Australia: Selected Issues
Author: International Monetary Fund
Publication Date:
17 Apr 1997
eISBN:
9781451802016
Language:
English
Keywords:
ISCR; CR; government; enterprise; OECD level; return on assets; financing burden; real interest rate; economic growth; Private savings; Income; Income tax systems; Social security contributions; Australia and New Zealand
Download PDF (1.3 MB)

This Selected Issues paper examines the role that government policy in Australia plays in influencing household saving, both directly through its own saving and the structure of the tax, social security and welfare systems, and indirectly through the influence of the policy environment on factors that affect saving such as economic growth. The determinants of household saving in a sample of 21 OECD (Organization for Economic Cooperation and Development) countries are also investigated, using both cross-section and panel estimation techniques.

Abstract

This Selected Issues paper examines the role that government policy in Australia plays in influencing household saving, both directly through its own saving and the structure of the tax, social security and welfare systems, and indirectly through the influence of the policy environment on factors that affect saving such as economic growth. The determinants of household saving in a sample of 21 OECD (Organization for Economic Cooperation and Development) countries are also investigated, using both cross-section and panel estimation techniques.

I. Public Policy and Household Saving: Evidence from OECD Countries and Implications for Australia1

A. Introduction

1. There has been wide recognition in Australia that the low level of national saving has placed an excessive reliance on foreign saving to finance the investment needed to sustain growth. This recourse to foreign saving has resulted in a significant buildup of net external liabilities, which imposes costs on the economy by making it more vulnerable to external shocks and swings in investor sentiment, and raises the interest premia on borrowing.2 The decline in domestic saving in Australia has been caused by both lower public and private saving; the bulk of the decline in private saving has been due to a trend decline in household saving. Economic policies have increasingly sought to raise national saving, including through the introduction of compulsory employer and employee superannuation contributions, although success has so far been limited.

2. In Australia, household saving has declined sharply since the 1970s. This decline has been greater than in most other industrial countries, and has left Australia with one of the lowest household saving rates in the OECD. The focus of this paper is the role that government policy plays in influencing household saving, both directly through its own saving and the structure of the tax, social security and welfare systems, and indirectly through the influence of the policy environment on factors that affect saving such as economic growth. In order to address these issues, the determinants of household saving in a sample of 21 OECD countries are investigated, using both cross-section and panel estimation techniques. The widely differing experiences in household saving performance among these countries, as well as within countries over time, suggest that such an international perspective may provide useful insights into the determinants of household saving.

3. The rest of the chapter is organized as follows: Section B looks at recent trends in household saving; Section C discusses the influence of government policy on household saving; Section D presents the empirical results of the cross-country and panel estimations; and Section E concludes by drawing some implications for Australia.

B. Household Saving in Australia in an International Context

4. In Australia, household saving has declined substantially over the past 20 years. On average, in the second half of the 1970s, Australian households saved 11 percent of their disposable incomes; during the 1980s, they saved only 7 percent on average, and in the first half of the 1990s, the household saving ratio fell to below 4 percent (Chart I.1). This decline was only partially offset by higher corporate saving; consequently, private saving declined from 19 percent of GDP in the late 1970s to 14 percent of GDP in 1995.

CHART I.1
CHART I.1

AUSTRALIA: SAVING RATES, 1975–95

Citation: IMF Staff Country Reports 1997, 023; 10.5089/9781451802016.002.A001

Sources: Australian Bureau of Statistics; OECD Analytical Database; and WEO Database.

5. At the OECD level, there is a substantial degree of variation of household saving rates, both among countries and over time (Table I.1). Three observations are evident from the data:

  • • At any point in time, household saving rates differ greatly among OECD countries. Over the period 1990–95, for example, the average household saving rate in 21 OECD countries was 10½ percent. However, while households in Belgium saved 19½ percent and those in the United Kingdom 10½ percent, in Australia and New Zealand they saved only 3¾ percent and 1¾ percent of their disposable income, respectively.

  • •There has been a slight trend decline in household saving rates in most OECD countries over the past two decades. The OECD average declined from 13 percent in the second half of the 1970s to 10½ percent in the early 1990s.

