Reynolds-Smolenski is a measure of fiscal redistribution.

There are several items in the paper that could benefit from some additional discussion. First, as noted in the paper, the present work is rather closely linked to a longstanding literature on the effective progressivity (or redistribution) of fiscal policy. The use of measures based on the Gini coefficient dates back at least to Musgrave and Thin (1948), while the intuitive measure frequently used – the market Gini minus the post-fiscal Gini – is attributed to Reynolds and Smolensky (1977). There are many analyses that use this index to measure redistribution, the most recent of which is IMF (2014) as cited by the authors. It would be quite useful to better understand the benefits and costs of regressing the post-fiscal Gini on the market Gini (as the authors do) instead of examining their difference (as is more commonly done). The latter approach is indeed intuitive and as noted in IMF (2014) there is a significant body of analysis available – much of which parallels and augments the key results of this paper. The authors emphasize that using the difference in Ginis in regression equations (say, by regressing the Reynolds–Smolinski measure on various fiscal policies) will effectively restrict the coefficient on the market Gini to be 1. While this may be true, it is not entirely clear why the restriction represents a significant cost, given the added intuition from using a measure of fiscal redistribution and, indeed, given the authors own empirical findings of a coefficient quite close to 1. 5