Since the International Comparison Program only produces PPP conversion factors for the benchmark year (2011 in the most recent release), we extrapolate backwards and forwards to create a time series from 1990 to the latest year available. The method we use is to apply the difference between the rate of inflation observed in the country over each period from 2011 compared with inflation in the United States over the same period to the benchmark PPP estimates. Extrapolation for the GDP conversion factor uses the change in the GDP implicit deflator. Extrapolation for private consumption uses the change in the consumer price index.
For example, let’s say Brazil’s CPI in 2014 divided by its CPI in 2011 is 1.19, which represents 19.0% inflation over the period; and the ratio for the US over the same period is 1.052, or 5.2% inflation. We can take Brazil’s private consumption PPP conversion factor in 2011, which was 1.66, and multiply it by 1.19/1.052 to get a 2014 conversion factor estimate of 1.88.
For OECD and Eurostat countries, we take OECD/Eurostat's extrapolated estimates, which are based on a rolling update method with more up-to-date survey information. We can also extend these data historically back to 1990, if necessary and where possible, using our regular extrapolation method explained above but measuring the comparable change from the last reported observation rather than 2011.