Elsevier

Journal of Monetary Economics

Volume 81, August 2016, Pages 21-24
Journal of Monetary Economics

Discussion
Comment on: “Income Inequality and Asset Prices under Redistributive Taxation” by Lˇuboš Pástor and Pietro Veronesi

https://doi.org/10.1016/j.jmoneco.2016.06.005Get rights and content

Highlights

  • Selection effects: entrepreneurs are more skilled and less risk averse than pensioners.

  • Taxes affect choice of becoming an entrepreneur or a pensioner.

  • Higher taxes reduce aggregate output but increase productivity.

  • Empirically negative relationship between entrepreneurial activity and tax burden.

  • Possible extensions: add workers or social planner that sets optimal tax schedule.

Introduction

The question of how an optimal tax system should be designed has spawned an enormous literature and is still subject to heated debates (see, e.g., Mankiw et al., 2009 or Piketty and Saez, 2013 for extensive overviews). Models to tackle this question usually posit that the chosen tax system should maximize a social welfare function subject to a set of constraints while taking into account the individuals׳ responses to taxes and transfer payments. In standard models this results in a tradeoff between equality and efficiency: on the one hand, taxes and transfers lead to resources being distributed more equally and, thus, to an increase in social welfare. On the other hand, redistributive taxation may negatively affect the incentives of individuals to exert effort and create income in the first place, ultimately resulting in a reduction in efficiency.

While Pástor and Veronesi (2016) do not tackle the question of optimal taxation, it is still instructive to relate their setup to some of the basic insights in the existing literature. Pástor and Veronesi focus on the channels through which business ownership and financial markets can affect inequality through the lens of a simple general equilibrium model with heterogenous agents. Taxes are exogenously given and their sole purpose is redistribution. Ultimately though, a better understanding of the interaction between taxation, entrepreneurial activity and financial markets may be useful for tax policy making.

The research field of optimal income taxation originated with Mirrlees (1971) who approaches the question by recognizing that governments who want to redistribute from high to low skill types can (at best) observe wages but not the heterogenous skills. As a result, taxes and transfers depend on the observed wages instead of unobserved skills. The tax system must then be designed such that high skill types are provided with enough incentives to still exert a high level of effort and not mimic low skill types.

While the Mirrlees model operates on the intensive margin, Diamond (1980) considers a tax system that operates on the extensive margin, i.e., where agents decide whether or not to participate in the labor market in the first place. He shows that a subsidy for work (or a negative marginal tax rate) may be optimal in certain cases.

What both approaches have in common is the underlying premise of a benevolent social planner. Mankiw (2013) challenges the utilitarian approach and makes a case for an alternative approach based on the principle that people should get what they deserve.1 Such a shift in the political philosophy also results in a more nuanced view of inequality, i.e., it makes a difference whether the high earners are simply more productive or whether they are able to exploit the system by extracting rents without producing anything of value.

As a thought experiment, Mankiw (2013) starts with a utopian society where everybody earns the same until an entrepreneur has an idea for, and launches a new product. If the product turns out to be something that everybody wants to buy (Mankiw uses the iPod as an example) then the entrepreneur becomes very rich without making the rest of the society worse off. The question then is whether public policy should be altered as a reaction to the fact that inequality has just risen, i.e., whether the additional profits of the entrepreneur should be heavily taxed and redistributed. Mankiw notes that this cannot be tackled by relying solely on economic reasoning as the question is fundamentally of a philosophical nature.

In their paper, Pástor and Veronesi (2016) study the choice of agents to become either an entrepreneur (and be subject to redistributive taxation) or a pensioner (and be completely unproductive). The level of taxation then has an effect on the decision of agents to become either entrepreneurs or pensioners. The effects of different (flat) tax rates are analysed via comparative statics.

Section snippets

Financial markets, business ownership, and inequality

The model of Pástor and Veronesi features heterogenous agents that differ along two dimensions: skill and risk aversion. They optimally choose to become either “entrepreneurs” or “pensioners”. Entrepreneurs take risks and invest their endowed capital. Their skill level determines their productivity and, ultimately, the return to their investment which is taxed. Tax revenues are equally redistributed among the pensioners who choose not to invest and, thus, lose their endowed capital.

The model is

Taxes and entrepreneurial choice

While the model of Pástor and Veronesi provides interesting implications for inequality, productivity and asset prices, one key implication that remains underexplored in the data concerns the relationship between taxes and entrepreneurial choice.

In the extant literature, the choice to become an entrepreneur is often either based on skill (as in Lucas, 1978) or risk aversion (as in Kihlstrom and Laffont, 1979).3

Conclusion

Pástor and Veronesi have written an elegant paper that spans several strands of the literature. In doing so they have managed to frame the discussion about inequality and redistributive taxation in a way that incorporates interesting implications for financial markets and asset prices. Their paper also provides a mechanism how taxation can influence the agents’ choice to become entrepreneurs. Given that the model setup is very simple, possible extensions abound and the paper is likely to spur

Acknowledgments

I thank Lorenzo Bretscher, Anca Dimitriu, Jack Favilukis, Juanita Gonzalez-Uribe and Christian Julliard for comments and discussions.

First page preview

First page preview
Click to open first page preview

References (19)

There are more references available in the full text version of this article.

Cited by (0)

View full text