democratization does prevent the worst type of growth episode that a country can experience. We also find that improving state capacity in the form of the quality of the bureaucracy can help in taking a country out of negative growth episodes but that higher state capacity does not increase the likelihood of rapid growth episodes. This finding suggests that previous research that has found a positive role of bureaucratic quality in fostering economic growth need to differentiate between phases of growth, and that the relationship between bureaucratic quality and economic growth may not be monotonic.

We find that the most important institutional determinant of switching to higher order growth episodes from lower ones, and in particular, to rapid growth episodes, is the nature of property rights institutions – that is, the extent to which investors trust the viability of contracts. In contrast to the previous literature on the determinants of growth accelerations, we find that not only does institutional quality matter in bringing about a growth acceleration, it is the case that the greater the quality of property rights institutions, the higher is the likelihood of a transition to a rapid growth phase.

Our findings have clear policy implications. For a country in a growth decline or collapse, it is important to stress improvements in both political and economic institutions, such as bureaucratic quality, viability of contracts and democratization to move into an episode of positive growth. However, once the country is in a stable or moderate positive growth episode, further movements into rapid growth episodes need larger emphasis on improving the quality of property rights institutions than enhanced democratization or state capacity. Economic institutions trump political institutions in bringing about rapid growth episodes, though they both matter in reversing growth collapses.

Dr. Selim Raihan (Professor, Department of Economics, Dhaka University, Bangladesh; Email:, Dr. Sabyasachi Kar (Research Fellow, Institute of Economic Growth, Delhi, India), Dr. Kunal Sen (Professor, IDPM, University of Manchester, UK)

First published at the Thinking Aloud on 1 August 2016

Published at The Financial Express on 2 August 2016

Published at ESID blog on 1 August 2016



Why do countries differ in total factor productivity?

Theoretical and empirical literatures on sources of economic growth emphasized on factor accumulation and factor productivity as two major sources of growth. Though factor accumulation can explain a significant part of economic growth, it can’t explain the sustained long run economic growth, as sustained long run economic growth is attributable to growth in productivity. Productivity is the cornerstone of economic growth. Increases in productivity allow firms to produce greater output for the same level of input, and thus result in higher Gross Domestic Product (GDP).

We should make clear the difference between labor productivity, which is output per worker, and Total Factor Productivity (TFP), which is the ‘ability’ with which all factors are combined to produce outputs. TFP is the part of output which is not explained by the amount of inputs used in production. Essentially, its level is determined by how efficiently and intensely the inputs are used in the production process. TFP growth is usually measured by the Solow residual. TFP plays a key role in economic fluctuations, economic growth and cross-country differences in per capita income.

The scatter-plot using the data for 110 countries in 2011 shows a very interesting association between TFP and log of per capita GDP. The TFP data are derived from the Penn World Table version 8.1 (PWT 8.1) with some required adjustments and extensions. Here the TFP level of USA is considered as 1 and other countries’ TFP levels are indexed against USA’s TFP level. For example, among the South Asian countries, the TFP levels of Bangladesh, India, Nepal, Pakistan and Sri Lanka in 2011 were 0.15, 0.48, 0.11, 0.28 and 0.42 respectively. Similarly, TFP levels of Malaysia and Thailand were 0.47 and 0.65 respectively. Singapore’s TFP level (1.1) was higher than that of USA. The trend line shows a very strong positive association between TFP and log of per capita GDP (the correlation coefficient is 0.9). Nepal and Bangladesh, though on the trend line, are at the lower end of the association. A straightforward policy suggestion for these countries is that a rise in the TFP level is required to raise their per capita GDPs.


Why do countries differ in TFP? How to improve the level of TFP? In order to answer these questions, we have run fixed effect cross-country panel regressions for 110 countries for the period 1995-2011 considering log of TFP as the dependent variable. The explanatory variables include log of human capital (an index of human capital per person which is linked to the average years of schooling and the return to education), log of public expenditure on health as % of GDP, and log of trade-GDP ratio. The data source of human capital is the PWT 8.1, and the data of health expenditure and trade-GDP ratio are taken from the World Bank WDI. The logic behind the formulation of this model is that we want to explore how cross-country differences in statuses of education, health and openness affect the cross-country differences in the TFP. The regression results suggest that all three explanatory variables are statistically significant with expected signs. 1% rise in the human capital index is associated with 0.39% rise in the TFP. Also, 1% rise in the ratio of public health expenditure to GDP is associated with 0.03% rise in the TFP. Finally, 1% rise in the trade-GDP ratio is associated with 0.03% rise in the TFP.

In extended models, we have found that the ratio of FDI to GDP is positively and significantly associated with the TFP. 1% increase in the FDI-GDP ratio is associated with 0.01% rise in the TFP. Furthermore, institutional variables like bureaucracy quality and investment profile (from the PRS database) are positively associated with the TFP with statistical significance. Improvements in the bureaucracy quality and investment profile by one unit are associated with the rise in TFP by 0.03% and 0.01% respectively.

