Abstract
This Selected Issues paper and Statistical Appendix on Pakistan looks at the worrisome trend of declining growth from a growth-accounting point of view. The paper provides considerable evidence that Pakistan’s striking “social gap” in education and health is indeed a main culprit for a weakening growth performance. It looks at the financial sector as an additional area that is central for growth and governance, and where reforms are well advanced. The paper also analyzes how to ensure a continuation of prudent fiscal policies in Pakistan that would reduce public debt to more sustainable levels.
II. Long Term Growth in Pakistan: Trends in Productivity and Factor Accumulation3
A. Introduction
7. From 1961 to 2001, Pakistan has always experienced positive real GDP growth. However, growth rates slowed from almost 7 percent in the 1960s to about 3 percent in the 1990s. In the regional context, Pakistan grew faster than South Asia by an average 2 percent through most of the 1960s and 1970s, and at similar rates during the 1980s. However, since 1993 Pakistan’s growth was below the regional average (Figure II-1).4 More generally, trends in growth in Pakistan has been steadily slowing from initially very high levels since the 1960s, whereas growth in South Asia has been steadily accelerating for the four decades.
Pakistan and South Asia: Real GDP Growth Trends, 1962–1999
(Three-year moving average, in percent)
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Sources: World Development Indicators (2002); and World Bank.Pakistan and South Asia: Real GDP Growth Trends, 1962–1999
(Three-year moving average, in percent)
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Sources: World Development Indicators (2002); and World Bank.Pakistan and South Asia: Real GDP Growth Trends, 1962–1999
(Three-year moving average, in percent)
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Sources: World Development Indicators (2002); and World Bank.8. This paper is organized as follows. Section B breaks down Pakistan’s growth trends into productivity and factor accumulation trends, using a standard growth accounting approach. Section C discusses the results. Section D speculates about some of the factors that may explain Pakistan’s total factor productivity (TFP) performance, Section E deals more specifically with the role of Pakistan’s social gap. Section F reviews medium-term growth prospects in light of the preceding analysis.
B. Growth Accounting
Cobb-Douglas production function
9. To analyze Pakistan’s growth of aggregate output, a standard growth accounting exercise is used, assuming that national production can be modeled by using a Cobb-Douglas production function with constant returns to scale:
Here Y denotes real gross domestic product, A—total factor productivity (TFP), K—real capital stock, L—labor force, and H—a human capital stock measure (it is assumed that H=sL, where s represents average years of school of total population aged 25 and over, as estimated by Barro and Lee (2000). Due to the unavailability of a single data source for the whole period, a variety of sources was used. Estimates of real capital stock for 1960–1994 were obtained from the study by Bosworth and Collins (1995). For the later years, real capital stock was estimated using real gross capital formation and an assumed depreciation rate of 3 percent. Labor force time series were generated from databases from Bosworth and Collins, International Labor Organization, and the World Economic Outlook. These data weaknesses imply that in this exercise only orders of magnitude, rather than exact estimates, should be considered as relevant. A further limitation of this approach is that it includes the large agricultural sector, which is significantly influenced by weather conditions. Thus, TFP fluctuations, especially in dealing with subperiods, may reflect variations in rainfall.
Equation (1) can be rearranged to:
where the lowercase variables denote the growth rate of corresponding uppercase variables. GDP growth is thus broken down into TFP growth and a weighted average of the growth rates of production factors.
10. For the purposes of this study equal factor shares α=β=(1−α−β)=1/3 are assumed. To assess the robustness of our results, TFP was also estimated using different factor share values. This did not significantly change the final results (see below).
Trend GDP—adjusted TFP
11. TFP growth estimates derived using actual GDP make it difficult to determine to what extent the TFP fluctuations are driven by short-term factors such as droughts or business cycles, as opposed to longer-term economic trends. To measure TFP growth adjusted to such cycles, trend GDP was estimated via a Hodrick-Prescott filter (HP filter), decomposing time series data into cyclical and growth components and interpreting the latter as trend GDP. Adjusted TFP growth was then calculated using trend GDP in Equation (2) (Figure II-2).
