II. Assessing Fiscal Vulnerability, Fiscal Sustainability, and Fiscal Stance in a Natural Resource Rich-Country2
5. Kazakhstan is a country rich in natural resources. with its crude oil reserves estimated at 30 billion barrels,3 Kazakhstan has slightly higher reserves than Mexico, representing roughly 45 percent of Russia’s reserves, 11 percent of Saudi Arabia’s and almost three times the size of Norway (Text Table 2). Kazakhstan’s daily production in 2000 stood at 0.7 million barrels which is about 10 percent of Russia, the second largest producer in the world after Saudi Arabia, and almost twice the level of Yemen. By 2017, Kazakhstan’s production of crude oil could quadruple to reach a peak of 3 million barrels per day, roughly the level of Kuwait’s daily production at its peak in 1972 and that of Norway in 2000.4 Revenues from the oil sector collected by the government in Kazakhstan has increased from about 5 percent of general government revenues in 1999 to 15 percent in 2000 and are projected to rise to 26 percent of general government revenues in 2001 before falling to 18 percent in 2002 as oil prices are expected to decline. These swings in revenue and the sheer size of the expected oil wealth in the ground pose significant challenges for fiscal and macroeconomic policies.5
|Country||Reserves as of end-2000||Daily production in 2000|
|(In billions of barrels)||(In millions of barrels)|
|Islamic Republic of Iran||99.5||3.70|
|United Arab Emirates||97.8||2.17|
|Western Europe and North America|
|Asia and Pacific|
|OPEC (average) 3/||77||2.52|
See the text for the definition of proven crude oil reserves.
Organization of Petroleum Exporting Countries (OPEC) consists of Algeria, Indonesia, Islamic Republic of Iran, Iraq, Kuwait, Libya, Nigeria, Qatar, Saudi Arabia, United Arab Emirates, and Venezuela.
See the text for the definition of proven crude oil reserves.
Organization of Petroleum Exporting Countries (OPEC) consists of Algeria, Indonesia, Islamic Republic of Iran, Iraq, Kuwait, Libya, Nigeria, Qatar, Saudi Arabia, United Arab Emirates, and Venezuela.
6. It has been recognized for some time that in economies endowed with rich natural resources the conventional assessment of fiscal vulnerability, fiscal sustainability and fiscal stance can often be misleading.6 This recognition is driven by the increasing awareness of the characteristics unique to these economies. These are: (i) natural resource wealth can be a significant source of government revenues at least for an extended period of time and is often not treated as part of overall government wealth and national wealth;7 (ii) natural resource is non-renewable and its size is often subject to considerable uncertainty; 8 (iii) prices of resource-based commodities (e.g., oil, copper) are volatile, with no discernible trends or cycles;9 (iv) economy can be subject to so-called “Dutch disease” phenomenon of real exchange rate appreciation, loss of competitiveness in non-resource intensive tradable sector and perhaps de-industrialization;10 and (v) extensive research has shown that, even after controlling for other factors, resource-rich countries tend to grow slower than resource-poor countries,11 and that land-locked countries grow at a lower rate than countries that are not landlocked.12 The last point is a significant finding as Kazakhstan is the largest landlocked country in the world. A growing number of studies of fiscal policy in resource-rich countries have started incorporating some of these characteristics in assessing various aspects of fiscal policy.13
7. An oil-rich country has to solve a difficult portfolio problem in the short run as well as the long run. Should it leave for future generations a significant reserve of oil deposits (implying a slow rate of depletion of reserves), financial wealth accumulated from years of sustained oil extraction (suggesting a fast rate of depletion), or a higher productive capacity as a result of a sustained build up of the economy’s capital stock, broadly defined to include human as well as physical capital. This chapter is an attempt to solve only part of the problem.
8. In particular, the purpose of the chapter is to conduct a comprehensive assessment of fiscal vulnerability, fiscal sustainability, and fiscal stance in Kazakhstan focusing on the government’s nonoil budget balance while taking into account some of the key characteristics of resource-rich countries. Simply stated, the key issue is how much the government should consume and save out of oil revenues and how these decisions should change in response to changes in key aspects of fiscal vulnerability in Kazakhstan.14 The private sector response is obviously of paramount importance in such a setting as oil wealth is part of national wealth, and fiscal policy would influence private sector activities and vice versa. However, an analysis of the private sector response and its role in the economy is beyond the scope of this chapter.
B. Measuring Fiscal Stance in a Natural Resource-Rich Economy
9. An important lesson that has emerged from the previous studies of fiscal policy in natural resource-rich countries is that non-resource fiscal balance (i.e., the overall balance excluding resource receipts to the government) or the nonoil fiscal balance (i.e., the overall balance excluding oil revenues) in the case of oil-rich countries is the key variable for assessing the fiscal stance which, as will be shown in the chapter, is an important indicator for assessing fiscal sustainability and fiscal vulnerability; see, for example, Tersman’s (1991) and Bascand and Razin’s (1997) analyses of Norway and Indonesia, respectively.
10. There are at least four reasons why the emphasis on nonoil fiscal balance is important.15 First, if during periods of rising oil prices, oil revenues are used to finance permanent expenditure increases, then a permanent demand is placed on the budget which would be difficult to reverse during periods of falling oil prices or when oil reserves are exhausted. In the absence of compensating fiscal adjustment, the resulting nonoil balance would be unsustainable which would be the case whether or not oil reserves are depleted. Cost of fiscal adjustment (e.g., a higher tax burden) may be shifted to future generations that have to live with lower oil reserves or none at all.
