II. The Non-Oil Sector in Kazakhstan: Links with the Oil Industry and Contribution to Growth23
27. Kazakhstan, as other CIS economies, saw its output collapse after the disintegration of the Soviet Union. From 1996, the economy started to grow at an average rate of 5 percent per year between 1996–2002.24 The initial recovery was mostly based on the emerging oil industry. The non-oil economy decreased or stagnated for most of the decade, but started to show significant growth from 2000 on.
Oil Production and GDP
28. Kazakhstan’s transformation into an oil rich country brings a series of challenges to the economy in general and in particular to the non-oil sector. Natural resource sectors tend to rely on economies of scale, with low levels of employment and slow technical progress. Their potential growth is also constrained by the existing reserves of non-renewable resources. Economies dominated by those sectors tend to have lower long-term growth,25 high income and asset inequalities, and face larger volatility associated with the changes in commodity prices, such as oil. To avoid the negative impact of increasing oil wealth, including the “Dutch disease” type of phenomenon,26 it is important to develop the policies and reforms that support sustainable growth of the non-oil economy.
29. The oil sector can have a positive impact on the overall economy through the linkages to sectors such as the services, and investment projects associated with oil extraction. Moreover, the growth of the oil sector has contributed to rising household incomes and consumption demand. In Kazakhstan, the recent surge in overall growth of the economy, with substantial gains in non-oil employment, is partially due to positive externalities from booming oil revenues. However, it is important that the non-oil economy be able to sustain its growth independent of volatile oil revenues, in order to achieve a balanced development, creation of jobs, and a reduction of poverty.
30. This chapter analyzes the recent strong growth in the economy, taking a closer look at the links between the oil and the non-oil sector, and assesses the contribution of each sector. Section B gives an overview of the sources of growth in the economy. In section C, an attempt is made to estimate the links between the oil and non-oil sectors. The last section presents a more detailed view of two large sectors in the non-oil economy, namely agriculture and manufacturing.
B. Sources of Growth in Kazakhstan
31. The economic revival was initially based on increasing productivity gains. Since 2000, factor accumulation has also played a significant role, even though productivity gains remain the major engine of growth.27 Transition economies typically have a first stage of recovery mostly based on large productivity gains, specially total factor productivity (TFP) gains,28 due to macroeconomic stabilization, structural reforms, and a more efficient use of existing capital stock and labor. While a similar trend can also be observed in Kazakhstan, there was an early move towards large investments in capital due to structural changes in the main industries.
Figure II-1:Sources of Growth 1/
Sources: National Statistics Agency of Kazakhstan and staff estimates.
1/ The investment values for the growth accounting exercise are based on national accounts except for 2002, which is based on surveys of enterprises and staff estimates, and may differ from the National Accounts. The graph with investment by sector is also based on the surveys data. The large increase in agricultural employment in 2000–01, and the decline in Services in 2000, reflects only in part a better measurement of the unofficial economy and some reclassification of employment between sectors.
32. The oil industry, which had only a small share in total output in the early 1990s, has become the dominant industry. Investment in the sector accounts for about half of all investment in the country. For the remaining sectors of the economy, the capital stock started to grow in 2001, but accounted for only a quarter of the overall growth of capital. The capital-output ratio has been decreasing since 1996 as obsolete capital has been replaced, and new investments prove to be more efficient. However, investment in the non-oil economy remains relatively low, particularly in agriculture and non-oil industry. Kazakhstan is a landlocked and sparsely populated country, which brings even greater challenges for trade and dampens investment. It becomes, therefore, more urgent to invest in public infrastructure, particularly roads and communications, create a stable legal framework and promote free trade inside and across borders, in order to improve the profitability of private sector investment, both domestic and foreign.
