Abstract:
Per-capita national income remains a primary indicator of wellbeing in an international comparison. Accordingly, the World Bank has popularised the concept of constant-dollar income of a given base year (say 2015), which converts income denominated in constant local currency units (LCU) to a common base year and uses the market exchange rate of that year to derive the '2015 constant-USD' income which is directly comparable across nations. The issue we address here centres on the evolving base year; with each change, a new constant USD income series is born embodying the base-year market exchange rate, which tends to be volatile. Predictably, the prior ranking of income of any two countries may well get reversed. To be concrete, as per 2010-base, Japan's per capita GDP appeared to have peaked in 1991 at 10.4 percentage points above the US level, but by the constant 2015 metric, the 1991 Japanese income was only about 3/4th. Was Japan actually richer than the US in 1991? In an attempt to answer this, we propose replacing the market rate by a smoothed exchange rate. We then compare the smoothed exchange rates as well as the resulting constant-dollar income series with the PPP exchange rate and PPP income. We show that some of the smoothed rates yield exchange rates that are at least as stable as the PPP, measure the exchange rate movement at par with PPP, and the predicted income levels appear mutually consistent with the PPP income.
