Information about Middle East Oriente Medio
Chapter

Chapter 1: Economic Growth

Author(s):
International Monetary Fund
Published Date:
March 2007
Share
  • ShareShare
Information about Middle East Oriente Medio
Show Summary Details

Iran faces the challenge of increasing its growth rate to reduce unemployment and improve the living standards of its population over the medium term. Growth performance in recent years (6 percent during 2000–03) has been satisfactory and has been driven by major economic reforms as well as transitory factors such as high oil prices and expansionary fiscal and monetary policies. However, questions arise about the determinants of growth in Iran and the long-term sustainability of relatively high growth rates. Because past experience shows that the Iranian economy can grow at relatively high rates over an extended period, the first step is to examine the historical sources of growth and discuss the relevance of various contributing factors for the medium term. The second step is to provide an analytical framework for the formulation of growth-enhancing policies.

This chapter uses a growth accounting exercise to quantify the historical sources of growth during the period 1960–2002, including human capital accumulation and the contribution of TFP to growth. The chapter also presents an empirical study to quantify the role of several other contributing factors commonly discussed in the cross-country growth literature, including macroeconomic stability, financial development, trade openness, and changes in the terms of trade.1

Historical View of Growth Performance

During 1960–2002, real GDP growth in Iran averaged 4.6 percent per year (2 percent in per capita terms). Non-oil GDP grew at a faster pace of 5.5 percent during this period.2 There are three distinct subperiods (Figure 1):

  • During 1960–76, Iran enjoyed one of the fastest growth rates in the world: the economy grew at an average rate of 9.8 percent in real terms, and real per capita income grew by 7 percent on average. As a result, GDP at constant prices was almost five times higher in 1976 than in 1960. This stellar performance took place in an environment of relative political stability, low inflation (Figure 2), and improved terms of trade, as evidenced by the rising oil price relative to import prices (Figure 3). Both oil output and oil prices increased significantly during the period: oil production grew at an annual average rate of 10 percent, while oil prices relative to import prices increased by 214 percent during this subperiod.
  • The growth trend was reversed during 1977–88, reflecting the turmoil in the aftermath of the 1979 revolution, the eight-year war with Iraq, the international isolation of Iran, the increased state dominance of the economy, and the plummeting of oil output and revenue. In 1988, oil production was only 36 percent of its 1976 level, and oil prices were 40 percent lower in real terms. This resulted in real GDP growth of Ó2.4 per year on average. Excluding oil output, non-oil GDP also declined, albeit at a more moderate pace (0.5 percent per year).
  • With the reconstruction effort and a partial recovery in oil output, real economic growth recovered during 1989–2002 to an average of 4.7 percent per year. This period, however, was marked by sharp fluctuations in the growth pattern, as the postwar economic boom (1989–93) was followed by the stagnation of 1993–94, when the economy was hit by lower oil prices, lack of external financing, and economic sanctions. The ensuing severe debt crisis, together with inappropriate macroeconomic policies, had an adverse effect on growth, which hovered around 3.6 percent from 1995 to 2000. During 2000–03, real GDP growth picked up to 6 percent as a result of significant progress in economic reforms—such as trade liberalization, exchange rate unification, an opening up to FDI, and financial sector liberalization—but also because of high oil prices and expansionary fiscal and monetary policies.

Figure 1.GDP Growth Rates, 1960–2002

(Annual percentage change)

Sources: Iranian authorities; IMF staff estimates.

Figure 2.Inflation, 1960–2002

(Annual percentage change; end-period)

Sources: Iranian authorities; IMF staff estimates.

Figure 3.Oil GDP and Oil and Import Prices Ratio, 1960–2002

Sources: Iranian authorities; IMF staff estimates.

The growth performance of Iran compares favorably with that of the rest of the countries in the Middle East and North Africa (MENA) region, which averaged 4.2 percent per year during 1960–2002 (Table 1). Among the 17 countries in the region, only four—Oman, the Syrian Arab Republic, the United Arab Emirates, and the Republic of Yemen—grew faster than Iran. However, historical growth in Iran also exhibits higher variability than in the rest of the region.