11. Second, in oil-rich countries the nonoil balance is a better indicator of the subsequent growth performance than the overall balance. Extensive research has shown that, after controlling for other factors, higher deficit is associated with lower economic growth.16 This finding would hold even stronger when the nonoil deficit is the measure of deficit rather than the overall balance; and in circumstances in which the budget relies heavily on oil receipts. The nonoil deficit in such oil-dependent economies would be higher than otherwise would be the case; hence, growth would be lower. The poor growth track record of resource-rich countries has shown that their natural resource wealth has not been the blessing that one might have expected.17
12. Third, the emphasis on the nonoil balance rather than the overall balance would lead to a closer scrutiny of the nonoil revenue and the spending sides of the budget, and points towards sources of fiscal adjustment. On the revenue side, the emphasis would direct attention of the policymakers to the issue of broadening the nonoil tax base, which is crucial as oil reserves are non-renewable. On the spending side, closer attention would be placed on scrutinizing the productivity of each additional spending which the country cannot afford to live without even when oil reserves have not been depleted.18 A high nonoil deficit can also be an indicator of unproductive and wasteful spending which not only does not enhance the productive capacity of the country but may even lead to persistently higher nonoil deficits in the subsequent budgets. This is a cause for concern in low-income, resource-rich economies, given their weak institutional capacity for screening projects for public investment, budget implementation and their weak governance structure.19
13. Finally, for policymakers who may have a shorter horizon, the nonoil balance is a useful indicator for assessing the short-run fiscal policy stance, particularly in the transition from an oil-rich state to an oil-poor state and vice versa. The nonoil balance can be used to assess whether a proposed fiscal policy is too loose or too tight at a given point in time or whether fiscal policy has been loose or tight over time. The latter is particularly important when a significant oil deposit is discovered or when oil reserves are close to being depleted. In terms of a positive analysis, movements in the nonoil balance are an important guide for policymakers for developing the subsequent fiscal stance during the transition process.
14. Any assessment of fiscal policy stance, whether using the overall balance or nonoil balance, requires a normative framework. Building on the lead of previous studies of resource-rich countries, the framework used in this chapter relies on the government’s intertemporal budget constraint and the Permanent Income Hypothesis (PIH) of consumption. After describing some essential elements of the economics of natural resources and optimal consumption of such resources under a PIH framework, the simple analytics of the PIH is presented and then applied to Kazakhstan. The framework is also used to formally derive the relationship between nonoil balance and government consumption out of oil wealth.
C. Economics of Natural Resources Within a PIH Framework
15. The framework used for the analysis of this chapter is the PIH of consumption. This framework has been applied previously to varying degrees to ten resource-rich countries.20 The PIH has several desirable properties, which makes it suitable for the purpose of this chapter. First, the hypothesis is forward-looking as it assumes that government does not spend out of its current income (or resources), but out of its permanent income or total wealth. In its simplest form, government’s permanent income is merely the annuity value of its net wealth, the discounted net present value of future flows of earnings available to the government. This approach imposes fiscal discipline on the government in that its spending is guided by the available total net wealth, sum of financial and non-financial wealth defined to be net of debt. Without this discipline, value of total wealth can be eroded over time if natural resource wealth is consumed too fast and financial wealth is accumulated too slowly. If this happens to be the case, interest of future generations will be sacrificed at the expense of the current generation. Moreover, it is also possible, as illustrated by the history of some oil exporting countries, that even interest of a large segment of the current generation may not be served if prudent macroeconomic policies and proper safeguards are not put in place to protect the integrity of the rent extracted from a finite and exhaustible natural resource.21
16. The second property of PIH is that the government is assumed to be able to freely borrow from and lend to international and domestic capital markets in anticipation of future oil revenues.22 This property, along with the assumption that the government is no more or less patient than the discount rate implied by the prevailing interest rate on international capital markets, ensures that government smoothes its consumption of oil wealth over time, thus providing a steady level of public services. These two assumptions are admittedly too strong, but they are needed to produce a simple rule and a benchmark for evaluating alternative government consumption of natural resource wealth. For example, these two assumptions imply that in response to a boost in its permanent income, say, due to the discovery of higher natural resource wealth (e.g., a new oil field), government can choose to raise its spending or lower taxes in line with permanent income equivalent of change in wealth and retire the debt thus accumulated with future oil revenues. Third, the path of fiscal policy dictated by the PIH is by construction fiscally sustainable. This feature allows for an assessment of alternative fiscal policies and point towards the nature and size of fiscal adjustment needed to bring the fiscal policy back on track in response to changes in the economic environment.
17. Given the PIH and assuming that the government is no more or less patient than the discount rate implied by the prevailing interest rate on international capital markets, government should consume at most the real interest rate on its total wealth or the permanent income.23 The assumption made on the rate of impatience (i.e., government’s rate of time preference) refers to the celebrated case of permanent income as discussed by Friedman (1957). It is the permanent rent obtained each period from extraction of the natural resource and is equal to the real interest rate times initial oil wealth. Therefore, the nonoil deficit that is indefinitely sustainable should be no higher than this rent. Essentially, government consumes at most the annuity value of oil wealth, thus keeping oil wealth constant forever and providing the same real level of services for future generations as the current one.
18. Alternative assumptions on the rate of impatience imply either a decreasing or an increasing path for government consumption out of oil wealth, which would imply that one generation is sacrificed at the expense of the other as far as use of wealth from natural resources is concerned.24 In this regard, intergenerational equity considerations play an important role, i.e., the current generation should leave behind sufficient resources to allow future generations at least the same level of consumption out of natural resource wealth as that enjoyed by the current generation. This path corresponds to Friedman’s so-called constant consumption path, the highest consumption that can be enjoyed indefinitely by all generations without increasing the country’s debt and depleting its total wealth.25
19. An argument can be made that the current generation does not need to leave the same level of resources for future generations (i.e., the constant consumption path) since future generations will be better off. Taking the simplest case of no population growth and zero total factor productivity (TFP) growth or zero technical progress, this argument essentially assumes that the decline in the stock of exhaustible natural resource capital by the current generation is more than offset by future accumulation of human capital and reproducible physical capital, the two proximate sources of growth.26 However, this assumption is not supported by the growth experiences of resource-rich countries.27 This experience is consistent with the observation that in resource-rich economies, the rate of depletion of natural resource has been too fast relative to the speed with which other factors of production have been accumulated.28 Moreover, future generations will be better off only if appropriate economic policies, incentives and institutions are put in place to ensure the required accumulation of factors of production.
20. Allowing for a positive and economically significant TFP growth would naturally weaken the case for intergenerational equity and constant consumption path, but extensive recent research has shown that economic policies, incentives and institutions that drive TFP growth are also the same factors that drive accumulation of factors of production.29 This new body of research has shown that TFP growth as well as factor accumulation are all endogenous variables. Therefore, in theory a positive and economically significant TFP growth weakens the intergenerational argument, but it does not eliminate it. In practice, the TFP argument is weak, given the historical growth performances of resource-abundant economies.
