Journal Issue

Republic of Poland: Selected Issuse

International Monetary Fund
Published Date:
May 2000
  • ShareShare
Show Summary Details

VI. Prospects for Private Saving in Poland1

A. Introduction

1. During the last decade, Poland has achieved a high rate of economic growth, with investment rising rapidly over time. The saving rate, however, has been flat after a strong rebound from the sharp decline at the beginning of the transition, thereby leading to a growing current account deficit (Figure 1). Looking ahead, investment ratios will likely have to rise further to maintain the high growth rate. This points to the need to raise the saving rate in Poland over the medium term. This chapter discusses the prospects that this increase will be forthcoming. It describes the recent trend in the saving-investment balance, draws on the literature on the major determinants of private saving, and presents the results of a simple econometric analysis based on a panel dataset of selected transition economies.

Figure 1.Poland: Saving-Investment Balances, 1990–99

(in percent of GDP)

Citation: 2000, 60; 10.5089/9781451831887.002.A006

B. Trends in Saving and Investment in Poland

2. The gross domestic saving rate was around 21 percent of GDP during the 1980s in Poland (Figure 1). A marked drop in the saving rate was experienced in the early transition period, followed by a strong, but short-lived, rebound. The saving rate has returned to about 22 percent in recent years. Private saving accounted for the bulk of the gross domestic saving. Available data for 1991–98 indicate that household saving rate has been on a declining trend over the last few years, whereas business saving rate increased dramatically over 1993-95 before stabilizing (Figure 2). Both household and business saving rates have since been flat at about 9½ and 12½ percent of GDP, respectively. Meanwhile, total investment increased sharply and consistently after the collapse during the early period of the transition, with foreign saving filling in the gap between domestic saving and investment. Saving and investment rates exhibited similar patterns in other transition economies over time, falling sharply at the beginning of the transition before rebounding (see Denizer and Wolf, 1998 for a broader coverage of the transition economies). However, while saving rates were flat in Poland in recent years, five other transition economies showed a slightly upward trend (Figures 3 and 4).

Figure 2.Poland: Composition of Private Savings, 1991–98

(in percent of GDP)

Citation: 2000, 60; 10.5089/9781451831887.002.A006

Sources: WEO and staff estimates.

Figure 3.Poland: Gross Domestic Saving Rates in Selected Transition Economies,1990–99

(in percent of GDP)

Citation: 2000, 60; 10.5089/9781451831887.002.A006

Figure 4.Poland: Gross Capital Formation in Selected Transition Economies,1990–99

(in percent of GDP)

Citation: 2000, 60; 10.5089/9781451831887.002.A006

Sources: WEO and staff estimates.

3. Saving rates have behaved very differently over the time and across countries. A recent World Bank research project on saving—based on a World Saving Database covering 112 developing and 22 industrial countries from 1960 to 1995—found that in OECD countries, saving declined from more than 25 percent of gross disposable income in 1965 to less than 20 percent in 1995. On the contrary, between 1984 and 1993 in East Asia, saving rates nearly doubled, reaching on average 30 percent of disposable income in the 1990s. A key question for Poland is whether the saving rate will increase as the economy grows, resembling the East Asian experience or decline, following the experience in the industrial countries. The following sections attempt to answer this question, drawing on the theoretical literature and an empirical study of saving behavior in selected transition countries.

C. Prospects for Private Saving in Poland

International evidence on the behavior of private savings

4. Saving represents economic agents’ choice between current and future consumption. The theoretical literature, underpinned by the life cycle hypothesis (LCH) and the permanent income hypothesis (PIH), emphasizes the role of private saving in the intertemporal smoothing of the consumption rate, based on expected lifetime income. However, empirical evidence, based on the experience of the industrial and developing countries, is mixed on whether saving is a result of full intertemporal smoothing of lifetime consumption or depends mostly on current income. In this respect, liquidity constraints and uncertainty may play important roles that prevent full intertemporal smoothing (see Aghevli and others (1990) and Masson and others (1995) for an extensive review of the literature).

