3.6  Young dependants and old dependants

The complements of measures of the population in the working ages are measures of the dependent population. The most frequently-cited such measure is the “dependency ratio”, generally defined as the ratio of the population outside the working ages to the population inside the working ages. A fall in the proportion inside the working ages implies a rise in the dependency ratio. As can be seen in Panel (ii) of Figure 5, the proportion in the working ages is projected to begin falling in New Zealand, and in all the other selected countries and regions, some time within the next 20 or 30 years. However, the projected declines are not large, and would only return most countries and regions to the levels of the 1970s.

As Panels (i) and (iii) of Figure 5 show, the reason the dependency ratio is relatively stable is that increases in old-age dependency are offset by decreases in youth dependency. This implies that the balance between young dependants and old dependants is shifting towards old dependants. Figure 9, showing old dependants as a percentage of total dependants, bears this out. In contrast to the changes in the share of dependants in the total population, the changes in the share of old dependants in total dependants are large. In 1950, young dependants predominated throughout the OECD, but by 2050 old dependants are projected to make up 60%-70% of all dependants in Europe and Japan and around 50% elsewhere, including New Zealand. If, as many demographers believe, the UN has underestimated the scope for future mortality declines, the share of old dependants will be even higher than Figure 9 suggests.

Under current patterns of expenditure, the shift to old-age dependency is probably increasing the amount of resources transferred to dependants by the state. Across the OECD, government social expenditures per old person are much larger than expenditures per young person (Weil 1997: Table 10). Illustrative data for New Zealand are shown in Figure 10. Expenditures increase sharply between ages 60-64 and 65-69 as people become eligible for New Zealand Superannuation, and climb steadily after that, reflecting health care costs. As the share of the population in these high cost age groups increases, overall social expenditures can be expected to rise.

Figure 9 – Estimates and projections for the population aged 65+ as a percentage of population aged 0-19 or 65+, OECD countries and regions

Source: Calculated from data from the UN Population Division’s World Population Prospects online database.

Figure 10 – Government social expenditure per person per year, New Zealand 2001

Sources: All data apart from health expenditure data were obtained from Creedy and Scobie (2002: Appendix Tables 4,5). Health expenditure data were obtained from the New Zealand Treasury’s Long-Term Fiscal Model (www.treasury.govt.nz/ltfm/default.asp). Combined male and female totals were obtained by weighting by population numbers from the 2001 Census, available from the Statistics New Zealand website www.stats.govt.nz.

A recent paper by Razin, Sadka, and Swagel (2002) challenges this conclusion. Razin et al construct a political-economic model in which, under certain parameter settings, an increase in the old-age dependency leads a majority of voters to support a decrease in social expenditures per capita. Razin et al carry out a panel regression using data for 13 OECD countries, and obtain results suggesting that rising dependency is indeed associated with falling social expenditures per capita. Razin et al’s regression does not, however, distinguish between changes in overall dependency due to rises in old age dependency and changes due to falls in youth dependency. When Razin et al’s regressions are adjusted to allow for this distinction, rising old age dependency is found to be associated with rising, not falling, social expenditure per capita (Bryant 2003b). The standard assumption that increases in old-age dependency imply increases in fiscal pressures appears to be vindicated.

Governments of countries where a shift towards old-age dependency is creating pressure for increased social expenditures face unpalatable choices: holding expenditure down regardless, cutting other areas of government expenditure, raising taxes, building up debt, or doing some combination of the above. There is, of course, extensive debate among economists about the extent to which these various measures reduce welfare or reduce participation and productivity. For instance, some critics of the World Bank (1994) report Averting the Old Age Crisis argue that the Bank exaggerates the harm done by higher tax rates (Beattie and McGillivray 1995). The Bank has, however, stood its ground (James 1996). Indeed, it is hard to believe that increased taxes would not have some effect on participation or investment decisions, and that increased public expenditure on pensions and health care would not partly crowd out productivity-enhancing public expenditure on infrastructure, schools, and research. To the extent that fiscal pressures do threaten participation and productivity, the somewhat lower old-age dependency rates in New Zealand, Australia, the RDCs, and North America are likely to give them a slight economic advantage over Europe and Japan.