33. Labor accumulation contributed negatively to growth during the 1990s, when employment declined substantially, and there was large emigration. After the Russian crisis, this trend started to change, and there has been a strong recovery in the last years, especially after 2000. However, the largest share of the increase was in agriculture, where productivity remains low. The increase in agricultural employment seems to be partially a response to government incentives,29 and also reflect a move from the unofficial to the official economy. After independence the rural economy collapsed, rural enterprises disappeared and a large segment of the population moved to subsistence agriculture. Since 1999, the state has introduced several programs to support agriculture which resulted in a progressive move towards market activities and the official economy. The surge in official agricultural employment in 2000 and 2001 was in large part a reflection of these changes.30 Services and construction have seen some gains in employment, while industrial employment has been declining, reflecting better use of labor resources. The oil industry employs directly a very small share of the work force, less than one percent, and has no significant impact on overall employment.
34. A decomposition of growth accounting by sectors, shows that agriculture has been the worst performer of all sectors, particularly in terms of productivity. In agriculture, growth has been based on extensive use of labor and land, in particular since 1999. The sector has an increasingly obsolete capital stock, and only recently have there been some productivity gains. The recorded large increases in agriculture production, are only in part related to better measurement of the non-official economy. Sectors where the links with the oil industry are strong, for example construction and transportation, are the fastest growing. Services have been soaring in the last three years following the increase in real incomes. The services sector has enjoyed both productivity gains and an increase in the capital stock.
|Value Added||TFP||Labor productivity||Capital-Output Ratio|
|of which: Transport, and Comm||5.0||5.1||1.2||−10.4|
C. The Impact of the Oil Sector
35. In Kazakhstan, perhaps in contrast to more mature oil economies, the impact of oil production goes much beyond the oil industry’s direct contribution to GDP. The industry is just emerging, with several new oil fields to be explored and developed, and will need large investments and services associated with the expansion of extraction and transportation of oil. In order to study the impact of oil on the economy, a broad measure of the oil sector will be used, including not only the value added of the oil industry, but also the associated services and goods, the industry’s investments, and the oil refining industry.32
Figure II-2:The Oil Sector: Size and Contribution to Growth1
Sources: National Statistics Agency of Kazakhstan and staff estimates.
1 The size of the oil sector has been steadily increasing in real terms in the recent years. However, in 2001 there was a decline in its share of nominal GDP due to lower oil prices and a strong performance of the non-oil economy, particularly agriculture.
36. The value added by oil extraction represents about one third of the total value of output. Among the inputs used services account for about 40 percent. While transportation is the largest component, there are also significant expenditures on real estate and related services, and on a lesser scale on financial services and trade. A large part of remaining inputs is related to the oil industry itself and related industries like oil refining. The metallurgical industry also provides inputs to the oil industry.
37. Overall the oil sector, which in 1998 represented less than 10 percent of GDP, has been responsible for almost half of economic growth since 1998. Oil extraction alone has contributed about 1.2 percentage points a year to overall growth. The contribution to GDP growth of services related to oil extraction, transportation and investment projects in the sector, has averaged 2.3 percentage points. The non-oil economy, while posting a respectable average rate of growth of about 6 percent between 1999 and 2002, is lagging the oil sector, which grew at 24 percent per year over the same period.
38. The impact of investments in the oil industry has been impressive, particularly in the construction sector. Construction associated with oil has grown at an annual rate of 50 percent over the last four years. The rest of the construction sector saw an average annual decline of 7 percent in the same period, notwithstanding a recovery in non-oil construction since 2001. Also, services associated with oil production and investment projects have grown substantially in the last years.
39. Oil sector growth depends not only on oil extraction but also on the oil price. Under higher oil revenues, due to higher prices or production, there is a spill-over effect on the rest of the economy. The rise in revenues translates into higher spending on related services and, to a lesser degree, accelerated investment projects. In 2000, as the oil price increased substantially, the oil sector expanded at an impressive rate of 35 percent, more than twice the rate of oil extraction. As the oil price declined slightly in 2001, there was a slowdown in growth in the oil sector, mainly due to a sharp decline in services provided to the oil sector. The impact of declining oil prices, however, was smoothed out by investment in the sector, less affected by temporary declines in prices, and increased oil production. There is also a role for policy in smoothing out the impact of oil prices volatility. Since 2001, the government sends a share of oil revenues to an oil fund to be invested abroad. In periods of high oil prices the fund works to sterilize part of the foreign exchange inflows. In periods of low oil prices, the state budget can receive revenues from the fund to prevent sudden shortfalls in the budget. The fund also helps to stabilize the exchange rate and eases the burden on the conduct of monetary policy.