5. The empirical studies identified a large number of factors that may influence private saving decisions. Recent work by Masson and others (1995), the World Bank (1998), Savastano (1995), Callen and Thimann (1997), and Dayal-Gulati and Thimann (1997), among others, identified income, growth, government saving, tax and pension systems, financial liberalization, demographics, external factors, and uncertainty as key factors that influence saving behavior.

6. Per capita income is found to be positively correlated with private saving rates, with the effect being larger for developing countries than for industrial countries (Masson and others, 1995). Most empirical studies found that growth and saving are positively correlated. The World Bank study further showed that the causation is from growth to saving, rather than from saving to growth. It also found that sustained increases in growth are associated with permanent increases in the saving rate, while the increase in the saving rate is followed by an acceleration of growth for several years and rapid growth then disappears afterwards. The effect is substantial: a 1 percentage point increase in the growth rate raises the private saving rate by 1 percentage point (Rodrik, 1998). However, other studies found a negative relationship between growth and saving: in Latin American countries (Dayal-Gulati and Thimann, 1997) and in industrial countries (Masson and others, 1995). These results are consistent with the predictions of the LCH, as workers will want to consume more now if they expect their income will grow in the future.

7. The response of private saving rates to changes in public saving varies widely across countries. In general, a permanent increase in public saving will raise national saving, as one unit of a government deficit increase is associated with a decrease in consumption of only half the amount, rejecting the Ricardian equivalence hypothesis. Callen and Thimann (1997), in a study on household saving behavior in OECD countries, found that a high reliance on direct income taxes and high government transfers to households tends to be associated with low saving.

8. Age structure plays an important role in saving: a high proportion of the population of working age boosts the saving ratio and a high dependency ratio lowers it. Countries that experience a demographic transition as the working age population increases may have a transitory increase in their saving rates. An improvement in the terms of trade should boost saving and improve the trade balance (the Hargerger-Laursen-Metzler effect) through the positive effect on wealth and income, depending on the permanency of the terms of trade shift.

9. The effect of higher interest rates on saving is ambiguous, because higher interest income reduces saving (income effect), while a higher interest rate makes saving more attractive than current consumption (substitution effect). Financial liberalization usually provides a wide selection of instruments for financial saving and results in higher real interest rates on deposits (McKinnon-Shaw hypothesis), with an ambiguous effect on saving. In addition, liberalization may reduce the liquidity constraint faced by households, thus reducing saving. Easier access to credit reduces precautionary saving too. The World Bank study (1998) found that in general, a 1 percentage point increase in the ratio of private credit flows to income reduces the long-term private saving rate by 0.74 percentage point. An increase in uncertainty should boost saving as precautionary saving grows, as the World Bank study showed that inflation (a measure of macroeconomic volatility) is positively correlated with private saving, other things being equal. However, a study by Dayal-Gulati and Thimann found that macroeconomic stability (using inflation variability as a proxy) tends to spur private saving, because it is conducive to growth and saving. Unemployment was found to lower the household saving rates in OECD countries, as the impact from lower incomes dominated the positive effect from the increased need for precautionary saving (Callen and Thimann, 1997).

10. The effect of public pensions on private saving has generally been inconclusive. While the wealth effect could lower private saving as public pension benefits substitute for private provisions, the retirement effect (earlier retirement that results from the availability of the public pensions), and the improved awareness of the need for retirement saving could increase private saving. While higher rates of return on market-invested portfolios than those on the payroll contributions under pay-as-you-go systems could lead to lower private saving under fully funded systems, higher uncertainty associated with the market rates of return could boost the private precautionary saving. The World Bank study (1998) found that countries with pay as-you-go systems tend to have lower saving rates than those with fully funded systems, other things being equal. The impact of pension reform on national saving by moving from pay-as-you-go systems to fully funded systems is also ambiguous theoretically, with mixed country experiences. The direct, short-term effects of pension reform on saving will likely be dependent on how much the transfers to the old generation from the young are affected-since the old generation has a lower propensity to save than the young group; and on how the government finances the transition deficit-higher taxes tend to depress saving. Pension reforms in Chile raised the national saving rate substantially over a ten-year period, driven by an increase in public saving, which in turn led to high investment and growth and high private saving. However, other countries experienced lower national saving rates during their pension reforms (Rybinski, 1999).