  • • Despite the general trend decline, most countries have kept their relative position through time. For example, households in Japan and many continental European countries have remained relatively high savers, whereas the United States and the Scandinavian countries have remained relatively low savers. The two most notable exceptions are Australia and New Zealand, which have fallen from a middle position to the bottom end.

Table I.1.

Household Saving Rates in OECD Countries

(In percent of disposable income)

article image
Sources: IMF, WEO database; OECD, Analytical database.

6. Of course, there are a number of definitional issues that may affect the measurement of household saving over time. Because household saving is generally defined to include the saving of unincorporated enterprises, any changes to the incentives to incorporate over time (for example, for tax purposes) will affect the split. Also, structural changes in the economy, such as the declining importance of unincorporated farms in the agricultural sector in Australia, may lead to a decline in measured household saving.3 Further, on a national accounts basis, data may not be strictly comparable across countries as the treatment of particular items can differ. However, only a relatively small share of the divergence in saving rates across countries is due to differences in national accounting (Elmeskov and others (1991)). Because of these data and measurement issues, several studies have chosen to focus on private saving as a whole. However, the drawback with this approach is that it does not distinguish between the factors affecting household and corporate saving, which are likely to be considerably different. Consequently, while recognizing the measurement issues, this paper focuses on household saving.

C. Public Policy and Household Saving

7. There has been an increasing interest in Australia, as in many other countries, about the influence of public policy on household saving behavior. Public policy may directly influence household saving incentives through a number of channels:

  • (i) Government saving: a direct link between government and private sector saving is often hypothesized. An increase in the government deficit due to a reduction in taxes may be offset by an increase in private sector saving, as higher tax liabilities in the future are anticipated (the well-known Ricardian equivalence argument). Most studies of private saving, (for example, Masson and others (1995)) have found that, while such an offset exists, it is only about one half rather than unity as implied by full Ricardian equivalence.

  • (ii) Policy toward private pension provision: policies that make contributions to private pension schemes compulsory will affect the composition of private saving, and, if voluntary saving is not reduced by the full amount of the compulsory contribution, also its level.

  • (iii) The structure of the tax, social security and welfare systems, and the extent of other government social expenditure. The following paragraphs elaborate on this:

8. The structure of the tax system may influence saving both by changing lifetime wealth and by affecting the rate of return on saving (Boadway and Wildasin (1994)). Under an income tax system, saving is effectively taxed twice—once on the income from which the saving is made, and once on the return (if it is included in the definition of taxable income)—whereas a consumption tax applies only once, when income is used for consumption. For a given total tax level, a shift from an income to a consumption tax therefore reduces the wedge between the pre- and post-tax rate of return on assets. There are two further reasons why income taxes may be detrimental to saving: first, since they are generally progressive, high-income households—generally the high savers—are affected more;4 second, the working age population—which comprises the high-saving age groups—pay the bulk of direct taxes. Indirect taxes, in contrast, are more evenly distributed across income and age groups.

9. The coverage and generosity of the social security and welfare system and the extent of other government social expenditures may influence household saving in a number of ways. Governments typically provide a range of benefit payments that reduce the need for private asset accumulation for retirement and that cushion individuals from potential income losses (for example, unemployment) or from large unanticipated expenditures (for example, health expenditures). Public pension benefits can substitute for private saving as long as the anticipated benefits, net of contributions, have a positive present value. The extension of public pension schemes could thus lower household saving (the wealth effect)—however, if it also leads to earlier retirement (the retirement effect) or increases the awareness to provide for old-age resources, it may actually lead to higher private saving. Existing empirical evidence on the impact of public pensions on household saving has generally been inconclusive. Feldstein (1980) argued that public pension schemes have a negative impact on private saving, but his findings have been disputed on empirical and theoretical grounds (Koskela and Viren (1983)). Others (for example, Kopits and Gotur (1980)) have found social security arrangements to generally boost saving. Other benefits, such as unemployment benefits or welfare benefits, tend to lower the need for private precautionary saving. Given that a number of studies have found a significant proportion of household saving to be of a precautionary nature (for example, Skinner (1988)), policies that alter the need for precautionary saving could have a significant impact on the level of household saving. Governments also provide substantial assistance to households through the free or subsidized provision of goods and services—“in-kind” benefits—such as free education, university scholarships, free or subsidized health care, and public housing. Such benefits lower the need for private saving to cover expenditures in these areas.

