This workshop reviews techniques for projecting inflation, output, and the composition of expenditure in Hungary in 1990. Such forecasts are central to a stabilization program since inflation and growth are usually among the major goals of policy. Similarly, the composition of expenditure affects the growth potential and the external balance of the economy.

This workshop reviews techniques for projecting inflation, output, and the composition of expenditure in Hungary in 1990. Such forecasts are central to a stabilization program since inflation and growth are usually among the major goals of policy. Similarly, the composition of expenditure affects the growth potential and the external balance of the economy.


1. Introduction

In 1989, 63 percent (compared with 56 percent in 1985) of consumer goods had free prices, i.e., prices were subject to only the general rule that they not exceed the domestic price of comparable imports. A further 20 percent of consumer goods prices were subject to a requirement of advance reporting or prior consultation with the authorities. The remaining 17 percent of prices were subject to a maximum limit (compared with 44 percent in 1985).

Repeated efforts to overhaul the price system through administrative price realignments, the linkage of raw-materials and energy prices to world market prices, and the adoption of rules to allow for greater flexibility in the pricing policy of enterprises, avoided the accumulation of major distortions in relative prices that characterized most centrally planned economies. Nevertheless, the adjustment of prices to market conditions was hampered by three main factors. First, the pricing rules and the close monitoring of free price formation were used as an important instrument of anti-inflationary policy. Second, ad hoc taxes and subsidies were used extensively to affect individual enterprise results: current government subsidies and transfers accounted for about one third of GDP in 1989, a proportion that changed little in the 1980s. Finally, Hungarian industry was characterized by a very high degree of concentration and little competition. This stemmed from the early days of central planning when agreements among enterprises in a branch of industry would subdivide the product range into a number of distinct items or groups, each to be produced by one or a few firms. Not only were production units designed to be large and to take advantage of economies of scale, but they developed the in-house capacity to produce certain essential inputs, in light of uncertain external supplies, rather than relying on smaller-scale suppliers of producer goods as in a typical market economy. Competition was also impeded by the traditional reliance on foreign trade controls.

2. Forecasting inflation

a. General considerations

Forecasting inflation involves an assessment of many different factors, some of which are listed below:

  • (1) What are the cost pressures tending to increase prices? An example of a major cost-push (or supply side) effect on prices was the steep rise in world energy prices in 1973–74 and 1979–80. In estimating cost pressures, an assessment needs to be made regarding the proportion of the increase in costs that will be passed onto prices rather than being absorbed by enterprises in the form of lower profits. This will depend importantly on the stance of economic policies (point 2 below) and the degree of competition among enterprises. Enterprises response to cost increases will also depend on the willingness of the government to extend tax concessions and/or subsidies to those sectors experiencing cost increases.

  • (2) What is the stance of economic policies and how has it changed from the previous year? For example, an easing of monetary policy is likely to put upward pressures on prices, especially if real output remains unchanged, as residents attempt to spend increased money balances on goods and services (demand-pull inflation).

  • (3) Does a change in one, or a few, prices constitute a change in the overall price level or does it reflect a change in relative prices that should be offset by changes in the opposite direction in other prices? Relative price changes act as a signal to redirect scarce resources to activities where they can be most productively used. By contrast, increases in the overall price level normally indicate the existence of excess demand pressures in the economy. In practice, the distinction is blurred because of a downward rigidity of prices in the short-run. Relative price changes are, thus, typically also reflected in upward adjustments in the overall price level, at least temporarily.

  • (4) How do people’s expectations about future movements in the price level affect today’s increase in the price level? While modeling expectations has generated considerable debate and empirical work, the underlying notion is relatively simple: people’s behavior today will be affected by how they perceive the future.

  • (5) The institutional arrangements in a country can also affect price developments. These include price control arrangements, competition policy, and the foreign trade system. The more liberal the latter, the greater the proportion of any excess demand that is likely to be dissipated through the balance of payments rather than in domestic price increases. Equally, price behavior may be strongly influenced by the exchange rate regime: the less flexible is the exchange rate system, the more the demand pressures will be reflected in a drawdown of foreign exchange reserves rather than in an increase in prices.

b. A practical approach

The approach described below is designed to provide the outlines of a possible methodology for estimating inflation based on a cost-approach. The estimates of cost increases can only be approximations since over time they could have wider ripple effects as they work themselves throughout the economy. Similarly, price changes observed in any given period may reflect not only changes in cost pressures in that period but also the continuing effects of cost changes from previous periods. For example, the impact of an increase in import prices may at first only be felt on the prices of goods imported directly and domestic goods using imported inputs. However, over time, other domestically produced goods are also likely to face higher input costs as these increases spread through the economy. This effect will be more pronounced if wages are raised to compensate for price increases.

