This paper relies on a dynamic, intergenerational and interregional demographic life-cycle model. The model has three regions — the United States, the European Union and Japan — which exchange goods and capital. We carefully model the fiscal institutions of each region, including their pay-as-you-go government pension systems.⁵

“Our projections are based on optimistic assumptions about fertility rates and health care costs — the reality is likely to be far worse.”

In general, the assumptions behind the model are similar to the “intermediate” assumptions used in the United States by the Social Security Trustees and the Congressional Budget Office. However, in two respects we adopt assumptions that are extremely optimistic: (1) We assume that the long-term decline in fertility rates will reverse over time and eventually converge at the replacement rate (2.1); and (2) We assume that the growth of health care costs per beneficiary will slow and eventually grow no faster than the rate of growth of real per capita wages — even though they have grown 16 times faster over the past three years.

As a result of these very conservative modeling assumptions, the results reported here err on the side of optimism. Reality is likely to be far worse.