10. The impact that the social security and welfare system will have on individual saving behavior is likely to be dependent on a number of features of the system including: the value of the benefit payments (the higher the replacement ratio—defined as the entitlement as a percent of previous earnings—the less incentive there is for private provision); the length of time over which payments are available; the certainty with which people regard the future payment of such benefits; and the general availability of such payments, that is, the extent of “means-testing.”

11. The financing of the social security and welfare system may also affect household saving in an important way. The system can be financed either from general tax revenue or from specific social security contributions. However, while social security contributions are generally levied as a fixed proportion of income, tax rates are generally progressive. Further, compulsory contributions generally begin at very low income levels and are capped at high incomes; by contrast, income taxes generally exempt very low incomes and are not capped at high incomes. Therefore, tax financing shifts the financing burden toward higher-income earners relative to contribution financing. Since low-income households have been found to save very little, and saving rates rise strongly with incomes, a higher reliance on tax financing may lower aggregate household saving. It is, therefore, important to distinguish between gross government transfers to households and net transfers (which are calculated as gross transfers less social security contributions, and measure the amount of benefits financed from general tax revenue and paid directly from the budget).

12. In Australia, there is a heavy reliance on direct taxes, with more than half of general government revenue being raised through this source.5 The United States, Canada, and New Zealand also have similar tax structures. The position is significantly different in many European and Scandinavian countries, where less than one-third of tax revenue is raised from direct taxes (Chart I.2).

CHART I.2
CHART I.2

AUSTRALIA: TAX STRUCTURE IN 21 OECD COUNTRIES

(In percent of GDP)

Citation: IMF Staff Country Reports 1997, 023; 10.5089/9781451802016.002.A001

Source: International Monetary Fund, Government Finance Statistics.

13. The Australian social welfare system is entirely financed from government tax revenue. Receipt of a welfare or pension payment is conditional on a range of qualifying conditions being met (a “means” test). The use of means-testing results in the greater targeting of welfare payments and, in general, the Australian system has as its main focus the alleviation of poverty. The system differs from that in most other industrial countries where specific social security contributions are levied and many benefits are paid from the social security system. The main purpose of this type of social security system is income maintenance, rather than poverty alleviation, and benefit payments are conditional on establishing a contribution history, usually based on some average (from a few years to a lifetime) of past earnings, and are not means-tested.

14. A large increase in both gross and net social security and welfare transfers to households is apparent in many OECD countries since the 1970s (Chart I.3). The increase in gross transfers indicates that a growing share of resources is being channeled through the public sector for income maintenance, while the increase in net transfers indicates that the share of social security and welfare spending covered by contributions has fallen, implying that financing from general tax revenue has increased. While some of this is due to demographic factors and the rise in unemployment, the bulk of it is due to the increase in the comprehensiveness and generosity of the welfare system. While gross transfers in the continental European countries are significantly higher than in other countries, in terms of net transfers the situation is reversed as those countries with high gross transfers also have high social security contribution rates. In countries such as Australia and New Zealand, where social welfare spending is made directly from general revenues, net transfers households (in percent of GDP) are the highest in the OECD and the difference between gross and net transfers is small (Chart I.3, bottom panel).

CHART I.3
CHART I.3

AUSTRALIA: TRANSFERS IN THE OECD

(In percent of GOP)

Citation: IMF Staff Country Reports 1997, 023; 10.5089/9781451802016.002.A001

Source: OECD Analytical Database.1/ Net transfers, defined as gross transfers net of social security contributions paid by households, indicate the amount of transfers financed from general tax revenues.