Table 1.MENA Region: Economic Growth, 1960–2002(In percent, average for the period)
1960–761977–881989–20021960–2002
Iran9.8–2.44.74.6
Oil-producing MENA countries1
Algeria5.92.32.03.5
Bahrain4.54.34.74.5
Kuwait4.1–2.11.61.4
Libya14.0–0.2–0.24.5
Oman9.86.04.86.9
Qatar7.0–0.34.43.8
Saudi Arabia7.72.02.54.2
United Arab Emirates12.5–0.56.25.1
Average, oil MENA countries8.21.53.34.2
Non-oil MENA countries
Egypt4.65.73.34.5
Jordan4.85.43.54.5
Lebanon–3.95.20.30.3
Morocco4.34.12.73.7
Pakistan3.96.13.84.5
Sudan2.92.44.33.2
Syrian Arab Republic7.93.84.35.5
Tunisia7.14.24.45.3
Yemen, Republic of9.55.75.06.6
Average, non-oil MENA countries4.64.73.54.2
MENA average (excluding Iran)6.62.83.44.2
Source: IMF, International Financial Statistics (IFS).

Excluding Iran.

Source: IMF, International Financial Statistics (IFS).

Excluding Iran.

Determinants of Economic Growth

The empirical studies on the determinants of growth can be broadly divided into two main categories. The first includes growth accounting exercises, which consist of estimating the contributions to growth of basic factor inputs—labor, physical capital, and human capital—and a residual that captures the efficiency at which physical and human capital resources are used, or TFP. The second comprises several empirical studies analyzing cross-country growth regressions to find the relationship between different explanatory variables and growth.

Growth Accounting

A standard growth accounting framework is used to discuss the historical sources of growth in Iran. We use the following Cobb-Douglas production function:

where Y, K, and H represent output, physical capital, and human capital, respectively; α represents the contribution of physical capital to output; and t is an index for time. The term A represents TFP, or the efficiency at which the economy operates, which depends on factors such as the domestic political and international environment, the legal and regulatory framework, the creation and diffusion of more efficient technologies through international trade or FDI, and the effect of structural reforms, such as financial sector or labor market liberalization. Physical capital is considered as a homogeneous capital good, with no distinction made between equipment and nonequipment capital goods or between private and public capital goods (implicitly assuming that the productivity of the two types of capital is the same).3

To account for the effect of education on economic growth, a human capital index is constructed as a function of both the labor force and average years of schooling. However, in Iran it is difficult to measure the contribution of schooling to human capital because of the lack of an education quality index that would account for the changes in the productivity of education during 1960–2002..4 Therefore, the paper considers two different specifications of human capital, which result in two different growth accounting exercises.

A basic specification is to make human capital equal to raw labor, that is, Ht = Lt. Under this specification, an increase in the average years of schooling of the labor force does not increase the productivity of labor. Given that the crosscountry empirical evidence points to a positive effect of education on the productivity of labor, under this simple specification, the contribution of TFP to growth is overstated because it implicitly takes into account the effect of the quality of the labor input on output and growth.

A different assumption is to consider that schooling increases the productivity of the labor force along the following specification of human capital (Lucas, 1988):

where L represents the labor force and e is the average years of schooling of the labor force.

The above-specified production function implies that human capital accumulation exhibits increasing returns to scale. This means that if we double both the number of workers and the average years of education for the labor force, human capital increases fourfold. Because anecdotal evidence—such as the increased proportion of college graduates with nonmarketable skills—points to a reduction in the quality of education in Iran over the period under study, the growth accounting exercise using this technology specification may result in an overstatement of the contribution to growth of human capital and an understatement of the contribution of TFP.