Once some benchmark estimate is derived, it must be adjusted to take account of the stance of economic policies. For example, a tightening of policies is likely to be reflected in less of a pass-through of costs into prices as demand conditions weaken and, possibly, wages and prices show more downward flexibility. By contrast, a loosening of policies is likely to magnify the ripple effects of the cost push pressures. Such adjustments are by their nature judgmental. Nevertheless, a more accurate picture of the stance of economic policies is likely to emerge on completion of forecasts of the fiscal and monetary accounts (Chapters VI and VII). This may require a revision to the initial estimates for inflation calculated in this workshop.

The two indices that this exercise focuses on are the consumer price index (CPI) and the deflator for gross domestic product (PGDP). The CPI is an index of prices of a typical basket of imported and domestically produced goods that are consumed by residents. The PGDP, by contrast, is a price index for all goods produced, both for the home market and for export. Differences in movements in the CPI and PGDP would thus be expected to reflect, among other things, differences in the movements of export and import prices. For example, if import prices were to increase by much more than export prices, other things being equal, inflation, as measured by the CPI, would be expected to be higher than as shown by the PGDP.

It should, however, be noted that in reality the separation of the effect of import prices on the CPI and of export prices on the PGDP is not clear cut. The ripple effect of an increase in import prices on domestically produced goods, and thus on the PGDP, has already been discussed. As regards an increase in export prices, some exportable products may be consumed domestically. To the extent that enterprises ask the same price of these products, whether they are exported or sold at home, the effect of changes in export prices could have a direct impact on the CPI.

Based on the underlying cost pressures in the economy, Table 1 projects inflation for 1989, for both the CPI and PGDP. Actual data available on an ex-post basis is used. Table 2 is a working sheet summarizing the calculations underlying the various cost components of Table 1.

The starting point is to assume that, in the absence of changes in cost pressures, inflation will remain at about its 1988 rate (line 1, Table 1). In 1989 there were at least three factors that should have worked in the direction of moderating inflation:

  • (1) The introduction of a value-added tax in 1988 is estimated to have led to a once-and-for-all jump in the price level in that year equivalent to about 7.0 percentage points (line 2, Table 1).

  • (2) Despite the fact that wage increases in the socialized sector exceeded those planned by wide margins—19 percent compared to a plan of 6 percent—wages in 1989 still rose by about 5 percentage points less than in 1988. The large wage awards of 1988 reflected the grossing up of incomes prior to the introduction of the personal income tax so as to preserve the net income of employees. With the wage bill equivalent to about one-third of the value of output, this moderation in wage growth could be expected to lower inflation by about 2 percentage points (line 3 of Table 1 and calculation 1 of Table 2).

  • (3) Government subsidies, in relation to GDP, fell by 0.7 percentage points less in 1989 than in 1988, which should have moderated pressures to pass through the reduction in subsidies to prices by a similar amount (line 4 and calculation 2).

Table 1.

Hungary: Estimates of Inflation for 1989 Based on Cost Factors

(percentage change)

article image
Source: IMF Institute data base; see also Table 2 for calculations underlying the above figures.
Table 2.

working Table: calculations underlying Table 1

article image
Source: Staff estimates.

The main cost pressures in 1989 appeared to stem from foreign trade prices. Import prices, in terms of forint, increased by some 15 percentage points more than in 1988. This reflected two factors: (1) the forint/US dollar exchange rate depreciating at a faster rate in 1989 than in 1988; and (2) import prices in US dollar terms rising in 1989, while they fell in 1988. With imports accounting for just over one-third of GDP, this could be expected to raise inflation, as measured by the CPI, by around 5.5 percentage points. Inflation, as measured by the PGDP, would be affected by the rise in import prices only to the extent of the ripple effects on the cost of producing domestic output. An arbitrary estimate of 3.0 percentage points was attributed to this factor (line 5, calculation 3).

As regards export prices in forint terms, they also increased faster in 1989 than in 1988 (by 13.5 percentage points). With exports accounting for a similar share of GDP to imports, this could be expected to raise inflation, as measured by the PGDP, by about 5.5 percentage points. An arbitrary 1 percentage point was added to both the increase in the CPI and the PGDP, to take account of the effect of the increase in the price of exportables consumed domestically.

Increases in interest rates can also generate significant cost pressures when enterprises’ outstanding bank loans exceed their deposits. A calculated cost pressure of just over two percentage points is based on the fact that the net outstanding credit position of enterprises with the banking system accounted for just over 30 percent of GDP in 1988 (line 8, calculation 4).