Pakistan: Actual and Trend TFP Growth for α = β = 1/3, 1962–2000
(Three-year moving average, percent)
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Pakistan: Actual and Trend TFP Growth for α = β = 1/3, 1962–2000
(Three-year moving average, percent)
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Pakistan: Actual and Trend TFP Growth for α = β = 1/3, 1962–2000
(Three-year moving average, percent)
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
C. Main Results
12. Overall, for the 1961–2001 period, Pakistan has experienced moderate TFP growth (0.5 percent annually on average, whether measured by actual or by trend TFP), indicating that trend GDP growth of 5.6 percent over the period was driven mostly by factor accumulation, especially that of human and physical capital. In the 1960s, TFP experienced negative growth (on average -2.2 percent). In the 1970s, TFP growth became positive, peaking in the 1980s (2.4 percent). This latter trend seems to have reversed in the second half of the 1990s when TFP growth declined to -0.4 percent (Figure II-3, Table II-1).
Trend GDP and Factor Input Growth Development
(Arithmetic average, in percent)
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Trend GDP and Factor Input Growth Development
(Arithmetic average, in percent)
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Trend GDP and Factor Input Growth Development
(Arithmetic average, in percent)
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Pakistan: Real GDP, TFP, and Production Factor Growth, 1961–2001
Pakistan: Real GDP, TFP, and Production Factor Growth, 1961–2001
| 1961-1971 | 1971-1981 | 1981-1991 | 1991-2001 | 1996-2001 | 1961-2001 | |
|---|---|---|---|---|---|---|
| (Average annual growth rates) | ||||||
| Real Actual GDP | 6.6 | 5.0 | 6.2 | 3.9 | 3.2 | 5.4 |
| Real Trend GDP | 6.7 | 5.7 | 6.1 | 3.8 | 3.3 | 5.6 |
| Actual TFP | −2.2 | 1.3 | 2.5 | 0.6 | −0.5 | 0.4 |
| Trend TFP | −2.2 | 2.0 | 2.4 | 0.5 | −0.4 | 0.5 |
| Physical Capital Stock | 13.1 | 5.2 | 5.6 | 4.4 | 3.7 | 7.2 |
| Labor Force | 2.0 | 2.8 | 1.5 | 2.4 | 3.4 | 2.3 |
| Human Capital Proxy | 11.6 | 3.2 | 4.2 | 3.0 | 4.0 | 5.8 |
Pakistan: Real GDP, TFP, and Production Factor Growth, 1961–2001
| 1961-1971 | 1971-1981 | 1981-1991 | 1991-2001 | 1996-2001 | 1961-2001 | |
|---|---|---|---|---|---|---|
| (Average annual growth rates) | ||||||
| Real Actual GDP | 6.6 | 5.0 | 6.2 | 3.9 | 3.2 | 5.4 |
| Real Trend GDP | 6.7 | 5.7 | 6.1 | 3.8 | 3.3 | 5.6 |
| Actual TFP | −2.2 | 1.3 | 2.5 | 0.6 | −0.5 | 0.4 |
| Trend TFP | −2.2 | 2.0 | 2.4 | 0.5 | −0.4 | 0.5 |
| Physical Capital Stock | 13.1 | 5.2 | 5.6 | 4.4 | 3.7 | 7.2 |
| Labor Force | 2.0 | 2.8 | 1.5 | 2.4 | 3.4 | 2.3 |
| Human Capital Proxy | 11.6 | 3.2 | 4.2 | 3.0 | 4.0 | 5.8 |
13. During the 1960s, Pakistan’s strong growth performance was driven by real capital and human capital stock accumulation: the country experienced an average trend GDP growth of 6.7 percent with major advances in agricultural and large scale manufacturing sectors (Table II-2). After near-stagnation of agriculture in the 1950s, during the 1960s the “Green Revolution” technology was introduced on a large scale. Industrial production was stimulated by import-substitution policies, encouraging private investment in this sector. The expansion of these two sectors was particularly capital intensive, resulting in an average physical capital stock growth rate of 13.1 percent per year. While the labor force grew by a modest 2 percent per year on average, levels of schooling improved significantly due to advances in basic education that resulted in an average human capital stock growth of 11.6 percent per year.