The aforementioned analyses point to some very obvious policy suggestions. Countries like Bangladesh need to attach decisive emphasis on improving their currently low levels of human capital. This can happen by enhancing investing on education and health quite a lot in order to increase the efficiency in using inputs in the production process thus raising the level of TFP. Also, larger trade and FDI orientations and improvement in the quality of institutions are indispensably important.

Published at the Thinking Aloud on 1 May 2016

Published at The Financial Express on 10 May, 2016

Is Bangladesh poised for leaping towards a middle income country?

Bangladesh has recently been upgraded from low income country (LIC) to lower-middle income country (LMIC) as per the World Bank’s classification. There is an aspiration for graduating from the LDC status to that of a middle income country by 2021 as per the United Nations’ classification. In this context, it is very important to assess the capacity of the country’s social infrastructure in achieving the desired level of economic growth rate and subsequently the targeted per capita income level.

As far as the social infrastructure, in terms of education and health, is concerned, there is a valid question to ask, whether Bangladesh is poised for leaping towards a middle income country. In this paper, we focus on two indicators, public expenditure on education as % of GDP and public expenditure on health as % of GDP. It is well established in economic literature that strong bases of education and health are needed for a country to accelerate and sustain economic growth.



In Figure 1 and Figure 2, we have presented two scatter-plots involving public expenditure on education as % of GDP and public expenditure on health as % of GDP respectively against the log of GDP per capita for the lower-middle income countries and upper middle income countries as per the World Bank’s classification. The data are the averages for the years between 2010 and 2014 and they are derived from World Bank’s WDI. In both scatter-plots, the horizontal line at the level of 8 of log GDP per capita separates lower-middle income countries from upper-middle income countries. Bangladesh’s position is shown by the red circle.

Even these simple scatter-plots present some very interesting insights. In Figure 1, the trend line suggests that the association between public expenditure on education as % of GDP and log of GDP per capita for 80 countries (involving both lower-middle income and upper-middle income countries) is positive. This evokes that countries with a higher public expenditure on education as % of GDP tend to be associated with a higher GDP per capita. It is interesting to see that all upper-middle income countries are well above the trend line; whereas, most of the lower-middle income countries are well below the trend line while some of them are just on the trend line. Bangladesh’s position is among the worst performing countries, as its public expenditure on education as % of GDP (around 2%) is much lower than those of all upper-middle income countries, and is even much lower than most of the lower-middle income countries. Similar pattern is observed, as depicted in Figure 2, in case of the association between public expenditure on health as % of GDP and the log of GDP per capita. Once again, Bangladesh appears to be among the worst performing countries with only 1.2% share of public expenditure on health in GDP. Interestingly, the positive association between health expenditure and GDP per capita (Figure 2) seems to be stronger than that of between education expenditure and GDP per capita (Figure 1), which is supported by empirical literature that health expenditure has a relatively more immediate positive effect than education expenditure on economic growth.

The aforementioned exercises suggest that Bangladesh has to improve social infrastructure significantly in her way towards a middle income country. There are two critical lessons for Bangladesh. First, the current allocations for education and health expenditure as % of GDP need to be almost doubled from their current meagre levels (China and Malaysia, depicted in Figures 1 and 2, are good examples). Second, the efficiency of the public expenditure on education and health in Bangladesh needs to be improved. There are some countries (both in categories of lower-middle income and upper-middle income) who have been able to achieve much higher levels of GDP per capita than that of Bangladesh even with similar proportions of education and health expenditures that Bangladesh has.

Published at the Thinking Aloud on 1 March 2016

Published at The Daily Star on 6 March 2016

How does improvement in infrastructure affect economic growth?

Selim Raihan and Sunera Saba Khan

Infrastructure plays a decisive role in stimulating long-run economic growth. An increase in the level of infrastructure stock directly helps in reducing poverty and accelerating productivity. Infrastructure also contributes to the development process through the provision of intermediate consumption items for production as well as final consumption services for households. It contributes to growth through generating new jobs, creating cohesive spillover benefits and attracting further investments through crowding in effects. Empirical studies also corroborate the relationship between different infrastructural indicators and growth.