21. More recent formal economic theories have also studied the intergenerational equity argument. Gerlach and Keyzer (2001), for example, analyze three policy prescriptions for a resource-rich country. First, a “zero extraction” policy; second, a “grand fathering” policy that endows the current generation with all the resources, and that ensures efficiency but cannot prevent a persistent decline in lifetime utility from one generation to the next. Third, a “trust fund” policy, where future generations receive claims for the natural resource. Of the three, only the trust fund ensures efficiency and protects welfare of all generations.
22. Although the constant consumption path provides a useful benchmark for analysis, and one adopted here as well, the framework analyzed so far does not allow for growth in consumption out of oil wealth. Under the version of permanent income framework discussed so far, oil wealth can decline relative to nonoil GDP. Indeed, under the PIH studied so far any positive consumption growth path (with a positive trend above the Friedman’s zero growth rule) will either deplete oil wealth too rapidly, lead to costly borrowing to finance the extra consumption, lead to unproductive investment or a combination thereof. None of these paths are sustainable.
23. In contrast, standard models of optimal growth, such as the neoclassical model with exogenous TFP growth, the steady state balanced growth path implies that consumption and output grow at the same rate as the exogenous TFP growth.30 In the next section this possibility is also explored. It is assumed that real consumption out of oil wealth grows in line with real nonoil GDP, an assumption that admits future generations would be better off by the mere fact that they would come later. Under this additional assumption and continuing to maintain the PIH, maintaining oil wealth constant relative to nonoil GDP requires that the government consumes each period not the real interest rate, but the growth-adjusted real interest rate. Hence, permanent income would now be the growth-adjusted real interest rate (i.e., real interest rate minus real nonoil GDP growth rate) times initial oil wealth. However, for a growing consumption to take place, it is still the case that appropriate set of economic policies and institutions needs to be adopted such that they do bring about the assumed “exogenous” TFP growth and result in the build up of non-financial wealth that not only replaces the foregone oil in the ground, but ensures that total wealth grows in line with real nonoil GDP. These are much stronger requirements than the benchmark permanent income case which illustrates perhaps the simplicity and power of the benchmark case.
D. Analytics of the PIH and its Application to Kazakhstan
24. The starting point of analysis is the government intertemporal budget constraint which can be written as:
where Ptr is the rent accrued to the government from each unit of oil production during period t, and Et, is the stock of proven oil reserves at the beginning of time t. Therefore, Et-1 - E1 represents oil production during period t. Per period total rent, Ptr (Et-1 - Et), represents taxes paid to the government which in the case of Kazakhstan correspond to corporate income taxes, personal income taxes, royalties, local taxes, withholding tax on interest and dividend, oil bonuses, payments related to production sharing arrangements, VAT, import duties, excises and social tax; NOt stands for nonoil tax and non-tax revenues (excluding interest income on savings on oil revenues); At for the stock of net government financial assets at the beginning of period t; it for the interest rate during time t; and Ct, for the government’s total expenditure and net lending. Equation (1) can also be written as:
that is, nonoil balance (or nonoil savings), the left side of equation (2), is equal to the change in total net government wealth, with the latter consisting of change in the net financial claims on the government and change in the oil wealth (oil in the ground). If nonoil balance was zero for all time periods, then total government net wealth would not change; only its composition would; oil is taken out of the ground and sold; various oil taxes are collected which are then converted to financial wealth and saved at the interest rate it to augment the stock of net wealth for the next period. This corresponds exactly to the permanent income framework discussed in the previous section as total wealth is kept constant and the government consumes only the permanent income each period (itAt-1). To a first approximation, this intuition is correct, but its exact solution is somewhat different and needs to be worked out. An important implication of the PIH of consumption is that it does not distinguish between oil and nonoil wealth, as marginal propensity to consume out of each is the same.
25. The case of a zero nonoil saving or a balanced nonoil fiscal balance corresponds to the permanent income framework under the twin assumptions of zero inflation and zero population growth. Positive nonoil savings (or lower nonoil deficits) are required for a positive inflation rate and a growing population such that the same level of services is offered to each generation. In other words, what should be nonoil balance such that total real wealth per capita stays constant indefinitely? In this case, the exact analytics of how much should be saved and consumed out of per period oil revenues is as follows. First, the concept of oil wealth and its dynamics needs to be defined and quantified.31
26. Value of oil in the ground at the beginning of period t, denoted as Wt, is defined as present value of cash tax revenues from the oil sector. Hence,
where Ot, denotes tax revenues from oil extracted during time t or Ptr(Et-1 - Et), T is the period when oil reserves are depleted, and it is the per period nominal interest rate which is assumed to vaiy over time.32 The base period for discounting tax revenues is time t, which corresponds to year 2000 in the calculations. Therefore, Ot and its present value coincide in period t. The present value calculation in equation (1) implies the following dynamics for the value of (residual) oil reserves in the ground:
27. Dynamic path of Wt depends on the following assumptions about the economy and the oil sector: path of interest rate, stock of proven oil reserves, per period extraction of oil reserves, price of oil, date of depletion of oil reserves (T), and the nature of tax regime and cost conditions in the oil industry. These assumptions have been incorporated in a detailed model of oil sector in Kazakhstan, which has been developed jointly by the staff of the World Bank and International Monetary Fund.
28. This model is used to generate data on Ot. Under these assumptions, the current proven stock of oil reserves in Kazakhstan will be depleted by 2048 and value of oil reserves or oil wealth available to the government as of end 2000 is estimated at about $59 billion (289 percent of 2001 GDP).33 The exact path of Wt, derived from equation (4), is shown in (Figure 2). Although oil reserves in the ground will be gradually depleted with each additional barrel of oil taken out of the ground, value of the residual oil in the ground can be increasing, as shown in (Figure 2), for two reasons.34
29. First, given the oil model, growth in tax revenues in Kazakhstan from oil extraction can and does exceed the interest rate used for discounting these revenues.35 Second, each point in (Figure 2) represents present value of the residual oil reserves in the ground where the year associated with that point, and not 2000, is the base year used for present value calculation.36 The hump-shaped pattern of value of oil in the ground mirrors the pattern of oil production (Figure 3), with both peaking in 2017. Despite the fall in oil production thereafter, tax revenues from oil continue would continue to rise because of the rising nominal oil prices. Value of the residual oil in the ground starts declining after 2017 as oil reserves are being depleted until year 2048 when tax revenues from the production in the year is exactly the value of the residual oil in the ground in that year. Hence, value of the residual oil reserve in 2049 is zero. From 2048 onwards, stock of wealth consists entirely of financial wealth from accumulation of savings from oil. In 2049 for example, wealth in current U.S. dollar represents 866 percent of 2001 GDP.