Summing up the various cost pressure components, one derives benchmark estimates for inflation in 1989 of 14.6 percent and 16.9 percent, as measured by the CPI and PGDP, respectively. Such estimates are inevitably crude and should be treated as first approximations to be refined in light of additional information. They do, however, point to a slowing of inflation based on cost developments. In fact, inflation in 1989 edged upward, to 17 percent and 18.7 percent, as measured by the CPI and PGDP, respectively. While this discrepancy may, in part, reflect the continuing influence of cost changes in previous periods, it is also likely to be attributable to the lax financial policies discussed in Chapter I.


1. The structure of output in Hungary

Industry is the single largest sector in Hungary, accounting for about one third of aggregate value added throughout the 1980s (Table 9). There was some small decline in this share in the second half of the 1980s as industrial output fell on average by 0.2 percent, compared with a positive growth of GDP of 1.6 percent. Nevertheless, industrial output moved generally in step with total GDP (stagnant in 1985–86 and 1988, growing in 1987), until 1989. In that year, the volume of output of nonmaterial services increased by one quarter, while both industrial and agricultural output fell. During the 1980s nonmaterial services were the fastest growing activity and by 1984, at 18 percent of GDP, were firmly established as the second largest sector. Together, all branches of the services sector, defined to include transport and communications, trade, and other material and nonmaterial branches, accounted for over one third of value added, or about the same size as the industrial sector.

Agriculture, the third largest sector, accounted for about 15 percent of total output on average during the 1980s. With favorable weather, agricultural output expanded contracyclically in 1986 and 1988 and decreased (also contracyclically) in 1987 and 1989. Growth in agriculture averaged two tenths of one percent during 1985–89.

2. Determinants of output: demand and supply factors

In a market economy, decisions by enterprises to produce certain goods or services, and by households to consume them, are typically voluntary. The level of public services provided by government will be determined administratively, in response to political processes. For the most part the volume of aggregate output will reflect decisions of individual enterprises. As a general rule, enterprises will increase output when profitability is high or increasing, and conversely.

Macroeconomic analysis does not attempt to track the behavior of individual firms but instead focuses on the overall level of output. The quantity of goods and services that can be produced will vary with the inputs that are used (land, labor, and capital) and the level of available technology, work attitudes, and market structure. 1/ In general, if the quantity of inputs increase or are used more effectively, the potential output of the economy expands.

In a market economy, whether firms produce at the maximum potential rate or at less than that rate depends on their actual and prospective sales—i.e., on the demand for their product. If sales fall below potential output, an individual enterprise will typically adjust downward the volume that it has been producing, possibly laying off some workers. At the macroeconomic level, when demand is weak unemployment in the economy will be higher than normal, and firms will operate at some fraction of their potential level of production. When employees, machines, and buildings are idle, resources are wasted. Potential supply is a kind of upper limit to GDP, but when aggregate demand is less then supply, actual output will initially tend to be limited by demand. 1/

An aggregate production function indicates the relationship between output and the various inputs that constitute factors of production. Formally, the production function may be written as follows:


Equation (1) states that real output (Q) can change because of increases in the availability of capital (K) and labor (N), and through technological and other improvements (A) that enable existing factor supplies to be used more effectively.

Over time, increases in output must reflect greater availability of factors of production or more efficient factor use. In the short run, growth in real GDP is more likely to reflect changes in the extent to which labor and capital are fully employed. The level of output that would be realized if the labor force were fully employed, and labor and capital were used at normal intensity, is referred to as potential or full-employment output. Potential output defined in this way is less than the physical maximum producible with a given capital stock and labor force. The unemployment rate will not be zero at full employment since, in a dynamic economy, some workers are usually between jobs. Similarly, equipment is normally taken out of operation for certain periods of time to allow for maintenance.

While aggregate output may be forced to a level above normal full employment, this is usually associated with increasing costs of production. For example, if laborers work more than the customary number of hours per week, say 40, they would in most cases be paid “overtime,” an hourly wage rate 50–100 percent higher than normal for the excess hours. Depreciation of the capital stock would also increase at a higher intensity of production. A firm would normally try to bring its work force into line with its medium-term rate of production so as to avoid these increased costs. However, if demand is very strong in the short run, it may be profitable for a firm to temporarily increase output, the higher costs being passed onto prices.

Because the potential level of output refers to use of factors at normal intensity, total output can be higher than potential output, at least for short periods. In general, if aggregate demand is increased, output will tend to be higher, and conversely. Production costs may also be higher, however, especially if the economy is approaching normal full-capacity. Thus, as demand is increased, both some additional output and higher inflation will generally result. The favorable output gains will tend to be large and the inflation effects small if the economy is well below potential output, but price effects will be large and quantity effects small or negligible if output is already at or above the normal full employment level. When aggregate demand is reduced, initially the effects tend to fall primarily on the quantity of output, although in time the inflation rate will also tend to decrease.

