Most economists agree on the broad criteria that are important for evaluating tax policy. Good tax policy should be efficient, equitable and simple (or, perhaps, it is better to say “administrable”). The task of describing the overall desirable structure of the tax system is not easy, because tax policy can serve many different purposes. It is obviously there to simply collect revenue, but it also plays a key redistributive role, it is often used to incentivize particular types of behavior (e.g., R&D, addressing externalities, or implementing paternalistic objectives), and it may play a role in macroeconomic stabilization. In this short note, I will provide a brief overview of the redistributive role.

There are two broad ways of thinking about how tax policy might interact with redistribution. One is to focus on taxation alone and postulate that it should be progressive on its own. That’s an approach that’s familiar to public finance economists: the idea is to tax and transfer based on ability to pay, possibly as one integrated system, where the bottom of the distribution faces negative tax liability. It is also the line of thinking that naturally fits with direct interest in reducing inequality (rather than simply increasing welfare – a subtle distinction that assigns independent value to equality over direct individual/utilitarian well-being), because it motivates the focus on taxation of the rich, perhaps pushing beyond what otherwise might be considered efficient.

The alternative is to separate taxation and redistribution, with social assistance and insurance spending financed using tax revenue, but with less of a focus on distributional aspects of how revenue is collected.

Obviously, the real tax systems have elements of both, with income taxation adding progressivity to the tax side and the safety net financed out of many sources of revenue. As Figures 1 shows, the explicitly redistributional income tax and, to some extent, corporate tax amount for an important but far from dominant chunk of revenue in the OECD countries, with Spain close to the average and the United States relying on them more despite its smaller overall revenue take. In contrast, instruments that are on their own flat or mildly regressive, chiefly excise, sales/VAT and payroll taxes, account for the majority of the revenue for the average OECD country, with Spain again close to the mean, and the US deviating from it greatly. Finally, very progressive and grasping popular attention inheritance, estate and wealth taxes play a trivial role in terms of their revenue take.

Somewhat paradoxically, countries that do a lot to redistribute rely less (or, in proportion to the revenue, much less) on progressive tax instruments than the US, while financing their spending in part regressively. In particular, the VAT – that does not even exist in the U.S. –  collects over 9% of GDP in revenue in Denmark, New Zealand, Sweden and Finland.

As Figure 2, replicated from Blanchet et al (2022), shows this translates into a less progressive overall tax system in high income European countries than in the United States, reflecting the important role in Europe of somewhat regressive indirect taxation and effective income tax rates that are fairly high already low in the distribution (and overall flatter than in the US).

What I take from these observations is that successful redistributive systems involve (1) large tax take, that’s (2) used to spend very progressively, but (3) overall financed in a reasonably flat manner using taxes that hit the middle class with only moderate progressivity at the top.

I believe that this reflects the pragmatic choice driven by the fundamental characteristics of different sources of taxation. Each tax instrument has its problems, but the Value Added Tax and payroll taxes are collected by relying on businesses, predominantly large businesses, and tax bases that largely reflect transactions with redundant flows of information that lend themselves to effective information reporting: wages and business-to-business sales. This makes them efficient and administrable. Individual income taxation works well when it goes after labor earnings, but starts having practical problems when it extends to businesses, capital income, rents, and - even more so - wealth. Observing and taxing these additional sources of resources is important when one wants to redistribute through the tax system and tailor tax liability to the “ability to pay”, but this is precisely where practical issues related to timing and realization of income, its location, assignment to ultimate owners, valuation and liquidity start to bite, because in most cases extending the tax system in this direction requires going beyond arm's-length observable market transactions. Moderate progressivity of income taxation is one response to dealing with the real world consequences of problems with measuring, tracing and taxing income comprehensively.

Firms loom large in any tax system. A corporate tax without integration with individual income taxation is a rough attempt at progressivity (because equity owners tend to be wealthier), but it does so without closely interacting with income taxes paid by the ultimate owners. A potentially more progressive solution is to tax business income as individual income — what’s known as the “pass-through” approach in the United States (see Kopczuk and Zwick, 2020 for more discussion). That solution however is problematic when dealing with publicly traded firms or firms that have foreign or institutional owners, so that – in practice – profits of large corporations need to be taxed through entity level taxation in a way that’s only loosely tied to individual incomes.