Figure 2.Kazakhstan: Value of Oil in the Ground, 2001–48 Figure 3.Kazakhstan: Oil Production 2001–48
30. At any point in time, total (net) government wealth (TWt+1) consists of the value of residual oil in the ground and financial net wealth:
which would be equal to Wt+1 + St0 (1+it+1) where St0 is per period saving from oil revenues Ot such that this saving and consumption from oil revenues, denoted as Ct0 exhaust oil revenues, i.e., Ot = St0 + Ct0. Equating the two wealth concepts defines savings,
31. Equation (6) illustrates that with zero initial wealth, saving is merely the present value of future net financial wealth. Writing out equation (5) for the next time period (t + 2) and carrying over oil savings from periods t and t + 1 to t + 2 implies:
Therefore, given that equation (5) holds for all time periods, equation (7) defines financial net wealth at the beginning of period t + 2 and is given by
Upon substituting for St0 from Equation (6) in equation (8) and rearranging:
which holds for the period t+1 onwards. So far these formulations respect the accounting identities and do not describe any behavioral relationship. In particular, they do not in any way ensure that the resulting saving path (St0, St+10, …) is that dictated under the PIH. The saving path under the PIH would ensure that total real net wealth is kept constant in per capita terms. Therefore, total net wealth has to grow in line with inflation and population growth. Thus,
where nt and πt are population growth and inflation during period t, respectively. For the initial period t it is assumed that37
and from period t onwards, evolution of total net wealth is governed by equation (10). Consumption of oil wealth is given by Cto = Ot - Sto which upon substituting Sto from (6) and using (4), (5), and (10) to put every term in terms of the initial value of oil in the ground, the following expression is obtained:
where TWi+j is given by equation (10). Equations (13) and (14) are the optimal consumption under the PIH. They can also be written in terms of real interest rate, rt. For example, equation (12) can be written as:
which makes use of the usual relationship linking inflation, nominal and real interest rates. Upon an additional assumption of constant inflation rate and constant real interest rate in the steady state (i.e., πt+1 = πt = π and rt+1 = r), optimal consumption of oil wealth under the PIH, equation (14), can be written as:
32. Thus, the government should only consume the real annuity value of wealth adjusted for populating growth so that per capita real wealth remains constant over time. Assuming a zero population growth rate in equation (15) produces the familiar textbook expression of optimal consumption under the PIH.38
34. As expected, equation (15) and (16) state that for wealth per capita to be constant in real terms, each generation should consume population growth-adjusted real interest rate or population-cum-inflation adjusted nominal rate, respectively.
35. Maintaining oil wealth per capita constant means that future generations would benefit the same as the current generations from oil reserves. If the country consumes more than what is dictated by the PIH, then oil wealth would decline faster and the current generation would naturally receive a higher share of wealth than future generations. Conversely, consuming less than the level indicated by the PIH implies that current generation is saving more and leaving more of oil wealth to future generations, thus wealth accumulation is higher.
36. Dynamics of wealth in Kazakhstan, corresponding to equations (4), (5), (8), (9), (10) and (11) is shown in (Figure 4). As expected, it shows that financial net wealth has to grow as oil reserves are being depleted such that total wealth grows in line with inflation and population growth rate.
Figure 4.Kazakhstan: Dynamic of Wealth, 2001–48
Source: The Kazakhstan authorities; and Fund staff estimates and projections.
37. As discussed in the previous section, there is nothing in the above framework that would preclude the value of total net wealth from declining when measured relative to real nonoil GDP. This is a simple mathematical implication and indeed the case in Kazakhstan as well. In this formulation, the emphasis is on real nonoil GDP rather than real GDP because the long-run balanced growth path would refer to growth in real GDP when oil reserves are depleted, i.e., it is the sustainability of real nonoil GDP growth that would matter in the long run when oil production and its associated real activities no longer contribute to real GDP and the entire real GDP is nonoil.
38. Therefore, allowing total net wealth, TWt, to grow in line with inflation and real nonoil GDP implies replacing equation (10) with:
where xt is growth of real nonoil GDP.
Optimal consumption can then be written as
39. This is the version of equation (15) in the steady state, which simply states that optimal consumption should take place out of growth-adjusted real interest rate so that enough real wealth is set aside to allow wealth to grow in line with real nonoil GDP. Under this rule, less consumption would take place; hence, more of oil revenue is saved.
40. Equation (18) can also be written in terms of nominal interest rate, making use of the Fischerian relationship:
41. Equation (19) is the counterpart of equation (16) when wealth grows in line nonoil GDP.
42. (Text Table 3) shows the summary of evolution of wealth, nonoil deficit and the implied saving rate over the 2001–48 period, given the PIH under the two cases of constant real wealth per capita, the benchmark case, and constant ratio of wealth to nonoil GDP. The table also shows two sub-periods, 2001–15 and 2016–48, to analyze the implications of the hump-shaped pattern of value of oil in the ground, which occurs in 2015.
|Case 1: Constant Wealth Per Capita 1/||Case 2: Constant Wealth to Non-Oil Nominal GDP Ratio|
|Growth in real per capita wealth (in percent) 2/||0||0||0||2.3||1.7||1.9|
|Growth in wealth-nominal non-oil GDP ratio (in percent)||-2.5||-1.7||-1.9||0||0||0|
|Oil revenue per capita (current $US)||125||362||288||125||362||288|
|Oil consumption per capita (current $US)||169||267||238||55||276||207|
|Implied saving rate out of per period oil|
|revenue per capita (in percent)||-35||26||17||56||24||28|
|Non-oil fiscal deficit to nominal GDP (in percent)||7.8||5.5||6.3||0.8||5.5||4.1|
|Non-oil fiscal deficit to nominal non-oil GDP (in percent)||11.2||7.3||8.5||1.7||7.1||5.4|
This is also referred to as the benchmark case in the text.
In 2000 prices
This is also referred to as the benchmark case in the text.
