4  Exploring the Cohort Patterns (continued)

Notes: Refers to lifetime saving rate relative to the 1910-14 reference group (by definition zero). See pattern of cohort dummies in the regressions presented in Section 3. Negative and positive refer to the cohorts that were typically significantly lower or higher in their lifetime saving rates.

In what follows we examine some selected aspects of the economic and social environment facing the different cohorts both over their working lives as a whole, and in particular during their peak saving years. The question posed is the following: do those indicators vary in a way that is consistent with the observed cohort patterns in saving rates? We would expect to find that the proxies chosen for the economic and social environment adopted values less favourable for household saving rates during the critical years of the low saving cohorts, while the same indicator should be more favourable in years corresponding to the high saving cohorts.

Because of the magnitude of the task of assembling annual data on a wide range of variables in a consistent manner for 70 years, we have chosen to use selected years to illustrate the results. We focus on three cohorts: 1 (born 1910-14), 4 (born 1925-29) and 9 (born 1950-54), and will refer to these as the reference, early and late cohorts. Typically we will look at the values of the indicator variable prevailing during their peak saving years (given in Table 5).

We start with some key indicators relating to the labour market. The extent of unemployment is a critical factor affecting the expected flow of earnings. Those cohorts facing a lower probability of unemployment would be expected to have less incentive for precautionary saving. The unemployment rates (based on the average of the census years) facing the reference and early cohorts were 1.2 and 3.3 percent respectively, while based largely on projections the late cohort could face an average of 6 percent[29].

Prior to the major reforms of the late 1980’s, the New Zealand labour markets were characterised by central wage fixing, limited flexibility and a high degree of unionisation (see Figure 7). Strong national unions were able to bargain particularly with state sector employers (public works, power generation and distribution, the Post Office, forestry and railways) and gain job security for their members.[30] It would seem plausible that the job security (at least perceived) which the unions were able to achieve, might have reduced the incentive for precautionary saving by their members.

Figure 7 – Union density and membership in New Zealand: 1936-1999

(Membership: RHS; Density (%): LHS)

A further important labour market indicator is the rate of participation.[31] The reference group had labour force participation rates of 94 and 35 percent for male and females workers. In contrast the early cohort faced 89 and 46 percent. The significant increase for women would be consistent with the lower saving rates of this group. The late cohort could face rates of 85 and 70 percent, suggesting a possible drop in the saving rates in future.

Those facing expectations of higher future income growth rates might be expected to have lower rates of saving.[32] It is certainly true that real income growth rates (both GDP and household disposable incomes) were higher for the early cohort and lower for the later (higher saving) cohort.

Increasingly in New Zealand over the last century the state has assumed the role of saving for the household through the provision of subsidised (or free) education and health services, social insurance (sickness and unemployment benefits), family support (family allowances, capitalisation of family benefits, support to solo mothers) and a universal system of retirement income support which is neither means nor asset tested.[33] It is to be expected that the incentive for voluntary saving by households would be reduced in the presence of these programmes; further, the more generous the programmes the greater the disincentive effect[34]. As a result, we would expect that the lower savers faced more and the higher savers faced less generous state subsidies in health, education and welfare.

Much work remains to develop consistent long-term series for health and education benefits, making any inferences about changes in these policies too speculative at this stage. As an increasing share of the costs of tertiary education has been shifted from the state to individuals over the 1990s, it is to be expected that younger cohorts would have more incentive to make provision for their children’s educational costs.[35] This tendency would be reinforced by the marked increase in tertiary participation rates.

Arguably one of the most significant elements of public policy that influence saving behaviour is that relating to the provision of public pensions. Between 1970 and 1979, the payment to a married couple under the New Zealand Superannuation Scheme rose by over 40 percent in real terms (see Figure 8). This rise corresponded to the peak saving years of the very cohorts that display lower lifetime savings. During the 1970s when the early cohort was at its peak saving period the real weekly pension for a married person rose from $135 to over $200 (in constant June 2000 terms). Over the next decade this fell and reached a low of $163 in September 1996. By June 2000 it had recovered to $174. In other words the early cohort faced the prospects of very substantial real increases in the state pension, a position that was not subsequently sustained. This is consistent with the early cohort making less retirement provision than the later cohort who face lower real pensions and greater uncertainty about their viability.

Another way to assess the real value of the universal superannuation is to compute the pension-to-wage ratio. This was low to medium for the reference group, was markedly higher for the early cohorts (typically over 80 percent for a decade from the late 1970s), and lower for the late cohorts (expected to average 65 percent)[36]. In addition, between 1985-86 and 1997-98 an income related surcharge on superannuation was in place, affecting up to one third of all recipients. Not only did this reduce the value of the pension to current retirees but it would have created uncertainty about the level of future payments and hence encouraged later cohorts to place less reliance on its eventual receipt.

National survey results from September 1999 indicate that 75 percent of those questioned agreed with the statement that they would have to rely entirely on their own efforts to ensure a comfortable retirement. Almost 90 percent felt that there would be significant changes to the present arrangements, and 88 percent felt that any future payments would be less than today’s levels. In fact 52 percent felt there would be no superannuation within 20 to 30 years or that it would be only available to the indigent on a social safety net basis UMR, 1999).

New Zealand has created a public superannuation fund to meet part of the future government liabilities under the universal pension scheme. Taxes are now higher than are needed to meet current obligations, and will be commensurately lower in the future when drawings from the fund meet a part of the costs of the public pension; in fact the scheme is best viewed as a tax-smoothing system[37]. If the costs of higher taxes today are offset in present value by higher expected pensions in the future (or greater certainty that the present levels will be maintained), then lifetime wealth of an individual would be unaltered implying no change in their consumption today. As a consequence while national saving may be unchanged by the introduction of such a scheme, personal saving rates could well decline reflecting the effect of tax and transfer programmes rather than any underlying change in household consumption and saving behaviour. Gokhale, Kotlikoff and Sabelhaus (1996) conclude that the social security scheme which transferred resources from the current young and future generations to the current older ones is a principal factor in explaining the post-war decline in US saving rates.