The corporate income tax is under pressure. A host of factors have contributed to economic, normative and political discussions on the need for corporate income taxation and its role in the architecture of national and international tax systems.
Among the most persistent calls today is the lowering or scrapping of corporate income tax (CIT) altogether. And that’s not surprising, given that statutory CIT rates around the world have been in free fall for decades:
Although arguments to support this call are in the media weekly, two recent posts brought them very clearly to my attention. First, Diego Zuluaga, of The Institute of Economic Affairs (a UK free-market think tank), wrote this post in City A.M., the City of London newspaper, arguing that the CIT should be abolished because it discourages investment and incentivises tax avoidance. Second, Scott A. Hodge of TaxFoundation (a US tax policy research organisation) re-iterated arguments that the CIT is harmful to economic growth.
These and other positions are heard often in the debate on corporate income taxation, so it’s fair to examine the arguments and the science behind, which is the aim of this post. I want to break down some of the most prevalent discussions:
Corporate income tax, investment and wages
One central discussion around the CIT concerns its impact on investment and wages. A particularly popular argument comes straight from the neoclassical econ textbook, and is well-established: The CIT discourages investments, harming wages. Capital and investments will flow where the costs are lowest and the returns are highest, relatively speaking. Thus, the higher the CIT, the lower the returns, and thus the less desirable investment climate. If your country has a high corporate tax burden, investors will find it less attractive to invest there. In turn, fewer investments lead to a greater labour to capital share (K/L), meaning workers have relatively less machinery, technology, etc. at their disposal, making them less productive. And since each production factor is rewarded according to their marginal productivity, labour’s wage compensation decreases as a result.
And it’s not just theory. A well-known empirical meta study on these behavioural effects (tax elasticities) tells us that in the average situation, a 1%-point decrease in the effective marginal corporate tax rate will increase the corporate tax base by 0.4% due to intensive investment decisions (e.g. increasing production), and a 1%-point decrease in the effective average corporate tax rate will increase the tax base by 0,65% due to extensive investment decisions (e.g. buying a new plant).
On the other hand, there are a few criticisms of the argument. First, tax is only one side of the fiscal coin – expenditure is the other. Wherever there’s a tax, there’s a corresponding expenditure (or saving for future expenditure). So if the corporate income tax is raised, current or future government expenditure increases consequently, or the tax burden is shifted somewhere else, and vice-versa if the corporate income tax is lowered. The actual or potential government expenditure and the structure of the tax burden could raise the investment incentive by increasing returns or lowering risk, for instance via spending on education, infrastructure, technology – and thus if it is de-financed, the effect might be a net deterioration in the investment climate. But that would require detailed contextual analysis. Some studies confirm that CIT rates are relatively unimportant factors in business investment decisions – trumped by market size, human capital, infrastructure, etc. Some studies on the investment effect of CIT deal with the ‘fiscal coin’ issue by investigating long time series (usually 10-20 years) across many countries. Yet, even 20-year timelines may not necessarily be sufficient to reflect public spending or tax burden shift impacts, as it may take several decades for investment benefits, economic growth harms (e.g. from increase inequality) or structural tax shifts to kick in. Still, in the current research, the results are often, but not always, a negative semi-elasticity of investment to corporate income tax:
Second, even if investment is harmed by the CIT – what kind of investment is it? If it is real, productive investment that is discouraged, it is obviously harmful. If it is virtual ‘on paper’ investment is that is discouraged, it does not hurt the real economy. The latter could be profit shifting, round-tripping investment, or white-washing of black/grey investments. Evidence on the tax elasticity split of these two is naturally extremely sparse, given the difficult task of assessing the ‘real’ or ‘virtual’ nature of investments. Some studies have found that corporate tax incentives are poor at attracting real investments, whilst others argue that there is no relationship at all between low CIT rates and increased foreign investment. Well-known tax economist Kimberly Clausing has also argued that virtual investment re-allocations are far more tax sensitive than real economic activities.
The CIT and economic growth
A related argument, that CIT harms economic growth, is probably the most prevalent and well-analysed position against corporate income taxation. The intensely cited OECD ‘Tax and Economic Growth‘ piece from 2008, also referred to by Hodge, found corporate taxation to be the most harmful tax to economic growth. The conclusion was partly based on a statistical analysis of OECD firm investment levels and corporate taxes 1996-2004. In another study, the OECD compared tax levels with growth rates for 21 countries over 35 years and found that “corporate income taxes appear to have the most negative effect on GDP per capita”. Theoretically, the backdrop is familiar: the CIT causes behavioural distortions (fewer investments, more tax avoidance, etc.), which are harmful to overall output.
As noted above, there may be reasonable questions of the universal “CIT is harmful” thesis in relation to the timeline and the strength of conclusion. Are economic and statistical models sufficiently robust and sensitive to long-term structural evolutions caused by CIT changes, such as the impact of changed public spending levels, inequality (which may also harm growth), to provide clear evidence that the CIT is generally, or even in the average situation, harmful to economic growth? Much of the underlying economic science, based on statistical modelling and econometrics, is taken as evidence of universal laws or generalisable cause-effect relationships, but there are fair questions whether causality is really evidenced. Rather than asking or answering whether the CIT is generally harmful, you might say the more apt question concerns the specific circumstances under which a certain CIT raise or reduction might contribute to economic growth. We know, for instance, that because the CIT acts as a backstop for personal income taxation, at a certain low point, the CIT will facilitate significant tax avoidance as individuals disguise personal income as corporate income. In recent years, even the OECD itself has distanced its 2008 analysis with a more balanced view. Still, we must recognise that a sizable economic literature today holds that the CIT is, generally, harmful to economic growth.
