Benchmarking is an essential component of any performance assessment. While the world of high finance is hardly the place to draw positive lessons these days, the rewards to portfolio managers give a simple and clear example. Managers are typically assigned a particular asset class, and are rewarded based on how well their investments perform relative to the overall market return for that class. Although this compensation scheme has its problems, it provides clear benchmarks rather than assuming that all managers could make the same returns whether they are betting on risky or safe assets.

After the financial crisis of 1997–98 and the democratic transition in 1999, Indonesia did not return to the rapid rates of economic growth that had prevailed during the Suharto era. Growth in GDP per capita, which came in at 5.9 per cent per year from 1987 to 1997, fell to just 3.7 per cent in 1999–2008. Hence growth was slower by roughly 2.2 per cent. Is that slower growth a ‘disappointment’? Are the slower growth rates an indictment of the policies or economic management of Indonesia's democratic governments? I argue that past growth is not an appropriate benchmark for assessing the country's performance between 1999 and 2008 because it ignores two relevant facts about economic growth.

First, the literature has documented powerful regression to the mean in growth rates. Economies that have grown fast are expected to slow down. If one benchmarks Indonesia's expected rate of growth based on cross-country estimates of the typical magnitude of regression to the mean, this factor alone places its growth performance right at the benchmark.

Second, it is plausible that a major political transition would affect a country's rate of economic growth through a variety of channels. To test this, I identify all national episodes of rapid democratisation and examine those countries’ growth rates during the 10 years before and after the political transition. For countries that went into the transition with above-average growth rates, there appears to be a ‘democratic transition’ effect that slows post-transition growth by about two percentage points per capita per annum. Again, this effect alone would account fully for Indonesia's slower growth.