Taking natural logarithms and differentiating with respect to time, the following decomposition of growth is obtained:

where g denotes the growth rate of the variable in the subscript. If factor markets are competitive, the first-order profit-maximizing conditions for the firm imply that α corresponds to the share of rental payments to capital in total income, that is, α = (rd)*K/Y, where r is the net rate of return on capital and d is the depreciation rate of capital. The derivation of the time series of the capital stock and the data sources are presented in Appendix 1.

The results of the growth accounting exercises under the two alternative specifications for human capital (Ht = Lt and Ht = Ltet) are shown in Tables 2 and 3, respectively.

The noninclusion of the effect of increased schooling on the productivity of the labor force and on growth results in a positive contribution of TFP to growth during 1960–2002, because changes in the quality of the labor force are implicitly included in TFP. Under the alternative specification, which considers that human capital increases linearly with average years of schooling, the growth accounting exercise suggests that the contribution of

TFP to growth becomes negative (–1.2 percent on average during 1960–2002) (Table 3).

Table 2.Sources of Economic Growth (Raw Labor), 1960–2002(In percentage points)
Average Growth RateContribution of CapitalContribution of Raw LaborContribution of TFP
1960–769.83.91.24.7
1977–88–2.41.71.4–5.5
1989–20024.72.31.51.0
1960–20024.62.11.41.1
Sources: Iranian authorities; IMF staff estimates.
Sources: Iranian authorities; IMF staff estimates.
Table 3.Sources of Economic Growth (Education), 1960–2002(In percentage points)
Average Growth RateContribution of CapitalContribution of Human CapitalContribution of TFP
1960–769.83.92.73.2
1977–88–2.41.75.5–9.6
1989–20024.72.34.3–1.8
1960–20024.62.13.7–1.2
Sources: Iranian authorities; IMF staff estimates.
Sources: Iranian authorities; IMF staff estimates.

Under both accounting exercises, the contribution of TFP to growth is positive during the high-growth subperiod of 1960–76 and becomes negative during the political turmoil and war period of 1977–88. This result points to the critical importance of political and external developments for Iran’s economic growth.

The results of the two exercises differ for 1989–2002. Under the first specification, in which human capital equals raw labor (Ht = Lt), the contribution of TFP to growth is positive; however, if we consider a linear effect of education on human capital (Ht = Ltet), the contribution of TFP to growth becomes negative. A more realistic TFP estimate may lie between these two extreme cases. In particular, it is likely that there was a very small (or even a negative) contribution of TFP to growth during the 1989–2002 subperiod,5 owing to slow progress in structural reforms and an increase in both inflation and its volatility.

Empirical Analysis

Cross-Country Empirical Evidence

Some stylized facts from cross-country growth regressions are relevant for Iran:

  • There was a positive relationship between the level of education of the labor force and economic growth. Barro (1991 and 1997) and Benhabib and Spiegel (1994) show that the initial level of education is an important factor in explaining subsequent growth. However, Bils and Klenow (2000) find that the causality goes from growth to increases in school enrollment rates.
  • Macroeconomic stability—usually defined as a combination of a low inflation rate, low budget deficits, and undistorted foreign exchange markets—improves the business environment and reduces the uncertainty on the return of investment projects and, therefore, has a positive relationship with economic growth. Fischer (1993) and Bleaney (1996) find that macroeconomic instability (measured by a combination of high inflation, fiscal imbalances, and high volatility of the real exchange rate) has a significant negative effect on economic growth and possibly also a negative effect on investment. More specifically, the literature finds a negative relationship between inflation and growth. Khan and Senhadji (2001) find that an inflation rate above 11–12 percent is associated with a significant reduction in growth in developing countries. Sarel (1996) finds that high inflation—above 8 percent per year—has a negative and statistically significant effect on growth, and that doubling the inflation rate would reduce the average growth rate by 1.7 percentage points. Barro (1997) finds a smaller effect: an increase in the inflation rate of 10 percentage points would reduce the growth rate by 0.2–0.3 percentage points.
  • Financial development reduces the cost of capital and generally has a positive correlation with growth, but the direction of the causality is difficult to establish. Demetriades and Hussein (1996) studied the experience of 16 countries: 4 displayed causality from financial depth to growth, 4 displayed causality from growth to financial depth, and 7 displayed a feedback relationship between finance and growth. Regarding financial repression and growth, both Roubini and Sala-i-Martin (1995) and Arestis and Demetriades (1997) find a negative relationship between financial repression indicators and growth. The only exception was Korea, in which financial repression favored the growth-leading export sector.
  • Trade openness generates technology spillovers and provides the economy with access to specialized inputs from abroad. The literature finds a significant effect of trade openness on growth. Greenaway, Morgan, and Wright (1998) cover 73 countries and use a dynamic regression framework to investigate potential lagged effects of openness on growth. They find that the positive effect of trade openness on growth becomes more significant over the long term, whereas in the short term this effect is much less important. Improvements in the terms of trade are generally associated with faster growth (see Barro, 1996 and 1997; Easterly and others, 1993; and Fischer, 1993).
  • Finally, other factors, such as political variables and income inequality, may also play an important role in economic growth. Alesina and others (1996) find a significant negative relationship between political instability and growth. This result is particularly strong when there are significant changes in the ideological position of the executive branch. In another empirical study, Mauro (1995) finds a negative correlation between political risk and economic growth. Other empirical studies show mixed results on the relationship between income inequality and economic growth.