Outlook for private saving in Poland

11. Taking the possible effects of these factors into consideration, the prospects for private saving in Poland over the medium term are subject to a large degree of uncertainty, since many factors are likely to have offsetting effects (Figures 5 and 6).

Figure 5.Poland: Trends in Income Level, Growth and Unemployment Rates, 1990–04

(In percent)

Citation: 2000, 60; 10.5089/9781451831887.002.A006

Sources: WEO and staff estimates.

Figure 6.Poland: Trends in Public Savings and Demography, 1990–04

Sources: WEO, World Bank World Development Indicators 1999, and UN Population Projections.

  • Higher per capita income and high growth are likely to raise the private saving rates, as most empirical studies found that income, growth and saving are positively correlated. However, LCH, supported by some empirical evidence, predicts a lower saving rate, as workers will want to consume more now if they expect their income will grow in the future. Privatization, which is expected to improve enterprise efficiency and income, will further boost household consumption.

  • The authorities’ plan to meet its medium term fiscal balance target will be likely to reduce private saving, but have a positive effect on total saving as the offset in household saving is likely to be incomplete, and reforms of the tax system and public expenditures will also be conducive to private saving increases2.

  • A growing working population will be likely to raise the private saving rate, while an increasing population of pensioners will be likely to have an opposite effect. The effect of the declining unemployment rates will depend on the relative importance of the income and precautionary effects on saving rates.

  • Financial liberalization, which will reduce the effect of the liquidity constraint on household consumption, tends to lower the saving rate.

  • The large improvement in Poland’s business saving rate during 1993-95, which has been the driving force underlying the improvement in the private saving rate, is unlikely to recur, because the room for its further improvement appears to be limited. Business profitability is high in Poland, as it has one of the highest shares of non wage income among the European countries: the share of compensation of employees and self-employment and property income received by households in GDP at factor cost is around 74 percent in Poland in 1999, compared with more than 80 percent in Hungary, France, Germany, Italy, Portugal, and Spain. The exceptions are the Czech Republic and Ireland, which have slightly lower shares (Figure 7 and Table 1). The high share of business profits has been a result of strong productivity growth, and attracted large amount of investment, including from foreign investors seeking highest rates of return on capital. However, strong domestic investment, continued large inflow of foreign capital, coupled with a limited supply of qualified labor, reduce the likelihood of further increases in the share of business profits in the future.

Figure 7.Poland: Compensation of Employees and Self Employment and Property Income Received by Households

(in percent of GDP at factor cost, average over 1991-99)

Citation: 2000, 60; 10.5089/9781451831887.002.A006

Source: OECD Economic Outlook database, 1999.

Table 1.Poland: Comparisons of Compensation of Employees and Self-Employment and Property Income Received by Households, 1991–99
(in percent of GDP; at factor prices)
Czech Republic61.963.565.165.465.567.669.668.669.366.3
United States87.787.186.485.885.985.685.686.487.086.4
United Kingdom96.495.691.590.
Source: OECD Economic Outlook database, 1999.

12. A simple econometric analysis is conducted to examine the possible effect of the set of determinants of saving on private saving rates in Poland.

13. A panel data set is used for the estimation, to take advantage of the variations both across countries and over time. A fixed-effect specification is used, allowing for country-specific intercepts and assuming that the slope coefficients are constant. The general form of the model estimated is:

Yiti+β’Xit+uit i=1,…, N, t=1, …, T,

where Yit is the matrix of the dependent variable, α the country-specific intercepts, β a vector of coefficients, Xit the matrix of explanatory variables, and uit the error term.

14. The dependent variable in the regressions is the ratio of gross private saving to GDP (PSAV). Estimates of the gross domestic saving used in this analysis are based on WEO data and calculated as gross capital formation plus the negative of the current account balance. Private saving is calculated as gross domestic saving less the general government surplus and gross public fixed capital formation. Therefore gross private saving (combining household and business saving) is used in the estimations3. The explanatory variables are the real GDP growth rate (Growth), the public saving in percent of GDP (PubSav), the income level per capita measured in relation to the United States at purchasing power parity exchange rates (Income), the dependency ratio defined as people aged 0-14 and 65 and over to the working age population (DepRatio), the unemployment rates (Unemploy), the ratio of bank credit to the private sector to GDP (Pcredit/GDP) as a proxy for financial liberalization, inflation rates (Inflation), and the terms of trade (TOT).