For an industrial country with extensive data, the secular growth of potential output may be analyzed with the use of production functions, while short-;run determination of output should take into account both supply and demand factors. Estimates of production functions require adequate data on labor and capital stocks, while formal consideration of the interaction of supply and demand factors depends on a properly articulated econometric model.

Many countries have neither the data nor the models to support such approaches. Consequently, output forecasts in financial programs tend to rely on essentially ad hoc procedures. A starting point could be to project output for the forecast year on the basis of trends in output growth in the recent period. This could be verified by an analysis of trends in the sectoral composition of GDP as, for example, presented in Table 9, but these should be modified to reflect factors specific to each sector and known interrelationships among different activities. For instance, forecasts for agriculture should reflect expectations about weather conditions, producer prices, and other special factors. Forecasts for manufacturing activity could incorporate the reliance of manufacturing on the agricultural sector for inputs. Factors affecting output during periods of structural change also need to be considered. Finally, given the role of demand factors in determining output, projections of output need to take account of the stance of aggregate demand. In particular, activity in the secondary and tertiary sectors is likely to be sensitive to demand arising from the domestic or external sectors.

Central-Planning and Market Approaches to the Analysis of Aggregate Output

Under central planning, the detailed numerical tables that comprise the plan are simultaneously a means of directing producers, forecasting output, and monitoring the results of government directives. Unemployment and underutilization of productive resources are, in principle, eliminated by keeping demand high relative to supply. The resulting upward pressure on prices is controlled, in turn, by the cost of increased waiting time, or queuing, and by increases in involuntary savings such as in unspendable checking account balances. From the point of view of forecasting it is therefore sufficient to focus on the supply of output in a planned economy. By contrast, aggregate supply and aggregate demand, or spending, need to be considered together in determining output in a market economy.

The elements of an input-output matrix in central-planning describe quantities of output of certain enterprises or branches that are required as inputs by other producers. Additional rows indicate inputs of primary factors (labor, capital, and so forth), and total output; columns show intermediate use, investment, export, and consumption. The division of output between consumption and investment is made administratively and reflects political considerations.

The production function is a major analytical tool in market economies in analyzing the growth of output. Although the voluntary behavior of profit maximizing enterprises is commonly assumed to underlie the activity represented by a production function, this is not essential; a production function could, in principle, be used to forecast output in a centrally-planned economy, just as input output analysis is sometimes used in market economies. The production function lacks, however, the details of the operations of individual enterprises or branch industries that is essential to the working of the planning mechanism.

The essential difference between these two approaches is that the production function sacrifices detail but can incorporate the possibility of varying the relative amounts of capital and labor used in the production process. From the point of view of economic theory there are two significant features of this variability that are not allowed for in input-output analysis. One is that the relative economic rewards of labor and capital may change—wages may rise or fall compared to the rate of return on capital. The second is that these relative income changes could induce enterprises to change their mix of factor inputs—to “economize” on labor and use more capital, or vice versa—so that there is a tendency for all available inputs to be fully employed.

3. Applications to Hungary

a. Historical trends in output growth

Forecasts of output and prices for Hungary are complicated by the unevenness of developments during the 1980s and the increasing pace of reform toward the end of the decade, GDP growth was close to zero in 1985–86, and again in 1988, but attained favorable rates in 1987 and 1989. (See Tables 9 and 10). The authorities were concerned to limit aggregate demand in order to reduce the deficit on current external transactions in convertible currencies because Hungary had, by the end of the 1970s, already accumulated a large stock of external debt. At the same time, there was a desire to stimulate growth and to increase living standards. The conflict between these two objectives resulted in periods of more expansionary monetary and fiscal policies followed by years of contractionary policies (sometimes called “stop-go” policies). 1/

Fluctuations in domestic demand, however, only partly explain the stagnation expansion pattern. Aggregate supply was influenced by changes in the share of output allocated to investment, by variation in weather conditions, and by fluctuations in foreign demand for Hungary’s exports and changes in their competitiveness. Major reform measures taken in this period may also have affected supply.

While output in the 1980s fluctuated considerably, the trend growth rate was considerably lower than in the previous two decades (Table 3). Notably, aggregate real output in Hungary grew vigorously (by around 6.5 percent) in the decade preceding the first oil-price increase, and at a reduced, but still decent rate (4.9 percent) during the remainder of the 1970s. During the 1980s, however, output growth slowed markedly, averaging 1.6 percent per year in the last five years.

Several hypotheses can be offered in explanation of this medium-term slowdown.