Empirical Analysis of Factors Affecting Non-Oil GDP Growth

This section focuses on the empirical relationship between non-oil GDP growth and some factors commonly referred to in the literature as having a significant effect on growth performance—namely, trade openness, macroeconomic stability, terms of trade changes, and financial development. Given the importance of changes in oil production and political developments in Iran, the empirical analysis also includes these two variables.

To examine the link between trade openness and economic growth, we adopt the ratio of imports to non-oil GDP as a proxy for trade openness because of the lack of data on average tariffs and nontariff barriers to trade for the entire period 1960–2002. Macroeconomic stability is proxied by the inflation rate, owing to the lack of data on government deficits and exchange rate distortions during 1960–2002. Terms of trade are proxied by the change in the ratio of oil prices to import prices of industrial goods. Because oil represents about 80 percent of Iran’s exports, and 95 percent of imports are industrial goods, the ratio of the prices of these two types of commodities is a good proxy of the terms of trade. Financial development is proxied by the change in the ratio of broad money (M2) to non-oil GDP. Finally, we include a dummy variable for the subperiod 1977–88 to take into account the effect of political instability and war on growth.

Using the econometric package PcGets, which allows for an automatic reduction of a general model to a parsimonious one, we establish an empirical relationship between non-oil GDP growth and its explanatory variables, shown in Table 4.

Table 4.Non-Oil Annual GDP Growth, 1961–2002
ExplanatoryhVariablesCoefficientT-Value
Constant4.280022.3848
Change in oil production (–1)0.056112.6605
Imports/GDP ratio0.322526.3954
Consumer price index–0.310064.7403
Oil imports/price ratio0.047942.9851
M2/GDP ratio–0.274083.4136
Dummy 1977/88–5.987645.3164
R20.82758
TestsValueP-Value
Chow (1982:1)0.48420.935
Chow (1998:1)0.58930.673
AR 1–4 test1.12250.364
ARCH 1–4 test0.11320.977
Hetero test4.67670.946
Source: IMF staff estimates.
Source: IMF staff estimates.

The t-statistics of the regression show that all variables are significant at the 95 percent confidence level and explain 82.8 percent of variance of growth; there are no structural changes during the period (Chow tests); no autocorrelation (AR test); and no heteroscedasticity (ARCH and hetero tests) of the residuals of the regression.

The above results are largely consistent with the cross-country evidence on economic growth. Increased trade openness and macroeconomic stability (measured by inflation rates) are positively correlated with growth. Also, improvements in the terms of trade are positively correlated with growth. Higher oil production stimulates non-oil GDP growth through higher demand of inputs from the non-oil sector and because higher oil revenues stimulate public expenditures, particularly public capital expenditures (see Chapter 5 on oil revenue management). Of all the variables studied here, political instability and war have the strongest (negative) effect on growth, reducing growth by about 6 percentage points per year during the 1977–88 subperiod.