A broad alternative to that is to implement progressive individual taxation of consumption rather than income, perhaps along the lines of Bradford’s (1986) famous X-tax proposal. Such an approach has some administrative advantages and efficiency benefits related to eliminating saving distortions, but has not yet been implemented anywhere at a large scale.

A different direction is to focus on alternative metrics of ability to pay such as wealth or transfers. Saez and Zucman (2019) advocate a progressive top-focused variant of such a tax. My view, expressed more extensively in Kopczuk (2019), is that such proposals have problems. Wealth taxation - especially when it is not just about the very top – reflects a mix of past consumption and saving decisions with corresponding inequities and distortions. Perhaps more importantly in practice, wealth taxes tend to struggle with observability, valuation and liquidity of assets, in particular in the case of the class that’s key for progressivity: private business. They also often run into political problems when expanding the base to owner-occupied housing. Practical approaches to dealing with these issues involve exempting or preferential treatment of hard to tax or unpopular components of the base that lead to a system that introduces cross-asset distortions, inequities and raises little revenue due to narrow base (see Alstadsæter et al, 2022, for a discussion and evidence in the Norwegian case). Finally, wealth tax is a cousin of capital income taxation that shifts focus toward principal and away from returns (and, especially, away from rents/extraordinary returns), raising additional efficiency and equity concerns.

Does that mean that wealth taxation should not be used? Economic theory does suggest that wealth tax is desirable when it is one-time, immediate, and unexpected, because it resembles lump-sum taxation. It is a hard task to credibly implement such a tax, but historically solutions resembling this idea have occasionally been used to finance wars or tax ill-gotten gains. Such situations do seem like the best case for it.

This is but a quick overview of main themes related to the role of tax policy in the design of a redistributive tax system. Countries that have large and generous safety nets collect their revenue in ways that are not extremely progressive, but rather correspond to fairly large and reasonably flat share of overall income effectively taxed throughout the distribution. I believe that this reflects the general administrative difficulty in implementing aggressive taxation based on ability to pay. Nevertheless, if the primary objective is safety net and well-being rather than just focus on metrics of inequality, those practical choices that combine large, but not extremely progressive tax take with very progressive spending are capable of delivering that objective.

References

Annette Alstadsæter, Marie Bjørneby, Wojciech Kopczuk, Simen Markussen and Knut Røed, "Saving effects of a real-life imperfectly implemented wealth tax: Evidence from Norwegian micro data", AEA Papers and Proceedings, 2022, 112, 63-67.

Blanchet, Thomas, Lucas Chancel, and Amory Gethin. 2022. "Why Is Europe More Equal than the United States?" American Economic Journal: Applied Economics, 14 (4): 480-518.

Bradford, David F. 1986. Untangling the Income Tax. Cambridge, MA: Harvard University Press.

Wojciech Kopczuk, "Comment on 'Progressive Wealth Taxation'", Brookings Papers on Economic Activity, Fall 2019.

Wojciech Kopczuk, "Reflections on Taxation in Support of Redistributive Policies", in Olivier Blanchard and Dani Rodrik (eds.), "Combating Inequality: Rethinking Policies to Reduce Inequality in Advanced Economies", MIT Press, 2021.

Wojciech Kopczuk and Eric Zwick, "Business Incomes at the Top”, Journal of Economic Perspectives, 34(4), 2020, 27-51

Emmanuel Saez and Gabriel Zucman, "Progressive Wealth Taxation", Brookings Papers on Economic Activity, Fall 2019.

Figure 1: Sources of revenue in Spain, the United States and OECD as a share of GDP

Source: OECD,  Revenue Statistics for 2019, retrieved from https://stats.oecd.org/Index.aspx?DataSetCode=REV

Figure 2: The Distribution of Taxes in Europe and the United States (Blanchet et al., 2022)

Source: Panel A of Figure 5 in Blanchet et al. (2022): “The Distribution of Taxes in Europe and the United States. Noncontributory taxes paid as a share of pretax income”