  • (1) The level of unemployment remained close to zero during this period, there was no major influx of laborers from neighboring countries, and the population actually declined slightly. Labor-force growth largely reflected increases in participation rates which limited the potential growth of total output.

  • (2) The tendency to reduce demand for stabilization purposes by cutting back investment relative to GDP during the 1980s is likely to have had negative effects on growth.

  • (3) The higher costs of energy tended to slow the Hungarian economy indirectly as weak demand in the world economy inhibited exports.

  • (4) Central planning was able to achieve rapid growth rates in initial years because of the substantial allocation of resources to production of capital goods. Gains from this source, however, eventually decreased, and failed to outweigh the costs of inefficient use of capital and the increasing administrative burden as economies became intrinsically more complex. Moreover, the reform measures introduced were often partial, inconsistent, and subsequently reversed.

  • (5) Finally, growth rates may eventually decline under central planning because such a system is less attuned to rapid adaptation and full exploitation of the opportunities presented by scientific discoveries and changes in consumers’ tastes.

Table 3.

Hungary: Growth of Output and Prices by Five-Year Periods

(Annual rates of growth in percent)

article image
Sources: Row 1 is from Alton’s regression averages, reported in Ed A. Hewitt, “The Gross National Product of Hungary,” World Bank Staff Working Papers, Number 775 (Washington, D.O.: World Bank, 1985), pp, 16, 19, and from Hungarian sources. Other rows are from IMF, International Financial Statistics, except the column for 1985/89, which is from Tables 1 and 2.

Average of rates in 1971-74.

Measurement of Historical Rates of Growth in Centrally-Planned Economies

Market-economy observers of developments in centrally-planned economies have often been skeptical of historical output growth rates officially reported by the latter. Omitting certain services in the measure of output conventionally used, net material product was thought to result in an upward bias in growth rates dominated by manufacturing sectors. Many of these economies now also calculate output in terms of the concepts used in the United Nations System of National Accounts.

Two additional sources of over counting were suspected. Because of the negative political connotations of price increases and the pressures to report growth, enterprises were said to have a strong incentive to mask any increase in their costs of production in terms of output increases. For example, if production costs were to rise and be passed onto prices, enterprises could disguise this by attributing the increase in the value of production to real output increases—possibly by redesigning their products slightly so they could be labeled as new and improved quality products—rather than to price increases.

A second potential component of overcoming was the inventories of partly produced and finished goods reported by firms. By accumulating nonexistent inventories, an enterprise could appear to have complied with a physical output target (perhaps a highly optimistic one) without producing the corresponding quantity of goods. Levels of inventories were typically much higher under central planning than competitive systems, although uncertain supplies of materials and parts in planned economies could account for some of the need to hold larger inventories.

b. Output during periods of transition

Periods of rapid structural change have often been associated with lower (higher)—than—expected output growth (losses). It is useful to review at least some of the systemic forces that could be at play in an economy moving from central planning to market orientation. While by their nature the effects of system factors are difficult to quantify, they are often the predominant influence over output developments in periods of transition. Some of the main such factors are listed below.

  • (1) The operating environment may not be conducive to the complete abandonment of old for new ways, which may inhibit market generated supply responses, including investment decisions. For example, the response to market signals will be weakened if enterprises believe that they will be bailed out of their financial difficulties, either with bank credits and/or budgetary subsidies, and when the initiation of bankruptcy proceedings is not perceived as a real threat.

  • (2) Delays in the clarification of property rights in a system undergoing decentralization may result in upward pressures on wages and a weakened output response to any given change. Take for instance the typical workers of an enterprise who have access to the rents on capital through their influence over the allocation of profits, but do not own the capital. The only channel through which they can share in the benefits of good decision made today that affects future profitability is to remain employed with the enterprise (which death, retirement or dismissal can prevent). Moreover, there is no way in which they can transfer their entitlement to the rents of capital to their heirs.

  • (3) The change in relative prices induced by liberalization, together with the effects of foreign competition, may render some of the output unmarketable at a cost-covering price. Given that it is easier to cut the output of those products that are not in demand than to increase production of those that are, a further transitory cut in output is implied. The need to shift resources from contracting industries to expanding industries as a result of a change in relative prices is similar to the reallocation that would be expected in any economy after, for example, a major devaluation of the exchange rate. In general, the fall in output should be less severe and protracted the smaller are the initial distortions, the more flexibility there is in the system, and the greater is factor mobility.

  • (4) Movement toward a market mechanism is typically accompanied by strong inflationary pressures as prices, exchange rates and interest rates are liberalized and subsidies are reduced. In order to minimize the disruptions that occur when inflationary pressures begin to create their own inertia, tight financial policies are needed. Their contractionary effects tend to initially fall more on output than on prices, reflecting a downward inflexibility of prices. This is particularly so when cost-plus pricing considerations predominate profit incentives.