In 2000 prices
43. Given the estimates of about $59 billion in the value of oil in the ground and the current Kazakhstan population of about 15 million, the level of wealth per capita that would be kept constant forever stands at $3,974.41 Keeping wealth constant at this level implies that oil revenue per capita would average to about $288 of which $238 is consumed by the government on average, implying a saving rate of 17 percent. By contrast, under the current NFRK rules and given the oil price assumption, 10 percent of oil revenues are saved. The path of saving is, however, completely different during the sub-periods as expected under the PIH. With a rising wealth in the ground during the first sub-period, and a low initial financial wealth, and assuming perfect capital markets, the government can borrow against the oil wealth, dissaving for almost the entire period 2001–15, and then start saving thereafter and paying off its debt. This behavior may stand in marked contrast to a simplistic interpretation of the life cycle hypothesis of consumption, according to which saving would occur as oil wealth is rising and dissaving would occur when oil wealth is falling. But this interpretation is not correct for two reasons. First, under a PIH, the government consumes and saves out of its total net wealth and not just oil wealth in the ground. Significant consumption smoothing occurs under a PIH as the government can borrow against the oil wealth in the ground since its financial wealth is too small by comparison. Second, in contrast to an individual, government lives forever, an assumption built into the PIH.42
44. Under the PIH and assuming that wealth is kept constant in per capita terms, the implied saving rate amounts to–35 percent and 26 percent during 2001–15 and 2016–48, respectively (Text Table 3). Therefore, under the current NFRK rules, too much is being saved in the first sub-period and too little in the second sub-period.
45. This saving and dissaving path under the PIH can also be shown in terms of GDP. The nonoil fiscal deficit would average at 7.8 percent of nominal GDP during the 2001–15 period and is expected to decline to 5.5 percent of nominal GDP in the 2016–48 period, averaging at about 6.3 percent of nominal GDP during the entire period. This figure is above the nonoil deficit in Kazakhstan as Kazakhstan’s highest nonoil deficit was at 5.7 percent of nominal GDP recorded in 1999, given available data on the brief history of oil in Kazakhstan.43 By implication, estimates of nonoil fiscal deficit in 2000, and in 2001 as projected, indicates that that the government is taking a conservative view of the size of the oil wealth at least for the 2001–16 period, a view that is justified given that estimate of wealth is sensitive to the many assumptions made, principally the size of the oil reserves in Kazakhstan, the evolution of oil prices, and the as-yet-to-be-announced discovery of the Kashgan field.
46. Under the benchmark case, the implied wealth will decline continuously in relation to nonoil GDP as the latter grows faster than wealth (see Text Table 3). Oil wealth represents about 401 percent of nominal nonoil GDP in 2000.44 Keeping total wealth constant at this level indefinitely would produce a path for and size of saving from each barrel of oil, which is entirely different from the benchmark case. For example, a much higher saving rate is required, estimated at an average of 28 percent out of every barrel of oil for the period 2001–48. The saving rate is much higher in the period 2001–15 (at 56 percent) than 2016–48 (at 24 percent) as initial financial nonoil wealth is too low for total wealth to be constant in relation to nonoil GDP. As a result, this shortfall has to be compensated out of higher saving from oil wealth and per period oil revenue. However, as saving is accumulated over time sufficiently from oil wealth and financial wealth has been built up accordingly, then the saving rate out of each barrel of oil can decline. As a result, consumption per capita (out of per period oil revenues) increases significantly, rising from $55 in the 2001–15 periods to $276 in the 2016–48 periods, a five-fold increase. This result is expected since by maintaining wealth constant relative to nonoil GDP, the objective of keeping a constant consumption per capita had to be given up. Nevertheless, consumption smoothing still occurs, as expected under a PIH, but the yardstick for this evaluation has changed to measuring it relative to real nonoil GDP rather than the population.
47. The objective of maintaining wealth constant in relation to nonoil GDP also implies a different size for and path of nonoil fiscal balance than the benchmark case. Nonoil fiscal deficit would average at about 4 percent of nominal GDP over the 2001–48 period and rising in the second sub-period, opposite of what occurs in the benchmark case. An important implication of keeping wealth constant in relation to nonoil GDP is that the government has to run a surplus for the first decade of the millennium and a balanced nonoil fiscal balance for the 2001–15 period. The temporary surplus in the first decade is consistent with the theoretical models of Alier and Kaufman (1999) and Engel and Valdes (2001), and with the interpretation that the government is perhaps more patient under the constant wealth to nonoil GDP case than the benchmark case.
48. Fiscal vulnerability in response to changes in oil prices and interest rate points towards considerable fiscal adjustment and changes in nonoil fiscal balance under the PIH. If the oil price drop $1 permanently from baseline oil price path (approximately a 5 percent annual reduction during 2001–06), then oil wealth would drop to about $51 billion (250 percent of 2001 GDP) from $59 billion (289 percent of 2001 GDP) and the sustainable nonoil fiscal deficit would fall to about 5.3 percent of GDP from 6.3 percent of GDP.45 Failure to adjust permanently would imply that the resulting fiscal path is unsustainable. These results also show that permanent income, hence, the nonoil fiscal balance, is more sensitive to oil price movements than is oil wealth. The analysis also shows an important aspect of the PIH; revisions in wealth would imply revisions to fiscal stance, which would not have been the case if fiscal stance had not been based on the permanent income.
49. If the interest rate is assumed to be lower permanently by one percentage point from the baseline 6 percent, then the estimate of wealth would rise to almost $72 billion (353 percent of 2001 GDP) from $59 billion (289 percent of 2001 GDP) and the sustainable nonoil fiscal deficit would fall to 5.6 percent of GDP from 6.3 percent of GDP. The drop in the nonoil fiscal deficit, in spite of the increase in wealth, is because permanent income declines in response to lower interest rate as the government consumes the annuity value of the wealth using a lower interest rate. The analysis also shows that nonoil fiscal balance is more sensitive to oil prices changes than changes in interest rate, with the elasticity of 3.2 and 0.7, respectively.
50. If oil prices drop permanently to $15 per barrel from the base line path, oil wealth would decline to about $29 billion (143 percent of 2001 GDP) from $59 billion (289 percent of 2001 GDP) and the sustainable nonoil fiscal deficit would be 3 percent, almost half the baseline scenario. This analysis underscores the point made earlier of the need to base fiscal policy on a conservative estimate of nonoil deficit. The elasticity of nonoil deficit with respect to this change in oil price is 2 which is much lower than the elasticity of 3.2 when oil prices dropped by only $1, implying that these elasticities depend on the initial value of oil price and presence of some non-linearities. It should be noted, however, that elasticities for such large changes in wealth, oil prices and deficit should be treated with caution.