In the present article, we have constructed an Infrastructure Index to observe the growth-infrastructure nexuses from a broader perspective. With a view to observing nexus we have constructed the Infrastructure Index for 133 countries over the period between 1990 and 2012 using four indicators namely Electric Power Consumption (per kWh per capita), Energy Use (kg of oil equivalent per capita), Fixed Broad Internet Subscribers (per 100 people) and Mobile Cellular Subscriptions (per 100 people). The indicators are selected based on the availability of data and importance. We have obtained the data of these selected indicators from the World Development Indicators (WDI) of the World Bank. In order to assign weight to each indicator to construct the Infrastructure Index we have applied the Principal Component Analysis (PCA) method as it enables to derive the weight for each variable associated with each principal component and its associated variance explained. In doing so, firstly, we have used normalized values of variables followed by the extraction of factors. Secondly, the Eigen values of the factors, which help to determine the significance of principal components, have been used to determine the factors that will be retained. Thirdly, the variables have been assigned weights, which have been calculated by multiplying factor loadings of the principal components with their corresponding Eigen values. And, finally, the index has been constructed using those weights. The constructed Infrastructure Index ranges from 0 to 100 where 0 depicts the worst case and 100 depicts the perfect case. The PCA suggests that the weights for electric power, energy use, internet use, and mobile subscriptions were 29.9%, 37.6%, 16.3% and 16.2% respectively in 1990; 30.1%, 36.5%, 15.8% and 17.6% respectively in 2000; and 31.4%, 33.8%, 19.4% and 15.4% respectively in 2010.

Table 1

Table 2

Table 3

Tables 1, 2 and 3 depict the top 10 and bottom 10 countries in terms of the Infrastructure Index for the years of 1990, 2000, and 2010 respectively. Norway ranked at the top in 1990 while Iceland ranked at the top in both 2000 and 2010. Among the 133 countries considered, Bangladesh ranked the lowest invariably in both 1990 and 2000 whereas, Ethiopia ranked the lowest in 2010. The ranking of the South Asian countries (Table 4) shows that Pakistan ranked 110th in 1990, the highest among the five South Asian countries; while Sri Lanka ranked 108th and 103rd respectively in the following two consecutive decades, the highest among the five South Asian countries. It should be noted that the South Asian countries’ rank as some of the bottom most countries clearly indicate dissatisfactory performance in their infrastructure development. This poor performance clearly depicts that the region has huge electricity shortages and very low energy use, which together takes into account more than 60% weight of the Infrastructure Index.

Table 4

In order to explore the association between infrastructure and economic growth we have run a series of fixed effect panel regressions where Infrastructure Index and its sub-components are treated as infrastructure capital. We have followed the production function approach in the cross-county growth regressions where aggregate output Y at time t is produced using other capital, infrastructure capital and labor. Our data covers the time period between 1990 and 2011 and we have a balanced panel data set. We have chosen a long panel over other models as infrastructure is expected to have a long-term effect on growth. Output is measured as real GDP at constant 2005 national prices (in million 2005 US$), other capital is measured as capital stock at constant 2005 national prices (in million 2005 US$), and labor is measured as the number of persons engaged (in millions). The data of real GDP, capital stock and labor is obtained from the Penn World Table 8.1. We have taken natural logarithm for all variables except the infrastructure variables.

We have carried out five individual sets of fixed effect regressions. The first set of regressions included real GDP, the Infrastructure Index, capital stock, and labor. The result shows strong, statistically significant and positive relationship of labor, capital stock, and Infrastructure Index with real GDP: a 10% increase in labor supply increases real GDP by 3.5%; a 10% increase in capital stock increases real GDP by 6.2% while a 10 unit increase in the Infrastructure Index raises real GDP by 1%. Analogous to the first set of regressions, in all of the successive regressions, after controlling for capital stock and labor, we find a highly significant influence of sub-components of Infrastructure Index over real GDP growth. It is observed that, a 10 unit increase in the electric power consumption raises real GDP by 1.3%; a 10 unit increase in the energy use raises real GDP by 1.7%; a 10 unit increase in the fixed broad internet subscribers brings about 1.6% increase in real GDP; and finally, a 10 unit increase in the mobile cellular subscriptions boosts real GDP by 1.6%.

Furthermore, to capture the regional differences between ‘South Asia’ (SA) and ‘East and South-East Asia’ (ESEA) with regard to impact of infrastructure over growth performances we have carried out regressions using a least squares dummy variable model (LSDV). It is observed that in the case of South Asia a 10 unit increase in Infrastructure Index results in a 3.1% rise in their real GDP, whereas, a 10 unit increase in Infrastructure Index results in a 1.2% increase in real GDP in ESEA. A reason for such difference in the size of the coefficients may be due to the differences in the level of development of infrastructure between SA and ESEA. As SA is well behind ESEA in terms of infrastructure development, improvements in infrastructure will bring about a larger positive effect on growth in SA than in ESEA.

The aforementioned analysis points to the fact that improvements in infrastructure significantly contributes to economic growth, and therefore, investment in infrastructure is an essential pre-requisite pediment. Hence, to opt for the ‘inclusive growth’ agenda, supply side bottlenecks should be addressed promptly. Priorities should be given to the development of infrastructures that can create highly adhesive ‘crowding in’ effect for private sector investment.

Published at the Thinking Aloud on 1 February, 2016

Published at The Financial Express on 1 March, 2016