40. The links between the oil industry and the rest of the economy will likely increase in the next years as production rises and the sector implements its investment plans. However the dimension and composition of the links are not completely clear. In terms of transportation, there have been large investments to move towards more efficient and less costly transportation.33 This will reduce the impact of increasing oil production on the transportation sector. In terms of other services to the sector, the links will depend on the development of competitive domestic suppliers. Such associated services would benefit from foreign investment and expertise at this early stage of development.
|Growth rates||Share of GDP|
A computable general equilibrium model
41. Over the next decade, the contribution of the oil sector to overall growth is likely to increase substantially. As the oil sector grows in size, the economy will become increasingly subject to oil price volatility. To complement the previous analysis, this subsection presents a general equilibrium model to illustrate some of the challenges that the Kazakhstani authorities will face. The model used is based on the work developed by Dervis, Melo, and Robinson (1982), usually known as DMR model.34 However, several adjustments are introduced to reflect the specifics of the Kazakhstan case. In particular, there is a detailed description of the impact of oil related flows on the balance of payments.35 The main advantage of multisector models, with linkages between the different sectors, is that it allows the estimation of the impact of changes in exogenous variables, such as oil prices, and reactions to changes in policy variables, for example exchange rate policy. However, the model only focuses on the real side of the economy, and has no explicit role for money.
Box II-1.Outline of the Model
A multisector Computable General Equilibrium (CGE) model is used, which is based on input-output tables and national accounts. Given the focus on the oil industry, the model incorporates a detailed description of this sector. The structure of the model is as follows:
1. There are 8 sectors: agriculture, oil extraction, machine building, other industry, construction, trade, transportation and communications and other services.
2. The economy comprises households, the government, and firms. Households supply labor and capital, receiving wages and capital rents. Their income is used for consumption and savings. The government collects taxes from firms and households, it uses the revenues to consume and save. Firms use intermediate goods, labor and capital to produce goods that can be sold domestically or exported.
3. Households spend a constant share of their income for each type of composite good (combination of imported and domestically produced). There is imperfect substitution between foreign and domestically produced goods. Households preferences assume a constant elasticity of substitution between the two.
4. There is limited labor mobility between sectors, and the model permits wages to differ among sectors by allowing for different factor productivities. Capital is assumed to be sector specific and is given for the period.
5. The economy is assumed to be a small economy, having no impact on world prices.
6. The model incorporates specific rules for the oil industry. In particular, it assumes that variations in oil revenues will lead to changes in payments to foreign shareholders. Also increases in taxes from this industry, except production taxes (VAT), will revert to an oil fund, and are invested abroad.
7. The model used is a differential linearized version of the DMR. The model is solved for percentage changes in the variables.
8. The underlying data are from the input-output tables and national accounts for 2000 for Kazakhstan, and staff estimates.
9. The results reported here were obtained using the GEMPACK economic modeling software (Harrison and Pearson (1996)).
42. The simulation carried out in here is an increase in oil prices by 10 percent, with two different policy responses.36 The first scenario assumes that the central bank does not intervene in the foreign exchange market, allowing the exchange rate to float. These results are compared with an alternative scenario, where the nominal exchange rate is kept constant, and any excess of foreign exchange is accumulated in the form of reserves by the central bank.
43. Under both scenarios, there is an improvement of the balance of trade due to the rise of the oil price.37 In scenario I, the rise in exports is partially compensated by increased imports, as the nominal exchange rate appreciates by 1.4 percent. Higher oil prices will also result in outflows of capital associated with the repayment of intra-company loans (by oil companies) and other payments to foreign shareholders, and flows to the oil fund. It is assumed that increases in tax revenues from the oil industry, except for VAT revenues, will be transferred to the oil fund and invested abroad.38
|Exchange Rate (appreciation -)||−1.4||0|
|Exports (foreign currency)||3.2||3.6|
|Imports (foreign currency)||1.5||0.6|
|Reserves (change in percentage of GDP)||0.0||0.6|
|Trade balance (change in percentage of GDP)||1.1||1.8|
44. Under scenario I, the economic growth rate increases by about 0.14 percent.39 The surge in oil revenues results in increased private expenditures, via higher wages and capital rents. This increase in domestic demand is in large part offset by a decline in real net exports which is a typical reaction to a positive shock to the terms of trade and the appreciation of the exchange rate. Exports rise in value, but because of the appreciation of the currency there will be a decline in the volume of non-oil exports and an increase in the volume of imports. Although nominal budget revenues increase, the increase in public expenditures in real terms is limited due to the flow of most of the increased oil revenues to the oil fund and to some extent also by rising prices.
