2 Overview of demographic trends
2.1 The sample and data
The paper compares only across countries in the OECD. Most policy discussion concerns New Zealand’s position in the OECD income distribution rather than the world income distribution. The paper’s approach of holding institutions, income levels, culture, and other determinants of long-run economic performance constant is also more sensible when comparison is limited to the OECD than it would be for a larger, more heterogeneous group.
The paper uses data for all 30 OECD countries. Results are shown for New Zealand, Australia, Japan, and the three groups defined as follows:
| Group | Countries included |
|---|---|
| Europe | Austria, Belgium, Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Luxembourg, Netherlands, Norway, Poland, Portugal, Slovakia, Spain, Sweden, Switzerland, United Kingdom |
| Recently developed countries (RDCs) | Korea, Mexico, Turkey |
| North America | Canada, United States |
Australia is treated separately because the Australian and New Zealand labour markets are closely integrated, and because Australia is the principal country with which New Zealand compares itself. The three groups have been assembled on the basis of geography and economic similarity. Japan forms its own group, as it does not fit into any of the others. The groups would be slightly different if they had been assembled on the basis of demographic similarity. France, for instance, would be included with the United States, and Canada would be included with the European countries, on the basis of current fertility rates (France, like the United States, has significantly higher fertility than Canada). The contrasts between the demographic prospects of different groups would be more marked if current demographic conditions, rather than geography or current economic conditions, had been used to assign countries to groups.
The paper uses “the OECD” as shorthand for “countries that are currently members of the OECD”. This leads to a few anachronisms, such as references to the OECD population in 1950, although the OECD was not established until 1961. The justification is convenience.
The demographic estimates and projections are obtained from the United Nations Population Division’s online database World Population Prospects: The 2002 Revision.[2] The UN estimates and projections are the only ones available that are sufficiently detailed, consistent, and comprehensive to enable the cross-country comparisons carried out in this paper.
2.2 Uncertainty about future demographic trends
The UN projections are not well suited, however, to properly representing uncertainty about future demographic trends. The UN represents uncertainty simply by calculating three different projection variants based on “high”, “median” and “low” fertility assumptions, with the median assumption being the UN’s preferred estimate. As demographers now emphasise, the traditional variant approach can give misleading results. The only internally consistent and readily interpretable way of representing uncertainty about future demographic trends is to construct fully stochastic population projections (Lutz, Vaupel, and Ahlburg 1998; Bryant 2003a).
Since stochastic population projections with the required level of detail are not available, this paper cannot represent uncertainty in a satisfactory way. The charts in this paper show only the UN’s median variant. Warning are, however, attached to results that demographic principles suggest are particularly uncertain. The paper also avoids spuriously precise statements of the form “Country X’s population will begin to decline in year 2037.”
Notes
- [2]http://esa.un.org/unpp/