51. This chapter used the permanent income hypothesis (PIH) to assess fiscal vulnerability, fiscal sustainability and fiscal stance in Kazakhstan. It was argued that in resource-rich countries, assessment of fiscal stance should rely on the non-resource fiscal balance, the overall balance excluding natural resource receipts accrued to the government, rather than the overall balance including these receipts. The chapter focused exclusively on the oil sector in Kazakhstan, but the analysis applies equally to other primary commodities produced in Kazakhstan, which are taxed by the government. In applying the PIH, two rather different objectives about future evolution of oil wealth were analyzed. The first objective assumed that the government would keep wealth constant in per capita terms, thus leaving future generations with the same real level of public services financed from oil revenues as the current generation. The second objective assumed that wealth would be kept constant in relation to nonoil GDP, thus assuming that society would be richer in the future as the economy grows at its “exogenous” positive TFP growth rate. However, it was also shown that for any positive TFP growth rate, pursuing the second objective would come at a cost of higher foregone current consumption out of the oil wealth and higher saving than indicated under the first objective. However, regardless of which objective is pursued or some variation of the two, appropriate macroeconomic and structural reforms need to be put in place to ensure a steady rise in per capita income, albeit a higher TFP growth is need under the second objective. Although this observation may seem to be straightforward at first, given Kazakhstan’s recent growth track record, the history of resource-rich countries has shown quite the opposite and the challenges are even more for Kazakhstan as its makes its transition to a market economy. The main lessons of the chapter can be summarized as follows:
The PIH provides a useful rule of thumb for evaluating fiscal stance; that the government should consume at most the annuity value of its total wealth. The PIH imposes fiscal discipline on the government by requiring the government’s total net wealth, hence, its permanent income, be the basis for its fiscal policy rather than current resources available to the government. The case of keeping wealth constant in per capita terms provides an easy rule for setting long-term fiscal policy and is the most-heavily used assumption in studies of economies with significant natural resources.
The PIH hypothesis provides a useful benchmark for the analysis, but its assumptions are perhaps too strong. However, these assumptions are needed to produce an-easy-to- use rule of thumb for policymakers. The rule should be used knowing its strengths and weaknesses along with the knowledge of the other aspects of the Kazakshtani economy (see below). The chapter provided a lengthy discussion of the limitations of the PIH.
Kazakhstan is undergoing a significant pension reform. Given the uncertainties associated with any implicit pension liability and size of the oil wealth, the PIH needs to interpreted with caution when deciding on the short to medium-term fiscal stance; higher saving than that implied by the PIH may be required, thus building an additional precautionary motive for saving. In fact, one motivation behind the significant build up of oil wealth in Norway’s State Petroleum Fund has been to save for the future pension liabilities that grow with an aging population. This could potentially be a problem in Kazakhstan, given its low birth rate and slow population growth rate.
The PIH shows that Kazakhstan’s nonoil fiscal balance in recent years has been well within the long-run sustainability path as measured relatively to GDP; this prudence in fiscal policy stance is welcomed, given the uncertainties associated with discovery of the new oil reserves, their timing, and future evolution of oil prices, to name a few. The extensive sensitivity analysis conducted in the chapter provides further support to adopting a cautious nonoil fiscal stance.
Long-term fiscal policy as dictated by the PIH was shown to be vulnerable to swings in oil prices and interest rate, as these require drastic changes in nonoil fiscal stance. This is indeed a major advantage of the PIH since swings in oil prices and interest rate result in revisions in oil wealth. The record of many current oil-producing countries suggests that governments may follow a “quasi-asymmetric” fiscal policy stance; they respond to upward revisions in wealth during price booms by spending more, but find it hard to respond symmetrically to downward revisions in wealth during price busts. Prudent short-run fiscal stance should also pay attention to other macroeconomic objectives such as inflation and the balance of payment.
Formation of the National Fund for the Republic of Kazakhstan (NFRK) is a welcomed step for saving oil revenues for the future and for sterilizing oil windfalls. In this regard, analysis of the PIH was conduced under the assumption of a constant real exchange rate. Deviations from this assumption will require reassessment of saving implied under the PIH and underscores adoption of complementary policies to prevent the occurrence of the Dutch disease.
Under the current NFRK rules, and assuming that these rules are maintained indefinitely into the future, the PIH implied that Kazakhstan is saving too little of the oil revenue in the NFRK. Future simplification of the complex NFRK rules can benefit from the PIH’s easy-to-use rule for thumb for saving.
Kazakhstan is the largest landlocked country in the world; it does not have easy access to markets for trade and has significant infrastructure needs, but it enjoys favorable initial fiscal stance, and a low debt-GDP ratio and. The challenge is how to build its capital stock and develop its infrastructure, while balancing in this regard the need to accumulate financial wealth and use the proceeds from the depletion of its oil reserves with borrowing (from domestic markets, international capital markets, international financial institutions), and entering in joint ventures to attract foreign direct investment.
A main lesson of economic theory is that given the larger share of physical capital in output than non-renewable natural resource capital, and given the low initial capital stock of a typical low-income, resource-rich country, earlier generations should use up the natural resource quite fast, while building the capital stock in turn (Solow, 1974; Stiglitz, 1974). But to do so also requires that the current and future generations adopt complementary policies that diversify the economy, create a productive capital stock, defined broadly to include physical and human capital, and a skilled labor force which bring about a sustained rise in TFP growth in the nonoil economy.
AmuzegarJahangir1999Managing Oil Wealth OPEC’s Windfalls and Pitfalls (London: LB. Tauris Publishers).
AutyRichard. M (1998) “Resource Abundance and Economic Development, Improving the Performance of Resource-Rich Countries” WIDER Research for Action #44 (Helsinki: The United Nations University).
BarroRobert J (1999) “Notes on Growth Accounting,” Journal of Economic Growth Vol. 4 (June) pp. 119–137.
BlanchardOlivier J. and StanleyFischer (1989). Lectures on MacroeconomicsCambridge: The MIT Press).
ChalkNigel1998 “Fiscal Sustainability with Non-Renewable Resources,” IMF Working Paper 98/26 (Washington D.C.: International Monetary Fund).