15. Commonwealth government social expenditures in Australia rose substantially in the mid-1970s, from around 9 percent of GDP in 1973/74 to 15 percent of GDP in 1977/78, primarily due to higher spending on social security and welfare (Chart I.4). Since then, expenditures have increased slightly father to about 17 percent of GDP in 1995/96. While efforts have been made to more effectively target social expenditures, these have met with mixed success. Evidence suggests that while cash transfers are well targeted at lower-income groups, expenditures in areas such as health and education are less well focused on those groups.6

CHART I.4
CHART I.4

AUSTRALIA: COMMONWEALTH GOVERNMENT SOCIAL SAFETY NET EXPENDITURE, 1973/74–1995/96

(In percent of GDP)

Citation: IMF Staff Country Reports 1997, 023; 10.5089/9781451802016.002.A001

Source: Australian Treasury, various budget statements.

16. Household saving is also likely to be influenced by a number of factors that are not directly controlled by policymakers, but on which the overall policy framework has an important influence. Of particular importance in this category is economic growth. While theoretically and empirically the impact of growth on saving is not clear,7 saving and growth are highly correlated over longtime horizons as well as for many regions and stages of development, with higher rates of growth being associated with higher saving (see Schmidt-Hebbel and others (1996)). Inflation and unemployment may also have an important influence on saving. Economic policies that are conducive to increasing the rate of economic growth and maintaining low and stable inflation should have a positive impact on saving. In particular, the application of consistent monetary and fiscal policies that engender a stable financial environment is important (Dayal-Gulati and Thimann (1996)). Further, microeconomic reforms (including labor market deregulation, trade policy, and privatization) that increase competition in domestic markets will also help to raise the rate of growth over the medium term.

17. Financial deregulation also influences household saving indirectly. Development of the financial system may increase the opportunities for financial saving. However, the liberalization of consumer access to credit may ease liquidity constraints faced by households (which had effectively limited their ability to borrow against future expected income) and could therefore, at least initially, lead to lower household saving.8

D. An Empirical Analysis of Household Saving

18. This section investigates the empirical determinants of household saving in the 21 OECD countries presented in Table I.1, through both a cross-section and a panel data analysis over the period 1975–95. The equation specification incorporates variables that have been found to be important determinants of saving behavior in previous empirical studies,9 and additional variables to capture the structure of the tax system, the structure and generosity of the social security and welfare system, and to proxy financial deregulation. The general specification is:

(1)Hsavei=α+β1GovSavi+β2CorpSavi+β3IncLevi+β4InGrowi+β5Unempi+β6Re⁡alri+βiInflati+β2OldDepi+β9TaxStructi+β10Transferi+β11ConsCrediti+∈

where I is the country index and the variables are as follows:10

article image
article image

19. In the cross-section analysis, regressions were run with each of the variables included as its average value over the time period of the regression. Consequently, there were 21 observations for each variable. Five sets of regressions were estimated: averages of the 21 countries over the full-time horizon (1975–95), over three sub-periods which coincide closely with the economic cycle (1975–81,1982–89,1990–95), and a pooled regression where the 21 observations from each of the three sub-periods were used to yield a total of 63 observations. Estimating the equation over each of these different periods provides a check on the robustness of the results and provides some information on the stability of the long-run relationships in the data.

20. The panel regressions add a time-series dimension to the analysis. As with any time-series regression, nonstationarity in the data may make coefficient estimates and t-statistics unreliable (Matyas and Sevestre (1996)). However, in the equation, the dependent and explanatory variables are either expressed as growth rates or ratios (which are bounded in the long run), rather than absolute values, to avoid nonstationarity, at least over longer-time horizons. Potential problems with heteroskedasticity should also be reduced by the relatively homogenous sample of OECD countries where the underlying data generation processes are likely to be relatively similar so that coefficients can be expected to vary randomly, but not systematically, across countries. The fixed effects model, which allows for different country intercepts, is used in the estimations.

21. The following principal results emerge robustly from the cross-section and panel estimations (Tables I.2 and I.3): 11,12

  • the coefficient on government saving is significant, indicating that there is some offset in household saving to changes in government saving. The size of the estimated offset is much larger in the cross-section than in the panel results.13

  • the offset coefficient on corporate saving is significant but smaller than unity, suggesting that in their saving behavior households partially see through the corporate “veil.”

