In the previous episode, we looked beneath the aggregate labor force participation rate and examined its age structure.
The basic idea was simple. Overall participation is not a single primitive number. It is formed from participation rates within different age groups, weighted by the population share of each group.
This raises a natural next question:
When participation changes over time, what is actually changing underneath?
Is the change driven by demographic structure? Or is it driven by changing participation within age groups?
For this episode, we use three age groups:
For each country and year, reconstructed participation is calculated as:
LFPR = Σ (population share × age-specific LFPR)
This reconstructed measure is a working-age participation rate for ages 15–64. It is not identical to the World Bank aggregate LFPR, which is based on the 15+ population. The purpose here is not to replace the aggregate measure, but to examine the internal structure of working-age participation.
Between 2015 and 2024, the change in reconstructed participation can be separated into three parts:
The word “behavioral” is used here in an accounting sense. It means that age-specific participation rates changed. It does not, by itself, identify the social or institutional reasons for that change.
| Country | Demographic Effect | Behavioral Effect | Interaction Effect | Total Effect |
|---|---|---|---|---|
| Canada | 0.172 | 1.547 | −0.015 | 1.705 |
| Germany | −0.654 | 2.960 | 0.248 | 2.555 |
| France | −0.807 | 3.784 | 0.124 | 3.101 |
| United Kingdom | −0.348 | 0.257 | 0.103 | 0.012 |
| Italy | −1.418 | 3.596 | 0.327 | 2.505 |
| Japan | −0.160 | 5.663 | 0.042 | 5.546 |
| United States | 0.081 | 2.282 | 0.005 | 2.368 |
The main result is clear. In every G7 country, the behavioral effect is positive. In most countries, it is also much larger than the demographic effect.
In Germany, France, Italy, Japan, and the United Kingdom, the demographic effect is negative. In these cases, changes in age structure alone would have reduced reconstructed participation. Participation nevertheless increased, or remained nearly unchanged, because age-specific participation rates moved upward.
This suggests that the main movement in participation lies in the behavioral effect.
The country-level decomposition tells us that behavioral effects dominate. But it does not yet tell us which age groups contributed most.
The next table decomposes the total effect by age group.
| Country | Age 15–24 | Age 25–54 | Age 55–64 |
|---|---|---|---|
| Canada | −0.384 | 1.973 | 0.116 |
| Germany | 0.839 | −2.755 | 4.471 |
| France | 1.398 | −1.015 | 2.718 |
| United Kingdom | −1.344 | −0.966 | 2.322 |
| Italy | −0.121 | −2.231 | 4.857 |
| Japan | 1.322 | 1.762 | 2.462 |
| United States | 0.043 | 2.015 | 0.310 |
Several patterns emerge.
Japan shows broad-based gains. All three age groups contribute positively, with the largest contribution coming from the 55–64 group.
Germany and Italy show large positive contributions from the 55–64 group, while the 25–54 group contributes negatively. Their aggregate increases are therefore strongly older-worker driven.
The United States shows a different structure. Most of the increase comes from the 25–54 group.
The United Kingdom is especially instructive. The total change is almost zero, but this does not mean that nothing happened. Negative contributions from the 15–24 and 25–54 groups were almost exactly offset by a positive contribution from the 55–64 group.
In other words, aggregate stability can conceal large movements beneath the surface.
Changes in participation are often discussed in connection with demographic aging. But the decomposition here shows that demographic structure alone does not explain participation change across the G7.
Between 2015 and 2024, changes in age-specific participation rates were the dominant source of change in reconstructed working-age participation.
At the same time, countries followed different paths. Japan’s increase was broad-based. Germany and Italy were strongly older-worker driven. The United States was more prime-age driven. The United Kingdom showed offsetting movements across age groups.
The aggregate participation rate is therefore useful, but incomplete. To understand participation dynamics, we need to look beneath the aggregate and examine how different age groups contribute to change.
This episode has described what changed. It has not explained why these age-group patterns differ across countries.
Why do some countries rely heavily on older-worker participation, while others show broader or prime-age-centered changes?
That question belongs to the next discussion.