Pakistan: Real value Added By Sector, 1961–2001
Pakistan: Real value Added By Sector, 1961–2001
| 1961-1971 | 1971-1981 | 1981-1991 | 1991-2001 | 1996-2001 | 1961-2001 | |
|---|---|---|---|---|---|---|
| (Average annual growth rates) | ||||||
| Agriculture | 4.3 | 3.1 | 4.8 | 4.5 | 4.9 | 4.3 |
| Industry | 10.1 | 5.6 | 6.8 | 4.1 | 2.2 | 6.2 |
| Manufacturing | 9.3 | 5.2 | 7.2 | 3.9 | 2.0 | 6.2 |
| Services | 6.4 | 6.1 | 6.6 | 4.6 | 4.0 | 6.0 |
Pakistan: Real value Added By Sector, 1961–2001
| 1961-1971 | 1971-1981 | 1981-1991 | 1991-2001 | 1996-2001 | 1961-2001 | |
|---|---|---|---|---|---|---|
| (Average annual growth rates) | ||||||
| Agriculture | 4.3 | 3.1 | 4.8 | 4.5 | 4.9 | 4.3 |
| Industry | 10.1 | 5.6 | 6.8 | 4.1 | 2.2 | 6.2 |
| Manufacturing | 9.3 | 5.2 | 7.2 | 3.9 | 2.0 | 6.2 |
| Services | 6.4 | 6.1 | 6.6 | 4.6 | 4.0 | 6.0 |
14. The 1970s witnessed a decline in growth rates of GDP and almost all factor inputs. The nationalization of industries and financial institutions in 1972, as well as surging oil prices may have contributed to the slowdown in the industrial sector. Similarly, difficulties related to land reforms, climate shocks, and a cotton virus depressed agricultural production. Physical capital stock accumulation slowed down to 5.2 percent per year. Although average annual labor force growth increased to 2.8 percent, human capital stock growth fell to 3.2 percent per year due to a quasi-stagnation in the average years of schooling for this period (it declined from 1.68 years in 1970 to 1.62 years in 1975, increasing to 1.74 years in 1980). Trend GDP grew by 5.7 percent per year on average, owing to a much improved TFP growth performance.
15. The lost growth momentum was partly recovered in the 1980s: average GDP growth of 6.1 percent was approaching the levels of the 1960s. Policies of deregulation and privatization may have contributed to improvements in private sector confidence and industrial sector performance. Agriculture showed an uneven performance during this period, experiencing negative growth during the first half of the decade. The average physical capital accumulation rate of 5.6 percent remained close to the level of the 1970s. Improvements in schooling led to an increase in human capital stock accumulation rate to 4.2 percent per year, despite a slowdown in labor force growth to 1.5 percent. Significant GDP growth with moderate production factor accumulation was made possible by a further improvement in TFP growth.
16. Economic growth decelerated again in the 1990s with average trend GDP growth of 3.8 percent per year and stagnant TFP. During this decade the industrial sector, especially the textile industry, performed poorly; it might have been affected, inter alia, by prolonged political instability and inadequate infrastructure and power supply. Agricultural growth similarly slowed down mainly due to resource degradation. Physical capital stock growth decelerated somewhat, to average 4.4 percent. Human capital growth also decreased to 3 percent, despite the acceleration in labor force growth.
D. TFP Growth Determinants
17. To determine the exact reasons for the deterioration in Pakistan’s TFP performance during the 1990s is beyond the scope of this exercise. However, a number of factors are likely to have contributed to lower TFP growth, such as a decline in foreign direct investment (FDI), trade openness, and the quality of governance.