  • the rate of income growth is positively correlated with household saving. In the cross-section results, the negative coefficient on the level of per capita income relative to the United States points to a convergence effect with households in less prosperous countries tending to save more than their counterparts in richer countries.

  • a standard demographic impact is found, with a higher old-age dependency ratio associated with lower household saving. The young-dependency and the middle-age dependency ratios were also tested, but were not significant.

  • the direct tax ratio is negatively and significantly correlated with household saving, but indirect taxes are not significant.14 The ratio of direct-to-indirect taxes was also significant in the regressions.

  • net transfers are negative and significant in the cross-section results, whereas gross transfers are negative and significant in the panel results.15 A possible interpretation of this difference is that over a short-time horizon (captured in the panel) households respond to changes in transfers as they affect their current incomes whereas, over the longer term, they realize that they have to finance the additional payments through higher contributions and only react to the net amount. Unemployment benefit replacement rates and pension replacement rates were also tried on a more limited cross-section sample, but the results were too sensitive to changes in the equation specification to yield any robust conclusions.

Table I.2.

Results of Cross-Country Regressions Dependent Variable: Household Saving/GDP

article image
Note: T-statistics in brackets; the constant term is not reported throughout.
Table I.3.

Panel Estimation Results Sample: 21 Industrial Countries, 1975–95

article image
Note: T-statistics in brackets. The instrumental variables used are the first lags of the general government balance, corporate saving, and income growth.

22. Some results are mixed:

  • The consumer credit variable was negative and significant in the panel, but its presence caused large changes in otherwise robust variables, particularly the old-age dependency ratio, suggesting a high degree of correlation with other explanatory variables.16 As with the old-age dependency ratio, the ratio of outstanding consumer debt to GDP has trended upward in many countries in the sample.17 The credit card variable was insignificant.

  • Inflation was only significant in the panel estimations where it is probably proxying for measurement biases in national account measures of saving caused by the nominal component in interest payments (see Masson and others (1995)). The unemployment rate was only significant in the cross-section, with a negative correlation with household saving. The real interest rate was insignificant in both regressions (insignificant variables in the cross-section are not reported in Table I.2).

23. As discussed earlier, there are a number of potential problems in the definition of household saving both over time in one country and across countries. As a check on the robustness of the results, the instrumental variable specification in Table I.3 was reestimated using private saving rather than household saving as the dependent variable. The results are largely unchanged (last column in Table I.3).

E. Conclusions and Policy Implications

24. The results presented in this paper suggest that the structure of the tax, social security and welfare system, as well as factors such as growth, demographics, and public and corporate saving, may have a significant impact on household saving. While it is obviously difficult to infer conclusions for any one country from a cross-country analysis in which all coefficients represent averages, the results suggest a number of factors that may help explain the decline in household saving in Australia:

  • while the Australian welfare system is relatively well targeted, and does not appear generous by OECD standards, expenditures have risen sharply since the mid-1970s. The system is also financed directly from the budget, rather than by social insurance contributions, and this may, because of high levels of net transfers, also have negative implications for saving as discussed earlier. Furthermore, there may be features of the social security and welfare systems that are important in influencing incentives, but that are not easily captured in an aggregate analysis. This is particularly true in terms of the impact of means-testing, how effectively work tests are enforced, and the duration over which some benefits are available.

  • the high reliance on direct taxes in Australia may act as a disincentive to saving, although the tax structure has changed little in recent years (Pender and Ross (1994)).

  • real GDP growth has declined from an average of 5 percent in the 1960s and early 1970s to around 3 percent over the past decade. The slowdown in per capita growth has been even more marked over this period, with Australia having had one of the lowest per capita growth rates in the OECD in the 1970s and 1980s. This paper, along with most other studies, finds an important relationship between growth and saving.