  • (5) Price liberalization may also release monopolistic forces that had previously been repressed by the mechanisms of central planning and are not as yet effectively checked by competition from alternative domestic or foreign sources of supply. In this case, enterprises would respond by carrying-out a one time increase in prices and reduction in output.

  • (6) A contraction in CMEA trade would be expected to have a significant effect on output, given the noted rigidities in redirecting/transforming the previously exported output to other sectors, e.g., to the convertible currency area. Hungarian officials recently estimated that, other things being equal, a 10 percent decline of exports to the CMEA would result in a decline of GDP of 1 percent.

  • (7) Disruptions to, and bottlenecks in, the distribution networks may initially be in evidence as the existing structures break down and the new ones are not there to replace them.

  • (8) Finally, officially reported statistics may not accurately report actual output developments, at least in the initial stages of reform, because they should largely be based on returns from the existing large socialized enterprises. They may not, therefore, adequately reflect the growing activity of smaller, particularly private, enterprises.

The consequences of the above factors may be exacerbated by the uncertainty of the economic situation and political environment, which may lead many individuals and enterprises to adopt a “wait and see” position—making them reluctant to engage in new activity requiring a longer-term investment in physical or human capital.

Composition of Expenditure

Output can be forecast from both the demand and supply side. By definition, actual expenditure on domestic output will always be equal to the supply of domestic output on an ex-post basis. In a market economy, if the demand for output initially exceeds the supply, some form of adjustment will take place, typically through some combination of higher prices and higher net imports. 1/ A comparison of projections from both the demand and supply sides thus provides a check whether the overall stance of policies, as reflected in demand pressures for domestic output, is consistent with initial supply-side projections for output.

Projections of the composition of expenditure indicate the expected division of domestic demand between consumption and investment as well as the generation of domestic savings. Such forecasts also show the implications of domestic absorption for the external balance.

Forecasts for government consumption and investment should be based on budgetary projections, as discussed in Chapter II. Similarly, the forecasts for exports and imports should be based on balance of payments estimates (Chapter V). Consequently, the sections below concentrate on a projection of household consumption and household and enterprise investment. An overall projection of output from the demand side may need to wait until projections of the monetary, fiscal and external accounts are made.

1. Determinants of household consumption

Household disposable income (YD) can be viewed from the standpoint of either its sources or its uses:



  • W = wages and salaries from the socialized sector and agricultural cooperatives

  • E = net earning from private activities

  • B = social benefits

  • T = taxes less transfers, i.e., direct personal taxes, such as income tax, net of personal transfers, such as pensions and unemployment compensation.

  • CH = household consumption

  • SH = household savings

Equation (2) shows that household consumption and household savings are simultaneously determined. For a given level of disposable income, once consumption is known savings is determined as a residual. It also suggests that disposable income is an important determinant of household consumption.

Table 4 summarizes the trends in the relationship between household consumption, on the one hand, and GDP and household disposable income on the other. Between 1980 and 1989 household consumption as a share of GDP was remarkably stable at around 61 percent (column 1). Changes in the share of disposable income consumed (column 2) were broadly offset by changes in the opposite direction in the share of household disposable income in GDP (column 3). For example, in 1985–86 when financial policies were eased and wage gains exceeded increases in profitability, the share of household disposable income in GDP rose substantially at a time when the proportion of income consumed fell. By contrast, in 1989, when profit increases exceeded wage gains, household consumption in relation to disposable income soared dramatically following the liberalization of regulations on personal imports and travel abroad, which released pent-up purchasing power.

Table 4.

Hungary: Household Consumption in Relation to GDP and Disposable Income 1/

article image
Source: IMF Institute data base.

In nominal terms.

As an alternative to extrapolating on the basis of simple ratios, regression techniques may be used to identify behavioral relationships. Specifically, a behavioral relationship between household consumption and household disposable income can be postulated as:




This specification of the consumption function implies that the marginal propensity to consume (MPC, or ΔCP/ΔYD) is constant and is equal to b, while the average propensity to consume (CP/YD) falls as YD increases, due to the declining relative importance of a positive constant term. These properties suggest that developments in consumption and savings will help stabilize the economy because the savings ratio tends to rise in booms and decline in slumps. Further, as income grows over time they suggest a long-term rise in the savings ratio.

Other approaches to estimating consumption functions recognize that the consumer is unlikely to be influenced only by income in a particular period. Rather the consumer will consider some longer run or normal income stream in determining consumption behavior. For example, during cyclical booms transitory income is positive and the ratio of consumption to actual (as opposed to long-run) income would be expected to decline. Similarly, cyclical slumps would be associated with negative transitory income and the ratio of consumption to measured income would be expected to rise. This also suggests that the MPC will be different in the short-run and the long-run. The permanent income hypothesis and the life-cycle theory represent two essentially complementary attempts to make longer-term income concepts operational.