The main difference from the cross-country evidence in other studies concerns the negative effect of financial deepening (measured by the change in the ratio of M2 to non-oil GDP) on growth in Iran.6 The lack of positive correlation between financial deepening and growth in Iran, which is in contrast with other countries, could be attributed to an inefficient financial system that channels resources to investments with very low productivity or to the sectors with lower growth potential. The low or even negative TFP contribution to growth in Iran is likely to have captured these financial system deficiencies as well.

Sectoral Growth

During 1960–2002, the industrial sector exhibited the strongest performance. Industrial output (mostly manufacturing and construction) grew at 7.6 percent per year on average and was 23 times higher in 2002 than in 1960. As a consequence, the share of industrial output to GDP increased during this period from 7 percent to 25 percent. In contrast, the oil sector grew by just 2.5 times, and its relative weight decreased from one-third of GDP at constant prices in 1960 to less than 13 percent in 2002. The output in the agricultural sector grew at 4.2 percent on average, a slightly slower pace than GDP, but well above the average population growth of 2.6 percent, whereas the services sector grew at a faster pace than GDP (5.4 percent per year on average during the period). Table 5 shows the average growth rates for each sector.

Table 5.Average Sectoral Growth, 1960–2002(In percent, in real terms)
1960–761977–881989–20021960–2002
Agriculture4.63.94.14.2
Oil and gas10.0–8.62.52.4
Industries and mines14.0–1.37.37.6
Services11.1–1.94.85.4
Non-oil GDP10.1–0.55.05.5
GDP9.8–2.44.74.6
Source: Iranian authorities.
Source: Iranian authorities.

Despite the rapid growth of the industrial sector, the low or even negative growth of TFP during the period under study—together with high physical capital investment—suggests a low productivity of investment in the industrial sector, possibly reflecting trade restrictions7 and inefficient public sector investment in the industrial sector.

Summary of Empirical Findings

The analysis covers a period of 42 years (1960–2002) that witnessed significant political and social changes as well as periods of instability (for example, the 1979 revolution, the war with Iraq, economic sanctions). The analysis in the previous sections shows the following:

  • Changes in the political environment have had a major impact on economic growth in Iran. The periods of relative political stability and absence of major external conflicts (1960–76 and 1989–2002) are clearly associated with high GDP growth, whereas the political turmoil and war period of 1977–88 was associated with negative growth. The paper shows that the average annual growth rate during the 1977–88 period was reduced by 6 percentage points as a result of these factors.
  • The fivefold improvement in the average level of education of the labor force since 1960 may explain up to half of total economic growth in the past 42 years.8 However, it is difficult to determine precisely the magnitude of the contribution of investment in education to growth owing to a lack of the data needed to measure the effect of educational attainment on productivity.
  • Trade openness is significantly associated with faster GDP growth. An increase in the imports-to-GDP ratio of 1 percentage point is associated with faster growth of 0.3 percentage points.
  • Regarding macroeconomic stability and growth, there is a positive link between growth and lower inflation in Iran, and this relationship is statistically significant. A reduction in the inflation rate of 1 percentage point would increase potential growth by 0.3 percentage points.
  • Given the inefficiencies in the financial sector, the link between financial deepening and growth in Iran is not clear. In fact, when financial development is proxied by changes in the ratio of M2 to non-oil GDP, the relationship between these two variables becomes negative. Given the cross-country empirical evidence of a generally positive relationship between financial development and growth, it is likely that changes in the financial system that would increase its efficiency would yield potentially large gains in terms of long-term growth, reversing the observed negative relationship between financial depth and growth in Iran.