  • demographic factors have also contributed. Although the Australian population is relatively young compared to many European countries and Japan, it has become older faster, with the increase in the old-age dependency ratio exceeding that in many other countries. A related issue is the impact on household saving of the increasing trend toward early retirement in Australia.

25. In an effort to boost national saving, the Australian government has recently introduced a system of compulsory superannuation contributions. Under the scheme, up to 15 percent of employees’ gross salary will be contributed to a superannuation fund by 2002.18 In effect, this policy will result in a pension system similar in many respects to those operating in most other OECD countries, although it will be privately operated, fully funded, and replacement rates will depend on the rate of return earned on contributions. While the policy can be expected to raise national saving over the medium term (the government estimates by about 1 percent of GDP by 2005) and reduce the future burden on the government budget, the actual size of the impact on saving will depend on a number of factors, including the offset in discretionary private saving and the preservation rules that are applied. Currently, private pension benefits are usually taken as a lump sum, rather than an annuity, and only have to be preserved in the superannuation fund until age 55, while the eligibility age for the public pension is 60 for women (this is being gradually raised to 65 by 2005) and 65 for men. This encourages some people to retire early, run down their lump-sum private payout, and then claim the old-age pension (so-called “double-dipping”).

26. The results here also suggest that there are a number of other ways the government could help to raise household saving. Policies that improve the climate for growth will provide a boost to saving. Macroeconomic policies that maintain a low and stable inflation climate, and microeconomic policies that increase competition in factor and product markets, are both important in this regard. Further, reforms to the tax and welfare arrangements may improve saving incentives as well as help raise public saving, thus providing an important boost to national saving. Of course, such changes may also have important social implications which need to be considered against the benefits of higher saving.

APPENDIX: Data and Sources

The following data and sources were used: IMF, World Economic Outlook Database (WEO); IMF, International Finance Statistics (IFS); IMF, Government Finance Statistics Database (GFS); OECD, Analytical Database (OECD); United Nations, World Population Database (UN); and individual country central bank bulletins.

Description of the variables (with the source in brackets):

article image

References

  • Australia, Department of Social Security, 1992, International Social Security Review, Vol. 45, pp. 83–100.

  • Bateman, H., J. Frisen, G. Kingston, and J. Piggott, 1990, Demographics, Retirement Saving, and Superannuation Policy: An Australian Perspective, Discussion Paper No. 241, Centre for Economic Policy Research (Canberra: The Australian National University).

    • Search Google Scholar
    • Export Citation

  • Bayoumi, T., 1993, “Financial Deregulation and Household Saving,” Economic Journal, Vol. 103.

  • Boadway, R., and D. Wildasin, 1994, “Taxation and Savings: A Survey,” Fiscal Studies, Vol. 15, No. 3, pp. 19–63.

  • Bosworth, B., 1993, Saving and Investment in a Global Economy (Washington: Brookings Institution).

  • Carnahan, M., and L. Camilleri, 1995, “A Reconsideration of the Saving Growth Relationship,” Paper presented at the twenty-fourth Annual Conference of Economists, September.

    • Search Google Scholar
    • Export Citation

  • Carroll, C, and D. Weil, 1993, “Saving and Growth: A Reinterpretation,” NBER Working Paper No. 4470 (Cambridge, Massachusetts: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation

  • Dayal-Gulati, A., and C. Thimann, 1996, Saving in Southeast Asia and Latin America Compared; Searching for Policy Lessons, Paper prepared for the IMF-Bank Indonesia Conference on Macroeconomic Issues Facing ASEAN Countries, held in Jakarta, November 7–8.

    • Search Google Scholar
    • Export Citation

  • Economic Planning Advisory Council (EPAC), 1988, Trends in Private Saving, Council Paper No. 36 (Canberra: Australian Government Publishing Service).

    • Search Google Scholar
    • Export Citation

  • Edey, M., and M. Britten-Jones, 1990, “Saving and Investment,” in The Australian Macro-Economy in the 1980s, pp. 79–157 (Sydney: Reserve Bank of Australia).

    • Search Google Scholar
    • Export Citation

  • Elmeskov, J., J. Shater, and W. Tease, 1991, “Savings Trends and Measurement Issues,” Economics and Statistics Department Working Papers No. 105 (Paris: OECD).