Foreign direct investment
18. A number of studies have found that FDI can contribute to TFP growth.5 FDI can facilitate technology transfer to developing countries from foreign to domestic firms, as well as better managerial and production practices in domestic firms. Pakistan has experienced declining FDI inflows since the mid-1990s (Figure II-4). FDI levels also trended lower in South Asia, but the trend was less pronounced than in Pakistan.
Foreign Direct Investment: Net Inflows, 1990–1999
(In percent of GDP three-year moving average)
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Source: World Development Indicators 2002, World Bank.Foreign Direct Investment: Net Inflows, 1990–1999
(In percent of GDP three-year moving average)
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Source: World Development Indicators 2002, World Bank.Foreign Direct Investment: Net Inflows, 1990–1999
(In percent of GDP three-year moving average)
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Source: World Development Indicators 2002, World Bank.Trade openness
19. Higher trade openness may benefit TFP growth, for instance, through efficiency gains from specialization, by enlarging the interaction with modern economies, and raising the scope for learning-by-doing (further discussed in Miller and Mukti (2002)). Similarly to the developments in FDI, the evolution of Pakistan’s trade openness, as measured by the GDP share of exports plus imports, compared unfavorably with that of the region overall, with a decline in the 1990s (Figure II-5). Trade openness in South Asia was increasing throughout the decade, albeit at a slower rate during the second half of the 1990s.
Trade, 1990–1999
(In percent of GDP, three-year moving average)
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Source: World Development Indicators 2002, World Bank.Trade, 1990–1999
(In percent of GDP, three-year moving average)
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Source: World Development Indicators 2002, World Bank.Trade, 1990–1999
(In percent of GDP, three-year moving average)
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Source: World Development Indicators 2002, World Bank.Quality of governance
20. A number of studies have indicated that governance may be an important determinant of economic growth. Olson, Sarna, and Swamy (2000) use International Country Risk Guide Rankings (ICRG) as a proxy for quality of governance and show that productivity growth rates tend to be higher in better-governed countries.6 Composite ICRG ratings that include ratings for political, economic, and financial risk were consistently lower for Pakistan throughout the 1990s, peaking in August 1997 with 63 (moderate risk) and declining afterwards to 56 in March 2002 (high risk). South Asia, on the other hand, maintained its moderate risk ratings (Figure II-6).7
Composite ICRG Ratings, 1990–2001
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Source: International Country Risk Guide.Composite ICRG Ratings, 1990–2001
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Source: International Country Risk Guide.Composite ICRG Ratings, 1990–2001
Citation: IMF Staff Country Reports 2002, 247; 10.5089/9781451830569.002.A002
Source: International Country Risk Guide.E. Pakistan’s Growth Performance and Its “Social Gap”8
21. The above analysis highlights the important role that weak growth of human and physical capital have played in reducing Pakistan’s growth rate. To determine why human and physical capital formation have slowed is beyond the scope of this section. However, Chapter III explores in more detail to what extent Pakistan’s comparatively poor health and education indicators, that is what has come to be called Pakistan’s “social gap”, have contributed to low growth. Poor performance on human development will not only directly slow the growth of human capital, but will probably also lower the investment rate and TFP growth, for example by adversely affecting governance. Pakistan’s growth slowdown may thus be in part attributable to its social gap, which has weakened its ability to take advantage of skill-based, technologically-driven changes in the world economy.
F. Medium-Term Growth Prospects
22. Current medium-term macroeconomic projections project an average annual real GDP growth of 5 percent by 2003/04. The above model and assumptions can be used to shed some light on medium-term growth prospects for Pakistan and the realism of these projections.
23. The attainability of this target depends on production factor and TFP growth rates. If their current growth rates would be raised to the average level of the period of 1961–2001, target real GDP growth can be achieved and even surpassed (Table II-7). If more recent trends were to continue, however, real GDP growth would be significantly lower: target GDP growth is not attainable if factor accumulation and TFP continue to grow at average 1991–2001 rates, and even less if their growth will remain at the 1996–2001 averages.
