Conference – 2006 Discussion

Richard T Freeman

I have heard it said that the intersection of demography and economics is a rich area for theorists, but much less so for econometricians. That observation highlights the fact that, while demographic variables are among the most reliable in many respects, econometric analysis of their effects often has to be filtered through economic frameworks where longer-run stability can be questionable. Given enough time, seemingly solid constants and supposedly stable relationships may be on unsteady ground. Some so-called ‘fundamentals’ even turn out to be partially endogenous.

Against that background, Philip Davis has done a nice job of surmounting many of the hazards of econometric work in this area by effective use of a very broad cross-country panel. It is reassuring that the findings on the whole tend to confirm theory (specifically the life-cycle hypothesis) across a wide range of financial structures and cyclical circumstances. The paper has a lot of good ideas, raises some important issues and adds to our understanding of some of the potential impacts of demographic change. The paper also raises some intriguing questions about how the growth of pension systems might contribute generally to financial development and, in turn, to real economic growth. My comments have to do mainly with how we might use the findings to draw out some broader implications – especially with regard to how saving may evolve over time and affect financial markets as global ageing advances.

Structural change in financial markets

My first point has to do with the interactions of ageing populations and structural change in financial markets. A noteworthy trend observed in the 1980s (and emphasised yesterday) was the widespread decline in household saving rates in major industrial countries that has been attributed, among other factors, to improvements in financial technology. At that time, more efficient structures arose for smoothing consumption and achieving life-cycle objectives. Opportunities for saving expanded. Potentially, such changes can operate in competing directions, but on balance households lowered saving rates – in fact, they lowered saving rates significantly in some countries where rates had been thought to be permanently entrenched at high levels.

This raises the question: should we expect something equally dramatic as engineering of financial instruments and frameworks for retirement-related needs continue to evolve and spread globally? Ongoing structural innovations could make the process of saving for retirement more efficient and, thereby, lower saving rates. But steps being taken in many countries to shift the burden of risk to individuals (moving them off defined benefit plans, for example) could work in the opposite direction. The net effect of these and other foreseeable structural changes is hard to pin down, and there may be many such changes in store. Of course, these changes not only have immediate impacts on saving and spending behaviour when they occur, they also change the entire incentive structure faced by the population as it ages over a period of indefinite duration. Accordingly, the responsiveness of saving to ageing populations, and the impact on financial markets, could be quite different from what has been revealed by the sample.

Home and property ownership

The papers presented so far at this workshop have been heavily focused on the provision of income for retirement via pensions and other essentially financial arrangements. But another key objective for household saving – and one we know to be importantly affected by demographic developments – is home and property ownership, which is the basis for my second point. In the United States, in interpreting movements in spending recently, we have seen the importance of households' capacity to extract income from housing assets as house prices rose to high levels. My reason for mentioning this is that home ownership is an alternative avenue for building up assets, and withdrawals from home equity could operate to smooth consumption over time for older cohorts as well. This is likely to be occurring to some extent already, through downsizing in housing and other such steps; the smoothing would be even more apparent if reverse mortgages and other similar instruments were more prevalent.

In cross-country data, a noticeable feature is how widely the share of home ownership in household wealth varies. That share is high here in Australia, for instance, and surprisingly low in countries like Switzerland. Such differences could arise in a number of ways, but one wonders about the extent to which home ownership may interact with decisions related to life-cycle saving. A broad comment that I might offer is: in this research, and in general in our discussions, impacts on saving and, thus, on financial markets through housing deserve more attention – especially if this is a sensitive margin across which individuals shift their personal saving portfolio.

Globalisation

The third area that I want to spotlight is globalisation. From our discussions, a basic point (in effect, nearly an axiom) seems to be: impacts of national demographic developments on local markets will be greater to the extent that an economy conforms to the closed model. As an economy becomes more globalised and open, the impacts of local ‘shocks’ – including those related to ageing – get dispersed into global markets and are thereby buffered; by the same token, external demographic shocks start to matter more locally.

If we ask what we can expect going forward, it seems safe to assume that globalisation will continue. But globalisation as it relates to ageing could mean changes across a variety of economic measures – and what particular aspects of globalisation should be stressed most are debatable. In fact, the degree to which the basic axiom holds up may vary, depending on our focus. Three aspects of globalisation or openness seem pertinent. At perhaps the level of greatest generality, openness might mean the extent to which a country is able to run an imbalance in its current account, with the gap financed by capital inflows or outflows. Analysis at this level, of course, links directly to broad, national measures of saving and investment, and in turn to how ageing might affect cross-country patterns of external imbalance. The paper by Philip includes a control variable that measures openness at this level of aggregation, in effect, as the share of trade in GDP. But openness at this level of generality may not capture well the aspects of globalisation that are most relevant to detecting how ageing populations might affect financial markets.

A somewhat narrower, more focused notion of openness refers to the degree to which flows in specific (but still fairly broad) compartments of financial markets can move across borders freely. To the extent that some financial markets (say, fixed-income securities and some stocks) are more subject to global arbitrage than others, local and global impacts of ageing populations will be felt differently across market sectors. And, of course, the margins of substitution that underlay such relationships could well shift over time.[1]

Finally, perhaps most narrowly, globalisation could refer to the extent to which particular institutions involved in intermediating intergenerational flows become more globalised themselves – either by wider portfolio diversification or by cross-border operations. Changes in the rules or practices governing these aspects of globalisation also could importantly affect how demographic developments get transmitted into local financial markets. Many pension portfolios already are widely diversified internationally, but the degree of diversification is uneven. In the United States, for example, there is a considerable range of variation in foreign asset shares. Rough calculations suggest that there still is scope for further decline in ‘pension-sector home-bias’ in many cases. I suspect that such estimates for other countries might lead to a similar conclusion.

In general, a trend toward wider global diversification likely would be a positive development, as well-managed, globally diversified portfolios offer potential for improved risk-return trade-offs. But there are some hazards as well. Lately, concerns have been expressed about pension funds becoming prominent investors in financial markets with high volatility, especially during the recent low-interest-rate period. We generally assume that pension providers are not high-risk investors, have effective governance structures in place and perform sufficient due diligence with respect to risk management. But these are assumptions for which it is worth reserving final judgment, and these concerns are a reminder of the need to have adequate supervisory safeguards in place as financial structures and strategies evolve.

Footnote

A theme of the recent compendium produced by the members of the G-20 on ‘Economic Reform in the Era of Globalization’ was the positive effects of openness on domestic financial development more generally (G-20 2003). Arguably, greater openness is likely to contribute to a more efficient, better-managed financial environment with improved options for meeting age-related investment demands. [1]

Reference

G-20 (2003), Economic Reform in the Era of Globalization – 16 Country Cases, October. Available at <http://www.g20.org/Public/Publications/Pdf/economic_reform_in_this_era_of_globalisation_16_country_cases.pdf>.