    • Search Google Scholar
    • Export Citation

  • Eslake, S., 1996, “The 1990s Economic Cycle and the Challenges Facing the New Government,” presentation to the Real Estate Institute of Australia, July.

    • Search Google Scholar
    • Export Citation

  • Feldstein, M., 1980, “International Differences in Social Security and Saving,” Journal of Public Economics, Vol. 14.

  • Fitzgerald, V., 1993, National Saving—A Report to the Treasurer (Canberra: Australian Government Publishing Service).

  • Knox, D.M., 1996, “Contemporary Issues in the Ongoing Reform of the Australian Retirement Income System,” The Australian Economic Review, pp. 199–210.

    • Search Google Scholar
    • Export Citation

  • Kopits, G., and P. Gotur, 1980, “The Influence of Social Security on Household Savings: A Cross-Country Investigation,” IMF Staff Papers, Vol. 29, No. 1 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Koskela, E., and M. Viren, 1983, “Social Security and Household Saving in an International Cross Section,” American Economic Review, Vol. 73.

    • Search Google Scholar
    • Export Citation

  • Masson, P., T. Bayoumi, and H. Samiei, 1995, “Saving Behavior in Industrial and Developing Countries,” Staff Studies for the World Economic Outlook (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Matyas, L., and P. Sevestre, 1996, The Econometrics of Panel Data: Handbook of Theory and Applications (Dordrecht and Norwell MA: Kluwer Academic Publishers).

    • Search Google Scholar
    • Export Citation

  • Morling, S., and R. Subbaraman, 1995, Superannuation and Saving, Research Discussion Paper RDP 9511 (Sydney: Reserve Bank of Australia).

    • Search Google Scholar
    • Export Citation

  • Organization for Economic Cooperation and Development, 1994, Taxation and Household Saving (Paris: OECD).

  • Pender H., and S. Ross, 1994, Taxation, Regulation and Private Saving in Australia, EPAC Background Paper No. 36 (Canberra: Australian Government Publishing Service).

    • Search Google Scholar
    • Export Citation

  • Poterba, J., 1994, Public Policies and Household Saving (Chicago: The University of Chicago Press).

  • Schmidt-Hebbel, K., L. Serven, and A. Solimano, 1996, “Saving and Investment: Paradigms, Puzzles, Policies,” The World Bank Research Observer, Vol. 11.

    • Search Google Scholar
    • Export Citation

  • Skinner, J., 1988, “Risky Income, Life Cycle Consumption, and Precautionary Savings,” Journal of Monetary Economics, September, pp. 237–55.

    • Search Google Scholar
    • Export Citation
1

This chapter was prepared by Tim Callen and Christian Thimann. The chapter was presented at a seminar during the 1996 Article IV consultation in Australia, and the authors are grateful to the discussants for helpful comments and suggestions.

2

The decline in national saving since the 1970s and its implications for the current account deficit and external liabilities were discussed extensively in the paper on Selected Issues for the 1995 Article IV consultation (SM/96/62,3/14/96).

3

Because the farm sector is generally thought to be a higher saver than the household sector as a whole. However, most of the decline in unincorporated farms in Australia took place in the 1960s and early 1970s. By 1975, the beginning of the sample period of this paper, unincorporated farms accounted for only a negligible share of household incomes.

4

Information from household surveys provided in Poterba (1994) shows that household saving rates in the highest-income quintile are, for example, 17 percent in Canada and Germany; 24 percent in the United Kingdom; and 42 percent in Japan. In contrast, the saving rates of households in the lowest-income quintile in these countries are close to zero and even negative in Canada and Germany. The age-specific saving rate peaks in the 55–59 age cohort for the G-7 countries, except Italy (65–69 years age cohort).

5

However, there are many exemptions in the Australian tax system which remove the double taxation on many different types of saving vehicles (particularly favored are owner-occupied housing and superannuation).