BascandGeoffrey and AssafRazin1997 “Indonesia’s Fiscal Position: Sustainability Issues,” in Macroeconomic Issues Facing ASEAN Countriesed. JohnHicklinDavidRobinson and AnoopSingh (Washington D.C.: International Monetary Fund).
BarroRobert1979 “On the Determination of Public Debt,” Journal of Political Economy 85 No.5 pp. 940–971.
BarroRobert1989 “The Ricardian Approach to Budget Deficits,” Journal of Economic Perspectives Vol. 39Spring) pp. 37–54.
BarroRobert J2000 “Notes on Growth Accounting,” Journal of Economic Growth Vol.4 (June) pp. 119–137.
BarroRobert J. and XavierSala-I-Martin1995Economic Growth (New York: McGraw-Hill, inc.).
BlanchardOlivier Jean and StanleyFischer1989Lectures on Macroeconomics (Cambridge: The MIT Press).
CashinPaulC.J.McDermott and A.Scott (1999) “Booms and Slumps in World Commodity Markets,” IMF Working Paper 99/155 (Washington D.C.: International Monetary Fund).
DavisJeffreyRolandoOssowskiJamesDaniel and StevenBarnett2001“Stabilization and Savings Funds for Non-Renewable Resources Experience and Fiscal Policy Implications” Occasional Paper no. 205 (Washington D.C.: International Monetary Fund).
De BroeckMark and KristinaKostial (1998) “Output Decline in Transition: The Case of Kazakhstan,” IMF Working Paper 98/45 (Washington D.C.: International Monetary Fund).
EngleEduardo and RodrigoValdes (2000) “Optimal Fiscal Strategy for Oil Exporting Countries,” IMF Working Paper 00/118 (Washington D.C.: International Monetary Fund).
EasterlyWilliam and RossLevine (2000) “It is not Factor Accumulation: Stylized Facts and Growth Models,” World Bank Discussion Paper Series no (Washington D.C.: The World Bank).
ChalkNigel and RichardHemming1999 “What Should be Done with a Fiscal Surplus?” International Economic Policy Review Vol. 1ed. PaulMassonPadmaGotur and TimothyLane (Washington D.C.: International Monetary Fund).
FischerStanley (1993) “Role of Macroeconomic Factors in growth,” Journal of Monetary Economics Vol. 32 (December) pp. 485–512.
GelbAlan H. and others (1988) Oil Windfalls: Blessing or Curse (New York: Oxford University Press).
GerlaghReyer and Michicl A.Keyzer2001 “Sustainability and the Intergenerational Distribution of Natural Resource Entitlements,” Journal of Public Economics Vol. 79 pp. 315–341.
GersonPhilip and DavidNellor1997 “Philippine Fiscal Policy: Sustainability, Growth and Savings,” Macroeconomic Issues Facing ASEAN Countriesed. JohnHicklinDavidRobinson and AnoopSingh (Washington D.C.: International Monetary Fund).
FriedmanMilton (1957). A Theory of the Consumption Function (Princeton: Princeton University Press).
HartwickJohn M (1977) “Intergenerational Equity and the Investing of Rents from Exhaustible Resources,” American Economic Review Vol. 67issue no. 5 (Dec.) pp. 972–974.
HemmingRichard and MurrayPetrie2000 “A Framework for Assessing Fiscal Vulnerability,” IMF Working Paper 00/52 (Washington D.C.: International Monetary Fund).
HomeJoycelyn1991 “Indicators of Fiscal Sustainability,” IMF Working Paper 91/5 (Washington D.C.: International Monetary Fund).
LiuksilaClaireAlejandroGarcia and SheilaBasset1994 “Fiscal Policy Sustainability in Oil-Producing Countries,” IMF Working Paper 94/137 (Washington D.C.: International Monetary Fund).
Organization of the Petroleum Exporting Countries2000 “Annual Statistical Bulletinavailable from http://www.opec.org
SachsJ. D. and A.Warner (1997) “Fundamental Sources of Long-Run Growth,” American Economic Review papers and proceedings Vol. 87 No.2 (May) pp. 184–188.
SachsJ. D. and A.Warner (2001) “The Curse of Natural Resources,” European Economic ReviewVo. 45 pp. 827–838.
SolowRobert. M (1974” “Intergenerational Equity and Exhaustible Resources,” The Review of Economic Studies Vol. 41 pp. 29–45.
StiglitzJoseph (1974) “Growth with Exhaustible Natural Resources: Efficient and Optimal Growth Paths,” The Review of Economic Studies Vol. 41 pp. 123–137.
TanziVito and Hamid R.Davoodi (1998) “Roads to Nowhere: How Corruption in Public Investment Hurts Growth,” Economic Issues no. 12 (Washington D.C.: International Monetary Fund).
TersmanGunnar (1991) “Oil, National Wealth, and Current and Future Consumption Possibilities,” IMF Working Paper 91/60 (Washington D.C.: International Monetary Fund).
Prepared by Hamid R. Davoodi.
This estimate assumes 15 billion barrels will be discovered in the offshore Caspian field of Kashgan; production is not expected to begin before 2005. The estimated 30 billion reserves correspond to what is known as “proven” crude oil reserves. The latter is “an estimated quantity of all hydrocarbons statistically defined as crude oil or natural gas, which geological and engineering data demonstrate with reasonable certainty to be recoverable in future years from known reservoir under existing economic and operating conditions. Reservoirs are considered proven if economic producibility is supported by either actual production or conclusive formation testing.” (Organization of the Petroleum Exporting Countries, 2000)
See Amuzegar (1999).
The analytical section of this chapter relies exclusively on the oil sector in Kazakhstan, but the same issues are also relevant for other primary commodities produced in Kazakhstan such as copper and zinc.
The literature on fiscal vulnerability, though closely related to fiscal sustainability, is rather new. Hemming and Petrie (2000), for example, provide an overview of the relevant issues and provide a framework for assessing sources of fiscal vulnerability. However, the literature on fiscal sustainability is much older. The conventional view of fiscal sustainability tests for departure of government’ observed fiscal policy from that implied by the present value intertemporal budget constraint; for this theory and some applications, see Barro (1979, 1989) and Chalk and Hemming (2000), respectively.
