Sarah Daway: The introduction is still not well-structured/well-motivated and has to be reorganized. It is still unclear why introducing a RST would be more advantageous to China. Is it in terms of ultimately lower administrative costs (after the transitional phase), its anti-consumption bias, something else? This should be more clearly spelled out.

Related to the previous comment: The paper should explain more clearly why raising the RST and reducing the VAT results in more capital accumulation.

The paper mentions in several places that VAT and RST are equivalent—in what sense, though? VAT and RST can be made equivalent only in terms of the revenues they raise. Otherwise, the RST is essentially a tax on consumption and VAT is a tax on (value-added) production. It is in this sense that raising the RST, which makes consumption relatively more expensive to saving and, ultimately, accumulating capital, would result in higher capital accumulation.

An interesting experiment (arguably, a cleaner experiment) would be to replace the VAT with a RST that yields the same tax revenues as the VAT and see how the transitional, steady-state, and welfare effects might change.

On a side note, how would the introduction of a RST, which favors capital accumulation, interfere with China's bid for a more consumption-driven growth? This appears to be one valid argument against a RST in China.

In Figure 6, it appears that some current generations—those who entered the workforce 25 or 24 years to one year before the policy change—also benefit from the reform. This should be explained in the text.

In the paper, λ is still in terms of consumption and leisure when the utility function does not include the labor–leisure choice. Labor is assumed to be inelastically supplied to the labor market.

In the earlier version of the paper, RST administrative costs (percent of GDP) were included, and provided a convincing deterrent for the adoption of an RST over the VAT. Why was this removed from the model?

In the previous referee report, we made the point about the incongruity of modeling China as a closed economy. Although we understand that this would involve a drastic change in the current framework, we think that the authors should at least justify the use of a closed-economy overlapping generations setup and how the results might change if the economy were to be opened up to trade.