The permanent income and life-cycle hypotheses of consumption behavior can both be empirically approximated as follows:


The coefficient B1 could be interpreted as the short-run MPC to consume and B11-B2 the long-run MPC.

Table 5 summarizes the results of estimating a consumption function in Hungary using annual data for the period 1975–88. Consumption and income data are defined in real terms. The year 1989 was excluded because, as is evident from Table 4 and was discussed above, the share of disposable income that was consumed in that year was extraordinarily high by historical standards.

Table 5.

Hungary: Consumption Functions

article image
Notes: t-value in parenthesisCH = Household consumption in real terms, the dependent variableYD = Household disposable income in real termsCH(-1) = Household consumption lagged one period.

Equation 5 assumes that the consumption function takes the form of equation (3) of the text. An MPC of 0.83 is estimated. A permanent income approach to Hungarian consumption behavior indicates a short-run MPC of 0.47 and a long-run MPC of 0.76 (Equation 6).

2. Forecasting investment

Gross investment consists of gross fixed capital formation, which includes machinery, equipment, and structures, and changes in inventory by enterprises. In recent years inventory increases have been equivalent to 0–1.5 percent of GDP annually.

Table 6 summarizes the composition of gross fixed investment by sector. Between 1985 and 1989 total fixed investment accounted for 20–25 percent of GDP. Household investment, 90 percent of which was in housing, remained relatively small at around 4.5 percent of GDP. The remainder of investment was done by the public sector. The distinction between the government sector and public enterprises was blurred by the fact that the government had the authority to force enterprises to carry out certain investment projects even if they were financed in part or entirely by the enterprises themselves. The classification of fixed capital formation in the accompanying tables is by the entity which paid for the investment, rather than by decision-making authority. The difficulties in attributing separate behavioral relationships to enterprise and government investment decision-making suggests that any discussion of investment should be in terms of the overall public sector. In forecasting a breakdown of investment by entity, however, the government investment figures should coincide with those in the general government accounts.

Table 6.

Hungary: Gross Fixed Investment

(In percent of GDP)

article image
Source: IMF Institute data base.

Among the major factors determining the actual level of investment in Hungary were the official plan; the amount of liquid financial assets held by enterprises; the availability of capital transfers, subsidies, and bank credit; fears of confiscation of assets; and expectations of tax changes. Evaluations based on profitability played a relatively minor role.

While the official plan strongly influenced the rate of real fixed capital formation in the 1970s, the correspondence between plan target and performance became looser during the 1980s (Table 7). This reflected the increased autonomy accorded to enterprises. In addition, the share of enterprise investments financed from their retained earning increased. These developments increased the relative importance of financial policies and discipline in determining the level and composition of investment.

Table 7.

Hungary: Annual Percentage Changes in Real Gross Fixed Capital Formation

article image
Source: IMF institute data base.

Simple average.

With the exception of 1982, real gross fixed capital formation declined in each year during 1980 to 1985, in part the result of stabilization policies that aimed to reduce domestic demand but protect household consumption levels.

The 1986–90 five-year plan envisaged a resumption of overall investment growth together with a shift in emphasis toward small- and medium-sized projects to promote structural change and introduce new technologies. Planned increases in real investment were exceeded by modest amounts in 1986 and by over three times the upper target in 1987. During the latter year investment in infrastructure by municipalities recovered, and state enterprises spent heavily in response to buoyant profits, swelled by a boom in consumer spending and easy access to credit and subsidies.

Following the surge of the previous year, a 1 percent reduction in real gross investment was planned for 1988, which proved to be an actual drop of over 8 percent. Large state and municipal projects were curtailed and enterprises’ access to state equity loans and credit was tightened. In addition, enterprises delayed investment decisions because of an expected increase, from 20–40 percent, in the rebate from the value added tax on investment expenditures. In the event, public enterprise investment fell by 18 percent in real terms. In 1989 the actual increase in real gross fixed investment of 5.2 percent approximated the plan target of 6.5 percent as credit conditions eased significantly for enterprises as the government adopted an expansionary stance. For 1990, a 3.9 percent increase in real fixed capital formation was planned.

It is interesting to note how the share of output devoted to investment changed relative to the growth of output over the last 10–15 years. An increase in this ratio, commonly referred to as an ICOR (incremental capital output ratio) would indicate that over time the attainment of a given output growth would require increasing amounts of investment. One implication that could be drawn from such a trend is that the efficiency of investment had fallen.