Second, again parallel to above, even if the CIT harms economic growth – what kind of growth is it? If national output growth is based on proceeds from paper shifted assets, the real economy impact might be negligible. If it’s based on actual economic activity shifts, then the story is different. There are reasons to believe real economic activity is less sensitive to tax than virtual capital re-allocation, as noted above. And if that is the case, then lowering the CIT rate merely facilitates an international negative-sum tax competition game, based on virtual capital reallocation, harmful to overall world welfare.
Moreover, it is well-established that CIT rate cuts harm national inequality, as tax burdens are shifted from capital onto labour and consumption, which is usually less progressive, and as public spending (where less wealthy citizens are often the target) is cut. Thus, economic growth from CIT cuts might come at the cost of inequality, less social cohesion and democratic harm.
The question of incidence
A third common discussion on the corporate income tax is its incidence. Who bears the burden of the corporate income tax? Some might find that a strange question, but in economic theory, the corporation is merely an intermediate vehicle. In the end, all flows move through the corporation and end up elsewhere, as dividends to shareholders, as products and services to customers, as wages to employees, etc. When speaking of the CIT incidence, the debate is usually whether it is labour (in the form of lower wages) or the shareholders (in the form of lower dividends or retained earnings) that bear the brunt of the burden. The answer to that question is instrumental to both economic, normative and political assessments of the corporate income tax. If, for instance, it falls primarily on labour, a CIT reduction could well be quite progressive; if it falls mainly on capital (shareholders), it is more likely to be regressive (depending on who those shareholders are).
Hodge’s post and the WSJ essay cited both argue that “workers bear the true economic burden of the corporate income tax through reduced wages.” This is a popular sentiment, and is usually employed in arguments for reducing the CIT. And certainly, scientific support is abundantly available. Theoretically, the incidence depends on the elasticities (yes, that word again) and mobilities of the production factors. In a modern open economy, capital is said to be more able than labour to ‘escape’ the tax incidence, by shifting investments abroad, thus reducing the marginal productivity of labour and incidentally leaving them to bear the burden. Empirically, you can pick any of a long list of studies that find labour’s share of the burden in an open economy to be significant (e.g. 1, 2, 3)
Both the theory and the empirics here are simplified, of course. Corporate tax incidence is highly complex and context-dependent, with differences in the short-, medium, and long-term, but there is no doubt that a large literature supports the notion that labour bears the main burden of corporate taxation.
On the other hand, you can find an equally large literature to support the notion that capital bears the main burden of corporate taxation. Yes, that is correct. Again, there are variations in the way incidence is studied and theorised, but there is a similarly long list of studies available (e.g. 1, 2, 3). In short, the argument here is that capital is not always more mobile and flexible than labour, partly because the ability of capital to ‘escape’ may have been decreased over time due to more sophisticated regulation.
In other words, there is a fundamental lack of consensus in the economics literature on the incidence of corporate taxation. There is no reasonable argument that the burden “clearly falls on labour”, nor that it “clearly falls on capital”. The only likely agreement is that there is some split, which is difficult to define and is context-dependent.
Corporate taxation and profit shifting
A final discussion, invoked by Zuluaga amongst others, is the extent to which the corporate tax burden incentivises tax avoidance. The argument that it does so significantly is, again, based in economic theory, as well as empirics. The CIT lowers the rate of return, and thus capital will flow elsewhere – a part of which will be virtual financial re-allocation (profit shifting). As De Mooij and Ederveen’s table 4 shows above, they empirically estimate that a 1% increase in the statutory CIT rate will decrease the corporate tax base by 1.2% due to profit shifting. The 1.2% estimate is based on a number of longitudinal and cross-sectional studies, mainly from the OECD countries or North America in the 1980s and 1990s. Whether one can generalise based on such evidence is, as discussed above, probably questionable. Still, there are many studies with similar findings. And the effect of profit shifting is obviously less corporate tax revenue, and thus public spending cuts or the shift of the tax burden to other areas.
On the other hand, and as noted elsewhere, over the past decades regulatory tightening of corporate profit shifting may have contributed to a notable reduction in the profit shifting elasticity. Subsequent international reforms have contributed to the lessened ‘virtual mobility of corporate profits. The OECD certainly seem to think so, at least. Furthermore, even if the elasticity is still substantial, ongoing reform attempts will likely squeeze the slipperiness more and more, so that in the future we might expect the effect on tax avoidance of corporate income taxation to become less and less important. So while lowering or abolishing the CIT might be advocated as one way to reduce tax avoidance, there are clearly regulatory alternatives, which may be more prudent.
To sum up: the impact of corporate income taxation on investments, wages, growth and profit shifting are hot, important and contested, and for each debate there are clear arguments for and against. Again and again, we see these arguments being used selectively by different discussants in debates. But it is important to maintain some balance. Ultimate claims on the “truth” about corporate income tax and the science behind it are simply untrue – there is no one answer. The state of our knowledge on these topics are not so that we can conclude decisively on cause and effect, unfortunately. Still, science can bring us far, and we should employ the best research available. But we should employ it with caution. I think we would all do well to keep that in mind.
September 2017 update: To all of the above, I think it is worth adding a recent discussion in economics on the nature of the contemporary economy, in particular economic concentration, and its impact on the investment, wages, growth and incidence. A number of economic researchers and commentators (1, 2, 3) have noted recently that the contemporary economy is marked by significant and increasing economic concentration and monopoly power. This has important implications for economic theory and reality on corporate taxation. In particular, monopoly power indicates that capital forces are able to extract significant rents from the economy. In turn, this might indicate that the incidence of corporate taxation falls more heavily on capital (shareholders), and less heavily on labour, which may also require a re-thinking of the economic effects on wages, investment and wages.