Some studies have also indicated that conventional national income accounting in resource-rich countries is also misleading since they do not incorporate natural resource wealth, changes in natural resource wealth and environmental degradation associated with extraction of natural resource. These aspects also expected to affect fiscal policy stance. These arguments date back to at least Kuznets; See Auty (1998) for a discussion.
See the strict requirements for the definition of size of proven reserves in footnote 3.
See Auty (1998) and Sachs and Warner (2001).
See Sachs and Warner (1997) and Auty (1998), for example, shows that the growth ate of real per capita GDP in large resource-poor countries in 1970–93 was 3.7 percent, while that of resource-rich countries was 1.3 percent, and that of oil exporters only 0.8 percent.
See Sachs and Warner (1997). This empirical regularity has been uncovered in many empirical studies of growth; see Sachs and Warner (1997) and Valdivieso et al (2001). These studies have found that being landlocked implies annual real per capita GDP growth would be lower by some 0.6 percentage point, a significant magnitude, enough to pose a challenge for a landlocked country to converge to the economies of countries that are not landlocked. These challenges can be overcome in part, as shown, for example, by the construction of the Caspian Consortium Pipeline carrying Kazakhstan oil to Russia. However, access to the sea, as in the case of Iran to the Persian Gulf is far superior in terms of costs, sales and profitability.
See Davis, Ossowski, Daniel and Barnett (2001) who provide a review of some of these studies.
See SM//00/260 for a detailed discussion of the key aspects of fiscal vulnerability in Kazakhstan.
In this chapter, nonoil fiscal balance and non-resource fiscal balance are used interchangeably as oil is more important than other natural resources in Kazakhstan, but otherwise the issues are the same regardless of the type of the natural resource.
See Fischer (1993) and Sachs and Warner (1997).
Sachs and Warner (2001) review this literature. They conclude that the reason for the poor growth performance is that natural resource abundant countries have systematically failed to achieve a strong export-led growth or other kinds of growth.
The nonoil balance should also perhaps exclude oil-related spending at least to the extent that some oil revenues are earmarked for special projects.
See Tanzi and Davoodi (1998) who provide a useful discussion of wasteful spending and corruption in public investment projects.
These countries are: Norway (Tersman, 1991), Egypt, Indonesia, Mexico, Nigeria, Saudi Arabia and Venezuela (Liuksila, Garcia and Basset, 1994; Bascand and Razin, 1997; Chalk, 1998; SM/01/31), Kuwait (Chalk, 1998), Yemen (SM/01/56) and Azerbaijan (EBS/01/91). For theoretical treatments, see Chalk (1998), Alier and kaufman (1999), and Engle and Valdes (2000).
See the discussion of Venezuela and Nigeria in Davis, Ossowski, Daniel and Barnett (2001); see also Gelb and others (1988).
Application of the PIH to Kazakhstan incorporates country-specific risk premium.
The empirical analysis would actually allow for net debt.
Note that an increasing or decreasing paths are both optimal under the permanent income framework, but imply different intergenerational equity considerations and certainly different paths from the Friedman’ case.
This path has also been referred to as Solow’ definition of intergenerational equity; see Solow (1974).
The case of constant consumption path corresponds to an exact offset.
Technically, Stiglitz (1974) has shown that with no technical progress sustained growth can still occur so long as marginal product of physical capital is higher than that of natural resource capital; see also Solow (1974) for the same conclusion. Again growth history of resource rich countries has shown that this condition may not have been met in practice either because the initial capital stock is not large enough to support sustained growth or escape a poverty a trap or negative TFP growth has dominated any positive contribution from physical capital or both. The brief history of Kazakhstan since independence has shown a strong negative TFP growth in aggregate and across all industries; see Mard De Broeck and Kostial (1998).
If physical capital has a higher marginal productivity than natural resource capital, as assumed by Stiglitz (1974) and Solow (1974), also known as the Hotelling rule, then earlier or current generation should accelerate depletion of natural resources and build up the stock of physical and human capital stock in turn. The former seems to have been the case in the history of resource-rich countries, but not the latter.
This is equivalent to writing government intertemporal budget constraint, equation (1), in its present value form.
Use of nominal vs. real rate for discounting depends on the assumption used for the oil price path. Given that Fischerian equation holds in the long run and base line oil price is assumed to grow in line with inflation, the two approaches are equivalent so long as real interest rate is used for a constant oil price path and nominal interest rate for the nominal oil price path. This is derived formally later in the chapter.
This represents only value of oil reserves to the government, given the stated assumptions and existing government’ stake in the joint ventures, and not the gross value of oil reserves or oil sales, which is much higher. For example, using the same path of oil prices, the extraction rate and the interest rate, present value of gross oil reserves as of 2000 amounts to $141 billion. Both sets of estimates are obtained using equation (3).
Over the 2001–17 periods when oil production is continually rising, oil revenues in current U.S. dollar grows on average at 14.6 percent rate each year vs. annual nominal interest rate of 6 percent.
If year 2000 was the year used for defining value of residual oil in the ground, this value will be declining monotonically until oil reserves are depleted. In this regard, it should be noted that equation (4) should not be taken as implying that all tax revenues from oil sector are consumed by the government in each period; a fraction of the revenue is saved each period. This issue in analyzed further below.
Note that country’s initial indebtedness can be subtracted from Wt which would imply that saving from oil has to be higher in order to allow future debt repayment, i.e., future generations share the benefit of oil wealth as well as cost of the past debt build up.
See Blanchard and Fischer (1989, pp. 285) and Engle and Valdes (2000).
Note that same deflators are used for converting nominal consumption, nominal GDP and nominal total net wealth from nominal to real. As a result, nominal and real variables are identical up to the inflation discount factor.
Even if it is assumed that the Fischer equation does not hold in the long run, the path of consumption is not affected since the interaction terms involving inflation and population growth rate are too small, which can be ignored.
This estimate uses the exact figures and not approximate value of each since the orders of magnitude can make a significant difference in per capita terms.
The life cycle hypothesis is equivalent to the PIH by building, for example, bequest motives in the life cycle hypothesis; hence, the term life cycle/permanent income hypothesis is also used quite often in practice.
Estimates of nonoil deficit in countries with a long history of oil extraction show that these governments have dissaved significantly as is the case in Yemen; see SM/01/56.
Choice of the year 2000 seems to be plausible as was the case for the benchmark case. However, this does not imply that other ratios cannot be entertained.
This and subsequent analyses are based on the benchmark case of constant per capita wealth.