45. The nominal appreciation of the currency limits the increase in the CPI, but also reduces the competitiveness of the non-oil economy. The exporting and import-competing sectors, namely agriculture and industry, lose competitiveness not only because of the nominal appreciation, but also because of the increase in prices of non-tradable inputs (mostly services).40 Machine building is the hardest hit industry due to import competition and the decline in demand from other industries. In contrast, construction grows fueled by larger demand for investment and services. The increased domestic demand also limits the decline in agriculture and non-oil industry.
46. Under the second scenario, the central bank intervenes to keep the nominal exchange rate stable. Given the outflows to the oil fund and foreign shareholders, the central bank would need to offset only a portion of the trade surplus. Because the exchange rate is kept constant, there is a smaller increase in imports and a lower decline in non-oil exports than under scenario I. By preventing an appreciation of the currency, price increases will be higher than in scenario I, leading to a relatively lower real domestic demand. Given the parameters of the model, real investment actually declines under the second scenario.41
47. The scenarios presented here are sensitive to the assumptions made. They constitute illustrative examples of policy responses, involving fixed and flexible nominal exchange rates, respectively. They highlight the difficult choices the authorities face when deciding on how to respond to oil price fluctuations.
D. Sectoral Analysis
The non-oil industry
48. The industrial sector has been undergoing a dual restructuring process, because of the transition to a market economy, and the emergence of the oil industry. After independence the sector saw production collapse faster than employment. The recovery in industry started with the extractive sector, particularly the oil industry, which now accounts for about 45 percent of total industrial output (versus a share of 11 percent in 1990).
49. Manufacturing revival began in its two largest industries, metallurgy and food processing. Growth has been associated with large gains in productivity, reflecting both structural and cyclical factors. As the industry had large under-utilized capacity and excess labor, increases in demand lead to large jumps in productivity. Structural adjustment has led to the release of excess labor and some replacement of obsolete capital with new investment. However, investment in the non-oil industry remains low.
50. While industries such as food processing and machine building have benefited from increasing domestic demand for consumer and investment goods, there was also pressure from import competition. The development of these industries has been based in part on import substitution strategies, with support from high tariffs and financing from the state through subsidies or loans. Such strategies tend to result in lower incentives for the development of competitive industries, as examples of other countries show.42
51. The developments in manufacturing since 2000 were also related to the oil sector. The continuing expansion of oil extraction and a period of high oil prices have resulted in larger demand for manufacturing products and the rebounding of the oil refining and the chemical industries. The development of a diverse and efficient industrial base will depend on complementary infrastructure investments and improvements in the conditions for internal and regional trade. The EBRD transition report for 2002, shows that there is still a need to strengthen competition policy. Improving access to regional and world markets would allow companies to enjoy economies of scale from selling to larger markets.
Sources: National Statistics Agency of Kazakhstan and staff estimates.
52. Agriculture was one of the hardest hit sectors during the initial transition process and has been lagging the rest of the economy during the recovery process. During the first years of independence, the sector went through the disintegration of the rural structures that existed during Soviet times and the collapse of animal and capital stock. The sector also suffered from the extensive use of low quality land during the Soviet times. In 1996 a gradual recovery of the economy began. However, agriculture lagged and had the lowest growth rates, and the worse performance in terms of productivity. Cattle breeding, in particular, showed a very poor performance until recently.