Table 8 calculates ICORs for the 1977–89 period. To minimize the impact of year-to-year fluctuations in this ratio, averages of the last three years were used. ICOR rose from about 6.0 in 1977–79 to a peak of about 21 in 1985–86 before falling back to a range of 10–15 in 1987–89. Developing country ICORs in the mid-1980s ranged from 4 for Asia to 20 for the Middle East and 25 for Africa; hence Hungary was in the middle of the spectrum.

Table 8.

Hungary: Calculation of ICOR

article image
Source: Staff estimates.

What factor could be responsible for the deterioration in the productivity of fixed investment in Hungary between the 1970s and 1980s? There was no similar doubling of ICOR in Europe as a whole over this time span. One possibility is that the data do not fully reflect investment and output growth in the expanding private enterprise sector, where fixed capital most likely was more efficient. Another explanation could be that the system of taxing successful enterprises heavily and periodically confiscating their retained profits so as to subsidize loss-making enterprises had increasingly adverse effects on efficiency.

Such practices had at least three different adverse impacts on the level and productivity of fixed investment. First, it was a deterrent to profit-making; hence, generating an operating surplus for financing investment was discouraged. Second, for firms that could generate an operating surplus, the threat of confiscation of financial assets created a strong incentive to transform these financial resources into retained fixed capital as quickly as possible, even if the marginal productivity of additional fixed investment was declining rapidly. Third, making transfers to inefficient enterprises channelled financial resources to either supporting current operations or financing low productivity investments.

Exercises and Issues for Discussion

1. Exercises

For the reference scenario:

a. Project inflation for 1990, as measured by both the CPI and PGDP, taking account of:

  • cost pressures

  • the stance of economic policies

  • institutional factors

  • inflationary expectations

In making these estimates, consideration should be given to the movement in foreign trade prices, provided below:

Total Trade (In Convertible and Nonconvertible Currencies)

article image

b. Project output from the supply side, using the format of Tables 9 and 10 and taking account of:

  • recent trends in output growth and factors particular to the current period

  • the stance of economic policies

  • the role of structural factors

c. Project household consumption and total investment in 1990. (An overall projection of output from the demand side, i.e., completion of Table 11 and 12, and verification that expenditure on domestic output is equal to the supply of domestic output may need to wait until projections 2 of the monetary, fiscal, and external accounts are made.)

2. Issues for discussion

a. Consider how the projections for output and prices might be affected by the following factors:

  • poor weather conditions

  • a depreciation of the nominal exchange rate

  • an increase in the price of petroleum on world markets

  • an easing of monetary and fiscal policies

b. Discuss possible differences in the behavior of inflation and output in response to the introduction of stabilization programs between economies in transition and market economies. Suggest ways to minimize adverse impacts of stabilization policies on prices and output.

c. Review the problems of interpreting official measures of prices and output during periods of transition when black markets exist.

Table 9.

Hungary: Gross Domestic Product by Sector

(In billions of forint, at approximately 1986 prices) 1/

article image
Source: IMF Institute data base.

The value of the deflator for GDP at market prices in 1986 is 101.5 (see Memorandum item).

Turnover taxes, customs taxes, and other business taxes, less government transfers to enterprises. Also includes valuation changes and a statistical discrepancy.

Table 10.

Hungary: Gross Domestic Product by Sector

(In billions of forint, at current prices)

article image
Source: IMF Institute data base.

Turnover taxes, customs taxes, and other business taxes, less government transfers to enterprises. Also includes valuation changes and a statistical discrepancy.

Table 11.

Hungary: Components of Aggregate Demand

(In billions of forint at 1986 prices

article image
Source: IMF Institute data base.

Standard definition: consumption of residents at home and abroad.

Standard definition: including foreign tourism.

Table 12.

Hungary: Components of Aggregate Demand

(In billions of forint at current prices

article image
Source: IMF Institute data base.

Standard definition: consumption of residents at home and abroad.

Standard definition: including foreign tourism.


Inputs are also referred to as “factors of production” or as “factor services”; workers comprise a factor of production, and the work that is performed is a factor service.


It should be noted that low domestic demand may free goods for export, if exports are determined more by domestic supply than by foreign demand. In this case, the extra foreign spending would substitute for domestic demand and the level of output could be sustained. Also, the presence of unemployed resources, especially in large quantities or if expected to persist, could slow the tendency of prices to creep upward, or even produce a decline in prices, which would generally stimulate demand, output, and employment.


For a more detailed account of developments in the 1980s, see Chapter I.


By contrast, in a planned economy such an imbalance may manifest itself in shortages, queues, holding of unwanted liquid balances, and proliferation of black markets.