15. The countries included in the dataset are the Czech Republic, Estonia, Hungary, Poland, the Slovak Republic and Slovenia. The data cover the period 1992–99 and are taken from the World Economic Outlook (WEO) database, except for demographic data, which are from the World Bank World Development Indicators Database.

16. As usual, there are problems related to data quality and measurement issues. In addition, the limited amount of data available for the panel regression makes the results subject to a wide margin of errors. Nonetheless, results from a general model with all variables included, and a restricted model with insignificant variables dropped are presented in Table 2, including results adjusted for heteroscedasticity. All the coefficients for the country specific intercept terms are significant and not reported in the table.

Table 2.Poland: Panel Regression Results from a Sample of Six Transition Economies, 1992–99

(Dependent variable-private saving as a share of GDP)

Fixed Effects
Number of observations48484848
R2 adjusted0.980.980.970.97
Equation Standard Error2.342.342.392.39
Source: Staff calculations.Note: T-ratios in italics.

17. While the results confirm that most of the determinants of saving identified in the literature also apply to the transition economies included in the analysis, growth notably did not exhibit a positive effect on saving rates. The relative income variable has a strong and positive effect on saving rates. Government saving is negatively associated with private saving, with about half of the increase in government saving offset by the decrease in private saving. Surprisingly, the dependency ratio is positively correlated with private saving rates. This result could be related to the substantial change in the education, social welfare and pension systems in the transition economies, where these benefits are expected to decline and people may have responded by increasing their own provision for education and retirement. Declining unemployment rates increase private saving rates, indicating that income effect dominates the precautionary need for saving. The ratio of credits to private sector to GDP is negatively related to the private saving rates, indicating the effect from financial liberalization on private saving is likely to be negative. Inflation has a small, but negative effect on saving rates. The coefficient for terms of trade is not found to be significant (at the 10 percent level).

18. The estimated equation provides some guidance for the prospects for private saving in Poland, though the results need to be interpreted cautiously. It suggests that private saving rate could rise by about 2 percentage points from their 1999 level, if projections for the regressors in the WEO and World Population Projections from the United Nations are used.4 The key factor raising the saving rate in these projections is the increasing income level relative to the US. But this is offset by the negative effect on private saving rates from financial liberalization, and the partial offset of the targeted improvement in public saving.

D. Conclusions

19. As Poland embarks on the road to EU accession, investment is likely to grow rapidly, as its income converges to the EU levels. However, the prospects for substantially higher domestic private saving to meet the investment demand are uncertain, since qualitatively, many factors affecting private saving rates are likely to have offsetting effects. The empirical analysis also projected a limited improvement in private saving rates over the medium term. In addition, the prospect for EU accession will improve Poland’s access to foreign saving, which could in turn replaces domestic saving at least partially. To minimize the danger of relying on excessive foreign financing these uncertainties related to the prospect for private saving need to be taken into consideration when the government sets its fiscal path over the medium term.


Prepared by Guorong Jiang.

See Chapters 2 and 3 for discussions of the authorities’ plans for tax and expenditure reforms.

Auerbash and Hassett (1991) found evidence that households in the United States pierced the corporate veil and took into account business saving when making their consumption decisions. This finding suggests that it is appropriate to model private saving rather than household or corporate saving separately, though the extent to which this is valid in transition economies is unclear.

These projections for 1999 to 2004 are: income per capita in Poland will rise from about 30 percent relative to the United States to 35½ percent; growth will accelerate from 4.1 percent a year to 5.6 percent; public saving will improve from -0.4 percent of GDP to 5 percent; the unemployment rate will decline to 10½ percent; inflation will be reduced from 7.3 percent to 4 percent; the dependency ratio will fall from 48 percent to 46½ percent, and the ratio of credit to the private sector to GDP will rise from 28 percent to 40 percent.

Other Resources Citing This Publication