53. Since about 1998, the agricultural sector has started to recover in large part due to the fast growth of private farms and, on a lesser scale, household plots. While private farms and household plots have shown better performance than the declining number of agricultural enterprises,43 growth was mostly based on extensive use of land and labor.44 There have been, however, significant gains in yields for plant growing, associated with the use of better land, improved management following a land code reform in 1999, and increased use of labor.45 Agriculture still faces the problem of over-employment, associated with the large share of subsistence farmers, specially in cattle breeding.46
54. Small- and medium-size farmers have a difficult financial position, a legacy of the initial years of transition. They lack access to credit not only for working capital, but also for investment. As a result, there has been little investment and the capital stock has become largely obsolete. The difficulty in accessing credits has been in part associated with a lack of clear property rights for agricultural land. The recently proposed land code appears to be a step towards introducing private ownership, but there are concerns with the transparency and length of the process. It will also be necessary to increase investment in rural infrastructures, improve water management and marketing institutions, particularly for accessing export markets.
55. In contrast to plant growing, where there has been a development of medium and large farms, cattle breeding is mostly related to household plots or small family farms. These farms tend to face even greater challenges in gaining access to credit or markets. They do not have strong associations to coordinate common interests. In plant growing, there is a much larger concentration of production among a smaller number of farms, specially for grain. There has been also a higher degree of support from the state, such as the creation of credit partnerships to improve credit to farmers. There is also a variety of other tools used by the state to provide inputs, such as seeds, and keep prices at above market levels.
56. The price support mechanism, introduced in 1999 to purchase wheat at above market prices, is now being expanded to other goods and increasing quantities. As a result, there has been a rise in the production of grain, much beyond domestic market needs, and increasing accumulation of stocks of grain. In response, the government, through state owned companies, has started to increase its intervention further, by buying more wheat and exporting it directly to foreign markets. This discourages the emergence of private distribution channels.
Figure II-4:The Agricultural Sector
Sources: National Statistics Agency of Kazakhstan and staff estimates.
AslundA.BooneP. and JohnsonS.2001 “Escaping the Under-Reform Trap,” IMF Staff Papers Vol. 48.
AutyR.2001 “The Political Economy of Resource-Driven Growth,” European Economic Review No. 45.
De BroeckM. and KostialK.1998 “Output Decline in Transition: The Case of Kazakhstan,” IMF Working Paper No.45.
Dervis. K.MeloJ.RobinsonS.1982General Equilibrium Models for Development PolicyCambridge University PressNew York.
DixonP.ParmenterB.PowellA. and WilcoxenP.1992Notes and Problems in Applied General Equilibrium EconomicsNorth-Holland.
EasterlyW. and FischerS.1994 “The Soviet Economic Decline: Historical and Republican Data,” World Bank Policy Research Working Paper No. 1284.
GuptaS.ClementsB.FletcherK. and InchausteG.2002 “Issues in Domestic Pricing in Oil-Producing Countries,” IMF Working paper WP/02/140.
HarrisonW. and PearsonK.1996 “Cómputing Solutions for Large General Equilibrium Models Using GEMPACK” Computational Economics Vol. 9 pp.83–127.
HavrylyshynO.2001 “Recovery and Growth in Transition: A Decade of Evidence,” IMF Staff Paper Vol. 48.
KrugmanPaul1979 “The Narrow Moving Band, The Dutch Disease and the Competitive Consequences of Mrs. Thatcher: Notes on Trade in the Presence of Dynamic Scale Economies,” Journal of Development Economics No. 27.
RobinsonS.Yunez-NaudeA.Hinojosa-OjedaR.LewisJ. and DevarajanS.1999 “From Stylized to Applied models: Building multisector CGE models for policy analysis,” The North American Journal of Economics and Finance No. 10.
SachsJ. and WarnerA.1995 “Economic Reform and the process of global integration,” Brookings Papers on Economic ActivityIssue 1Washington.
SachsJ. and WarnerA.2001 “The curse of natural resources,” European Economic Review No. 45.
Prepared by Paulo Medas.
1996 has been identified, previously, as the start of the recovery after the disintegration of the Soviet Union. According to De Broeck and Kostial (1998) the economy shrank at an average of about 10 percent a year between 1991–95. However, others like Aslund, Boone and Johnson (2001) argue that the large decline in the official economy was partially offset by an increase of the unofficial economy.