6

Eslake (1996) calculates that the top 40 percent of households by income received nearly $A 12 billion (2¾ percent of GDP) of “in kind” benefits from the Commonwealth government, a large part in the form of health spending.

7

The Life Cycle Hypothesis suggests that higher income growth would, for a given saving rate in each group, raise aggregate saving by increasing the incomes of those in work relative to those not working. However, it is also possible for saving rates within the working population to decline if workers anticipate higher future income and thus increase their current consumption. At an empirical level, Carroll and Weil (1994) find growth to Granger-cause saving, but saving not to Granger-cause growth. Carnahan and Camilleri (1995), by contrast, find little evidence of causation in either direction.

10

Data sources are described in the Appendix.

11

The fit of the cross-section equation for the 1990–95 sub-period is significantly worse than in the two earlier sub-periods. Masson and others (1995) found a similar result for private saving. Heteroskedasticity was rejected in all specifications on the basis of a White test.

12

For the panel estimation, three sets of results are presented: a general equation; a restricted equation (dropping the insignificant variables); and the restricted equation reestimated using instrumental variables to allow for the possibility that household saving is determined simultaneously with some other variables in the regression, particularly income growth, the government balance, and corporate saving (the first lag of each of these variables were used as instruments). The common intercept model was also tested and rejected for all specifications.

13

The lower offset of -0.3 in the panel estimates is much more consistent with some studies for Australia which have suggested there is little offset in private saving to changes in public saving (see Edey and Britten-Jones (1990)). Fund staff estimates from time-series analysis suggest an offset of about one-fourth for Australia, although the significance of the effect is sensitive to the equation specification.

14

Several studies have found that indirect taxes are not positively correlated with saving. The finding here is consistent with this, but suggests that, to the extent they allow a decline in direct taxes, saving will increase.

15

In the cross-section, a possible correlation between the unemployment rate and government transfers is largely excluded by the use of net transfers (contribution rates are generally raised with higher unemployment, so that net transfers exhibit little correlation with the unemployment rate). However, in the panel, the inclusion of the unemployment rate biases downward the coefficient on transfers, suggesting some positive correlation between unemployment and gross transfers. The unemployment rate is dropped in the restricted model.

16

The estimation for the equation including the proxy for financial deregulation was restricted to 1985–95, because only limited data was available prior to this. Household debt was lagged in the estimations, to avoid correlation between contemporaneous changes in debt and changes in the saving rate.

17

Including consumer credit significantly reduces the size of the fixed effect dummy for the Australia, the United Kingdom, and the United States, suggesting that some important information is added by the inclusion of this variable for these countries.

18

The 15 percent contribution will comprise 9 percent from employers, 3 percent from employees, and a means-tested 3 percent from the government. See Bateman and others (1990) and Knox (1996) for a discussion of the changes in superannuation policy.

Same Series

Other IMF Content

Other Publishers

Cover IMF Staff Country Reports
Australia: Selected Issues
Author: International Monetary Fund
Volume/Issue:
Volume 1997: Issue 023
Publisher:
International Monetary Fund
ISBN:
9781451802016
ISSN:
1934-7685
Pages:
104
DOI:
https://doi.org/10.5089/9781451802016.002
  • View in gallery

    AUSTRALIA: SAVING RATES, 1975–95

  • View in gallery

    AUSTRALIA: TAX STRUCTURE IN 21 OECD COUNTRIES

    (In percent of GDP)

  • View in gallery

    AUSTRALIA: TRANSFERS IN THE OECD

    (In percent of GOP)

  • View in gallery

    AUSTRALIA: COMMONWEALTH GOVERNMENT SOCIAL SAFETY NET EXPENDITURE, 1973/74–1995/96

    (In percent of GDP)

AUSTRALIA: SAVING RATES, 1975–95

AUSTRALIA: TAX STRUCTURE IN 21 OECD COUNTRIES

(In percent of GDP)

AUSTRALIA: TRANSFERS IN THE OECD

(In percent of GOP)

AUSTRALIA: COMMONWEALTH GOVERNMENT SOCIAL SAFETY NET EXPENDITURE, 1973/74–1995/96

(In percent of GDP)