Policy Lessons

Three main policy lessons can be derived from the Iranian growth experience:

  • Structural reforms in a stable political environment are key to improving growth performance over the medium and long term. To increase the long-term growth rate of the economy above its historical mean of 4.6 percent per year, policies should be directed at increasing productivity (measured by TFP). Moreover, the crosscountry empirical evidence and the empirical findings for Iran show that growth is directly associated with factors such as trade openness, macroeconomic stability, and political stability. These findings call for stepped-up implementation of structural reforms—trade and FDI liberalization, privatization and deregulation to increase the size and role of the private sector, and further financial sector reform to eliminate financial repression that harms long-term growth. Other reforms, such as the elimination of subsidies and the implementation of fiscal, monetary, and exchange rate policies aimed at increasing macroeconomic stability, would also play a critical role in enhancing growth performance.
  • Increases in the efficiency of human capital resources through investment in education appear to be an important reason for Iran’s growth. In this respect, achievements in Iran since the 1979 revolution have been very important: the average level of schooling for the working population has more than tripled (from 1.5 years of schooling in 1979 to about 5 years in 2002). Education policies aimed at allocating increased resources to primary and secondary education and promoting on-the-job training programs would, in all likelihood, further enhance growth prospects. The need for further efforts in education is evident when we consider that Iran has an illiteracy rate of about 20 percent, despite the substantial progress since 1979.
  • Finally, with respect to the contribution of physical capital to economic growth, Iran’s investment rate (which averaged more than 30 percent during 1960–2002) is already high by international standards, even when compared with the high-growth countries of East Asia (see Table 6). Its payoff, however, as measured by average incremental capital-output ratios, does not suggest that it should be increased further, but that the efficiency of investment projects needs to be improved. The low efficiency of many investment projects undertaken in Iran, especially in agriculture, industry and mining, and housing, could be explained in part by subsidized energy and inputs and negative real interest rates on bank financing. Nonetheless, despite the high rates of investment over the past years, physical infrastructure is in need of upgrading and modernization.
Table 6.Comparison of Investment and Growth Performance of Iran with Six High-Growth Asian Economies, 1962–2002(In percent)
Average GDP GrowthAverage Investment/GDPInvestment/GDP Growth
Growth
Hong Kong SAR7.026.13.7
Indonesia5.421.74.0
Korea7.827.93.6
Malaysia6.728.64.3
Singapore7.936.04.5
Thailand6.726.54.0
Average6.927.84.0
Iran4.630.56.6
Sources: IMF, IFS; IMF staff estimates.
Sources: IMF, IFS; IMF staff estimates.
1Under Article VIII, Sections 2 (a) and 3 of the IMF’s Articles of Agreement, Iran committed to eliminating all exchange restrictions on current account transactions and refraining from introducing new ones.
Note: Prepared by José Bailén.
1Recent growth studies on Iran include Jalali-Naini (2003) and Mojaverhosseini (2003).
2Since the relative price of oil GDP increased by an average of 3 percent per year during the period 1960–2002, the ratio of nominal oil GDP to total GDP increased from 12.8 percent to 22.1 percent, even though real oil output increased at a slower pace than real non-oil output.
3This assumption is made because of the difficulty of measuring the productivity of public capital goods in Iran.
4Other proxies of the quality of human capital, such as an increase in productivity of workers (measured by their relative salaries, presumably reflecting relative educational attainment), are not available in Iran.
5The TFP growth estimates are subject to measurement errors of physical and human capital. However, using the same growth accounting methodology, the TFP growth estimates for Iran are systematically lower than those in most other developing countries.
6With alternative specifications of the model, such as using changes in real money, the relationship between financial deepening and growth becomes statistically insignificant.
7Despite recent reductions in import taxes and nontax barriers, Iran’s trade regime is very restrictive: the average (unweighted) tariff rate in 2002 was 30 percent, the eleventh highest tariff rate out of 193 surveyed countries, according to the Trade Restrictiveness Index Ratings in the Trade Policy Information Database (TPID) maintained by the Policy Development and Review Department, IMF.
8The quantification of the contribution of education to growth depends on the specification of the production function.

    Other Resources Citing This Publication