There is substantial evidence that natural resource-rich countries tend to have lower long-term growth. However, the reasons for such a “curse” are not fully understood. Sachs and Warner (2001), and Auty (2001), discuss some of the potential reasons. These include “Dutch disease” effects, dominance of rent-seeking activities over productive ones, import-substitution strategies, and lower investment in human capital, among others.
Krugman (1979) presents some arguments why temporary losses of competitiveness, due to an appreciation of the exchange rate, can have long-term effects.
Figure II-1 shows the results of an analysis of sources of growth, based on a growth accounting exercise. The analysis uses the standard assumption that output follows a Cobb-Douglas production function, with employment and capital as factors of production, such that:
Δy=Δa + αΔ1+(1−α)Δk,
where y, a, 1, and k stand for GDP, total factor productivity, employment, and capital respectively; A represents percentage change, α is assumed to be 0.55, an estimate based on the labor share in total income.
In practice, changes in total factor productivity (TFP) reflect not only changes in technology, but also improvements in efficiency due to diverse factors, such as improvements in infrastructures or reforms in general. Havrylyshyn (2001) presents a survey of several studies on transition economies, which show that the initial recovery was generally based on efficiency gains.
These incentives to agriculture production and expansion of land use, consist of several instruments, from loans and subsidies, to schemes to sustain high prices in agriculture.
The analysis of the labor market is also complicated by changes in methodology in the last years. However, the trend seems to be for a recovery in employment in the last 3 years.
The estimates for capital stock changes are based on the data for the capital stock in 1994 and amortization until 2000 from the Statistics Agency of the Republic of Kazakhstan (2002). The data on investment are from the National Accounts and surveys of enterprises. The computation of sectoral total factor productivity assumes a Cobb-Douglas production function with labor and capital as factors of production. Land was also included as a factor of production for agriculture.
The estimates are based on data from the National Statistics Agency of Kazakhstan. The input-output tables for the period 1998–2001, that have been developed by the NSA were an important source. The estimates do not capture the total multiplier impact associated with oil production, but represent the main effects.
Mainly from railway to pipelines.
The DMR model is a multisector general equilibrium model, and has been applied to several countries and is used by the World Bank. For a more detail discussion on the model see Dixon etal (1992).
See Box II-1, which describes the main features of the model. For further information on the model and data used contact email@example.com.
Domestic prices are not allowed to increase by more than half of the world price increase. This reflects the restrictions in the domestic market, which keep domestic prices below market prices, using different mechanisms such as export restrictions and subsidies. Such policies are common to oil producer countries, including Kazakhstan. For a more detailed discussion on these issues see Gupta, et al (2002).
There is also a small increase in exports of oil. Changes in the oil price have a limited impact on oil production because the oil industry in Kazakhstan has only limited capacity to increase production and exports in the short run, and oil prices currently exceed, in general, production costs. See also the earlier chapter on the petroleum sector. Any production increase will come mainly from marginal producers that have higher production and transportation costs.
The Kazakhstani government started to transfer oil revenues to the oil fund in 2001. In 2001–02, the rule has been that 90 percent of oil revenues up to a price of $19/bbl go to the budget, while revenues from higher prices go fully to the oil fund.
The impact of changes in the oil price on GDP is limited partly because the oil industry is still emerging and has limited capability to increase production and exports in the short run. The oil fund and payments to foreign shareholders of the oil companies also contribute to dampening the impact of changes in the oil price.
The model does not capture the potential for large long-run gains in productivity in transition economies, such as Kazakhstan.
The exact impact on consumption and investment depends on the specifications of the model.
Sachs and Warner (1995) present an extensive analysis of the impact of reforms linked with international trade, particularly trade policies, on economic growth for several countries.
Which are the remains of the state enterprises.
After independence, there was a collapse on the use of arable land. Since 1999, there has been an increase in planted areas close to the levels that are though to be economically viable.
The change in coverage of agricultural employment statistics makes an accurate measurement of changes in labor input difficult.
The sector also has a large proportion of part-time and two-jobs workers, who have their household plots for self-consumption or to supplement income.