24. The “structural break” implied by the growth targets can also be highlighted by calculating the increase in the average growth rate of one factor necessary to achieve target growth while keeping other factor growth rates constant at 1991–2001 averages. In this case, the target GDP growth will be attained if, for example: (a) the real capital stock will grow at an average rate of 7.7 percent per year; (b) average labor force growth will be 5.6 percent; (c) average human capital stock growth rate will be 6.3 percent (equivalent to achieving three years of schooling by 2003, compared with 2.45 years in 2000); or (d) average TFP growth will grow at 1.7 percent.
References
Ahmed, S. M. and M.I. Ansari, 1998, “Financial Sector Development and Economic Growth: The South Asian Experience,” Journal of Asian Economics, Vol. 9, No. 3, pp. 503–517.
Barro, Robert, and Jong-Wha Lee, 2000, “International Data on Educational Attainment: Updates and Implications,” NBER Working Paper No. 7911 (Cambridge, Massachusetts: National Bureau of Economic Research).
Blomstrom, M., Lipsey, R., Zejan, M., 1992 “What Explains Developing Country Growth,” NBER Working Paper No. 4681 (Cambridge, Massachusetts: National Bureau of Economic Research).
Borensztein, E., J. De Gregorio, and Jong-Wha Lee, 1998, “How Does Foreign Direct Investment Affect Economic Growth,” Journal of International Economics, Vol 45, pp. 115–135.
Bosworth, Barry, and Susan Collins, 1995, “Accounting for Differences in Economic Growth,” Brookings Discussion Papers in International Economics, Brookings Institution, No. 115.
De Mello, L., 1997, “Foreign Direct Investment-Led Growth: Evidence From the Time Series and Panel Data,” Oxford Economic Papers, Vol. 51, No. 1, pp.133–151.
Ericsson, J., and M. Irandoust, 2001, “On the Causality Between Foreign Direct Investment and Output: A Comparative Study,” The International Trade Journal, Vol. 15, No.1, pp. 1–26.
Miller, Stephen M., and Mukti P. Upadhay, 2000, “The Effects of Openness, Trade Orientation, and Human Capital on Total Factor Productivity,” Journal of Development Economics, Vol. 63, pp. 399–423.
Olson, M., N. Sarna, and A.V. Swamy, 2000, “Governance and Growth: A Simple Hypothesis Explaining Cross-Country Differences in Productivity Growth,” Public Choice, Vol. 102, pp. 341–364.
Sadik, Ali T. and Ali A. Bolbol, 2001, “Capital Flows, FDI, and Technology Spillovers: Evidence from Arab Countries,” World Development, Vol. 29, No. 12, pp. 2111–2125.
Senhadji, Abdelhak, 2000, “Sources of Economic Growth: An Extensive Growth Accounting Exercise,” Staff Papers, International Monetary Fund, Vol. 41, No. 1.
Prepared by Alina Milasiute (MED).
Data for South Asia represent weighted averages of growth rates for Afghanistan, Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, and Sri Lanka.
Previous research on the effects of FDI on economic growth includes Borensztein et al (1998) and Blomstrom et al. (1992). Some examples of the effects of FDI on TFP are discussed in Ericsson and Irandoust (2001), De Mello, L. (1997), and Sadik and Bolbol (2001).
The ICRG Risk Rating System assigns a numerical value (risk points) to a predetermined range of risk components according to a preset weighted scale for each country covered by the system. Each scale is designed to award the highest value to the lowest risk and the lowest value to the highest risk: very low risk (100–80), low risk (70–79.5), moderate risk (60–69.5), high risk (50–59.5), and very high risk (0–49.5).
The ICRG ratings for South Asia are calculated as the arithmetic average of ICRG ratings for Bangladesh, India, Pakistan, and Sri Lanka.
See Easterly (2001) for an approach along these lines.
