Investment and corporate taxes

Thanks to Stephen Gordon, who made a link to a new unpublished study (fourth draft, 2009), The effect of corporate taxes on investment and entrepreneurship, by Djankov, Ganser, McLiesh, Ramalho and Shleifer. Stephen claims this study settles the matter that Canada should not reverse corporate tax cuts made in recent years. That discussion was happening deep in the comments section of Toby’s recent post, so I figured I would start a new thread (and I have pasted a few of my observations from there over here).

My contention is that there is little impact on efficiency, growth or investment from corporate income tax cuts. There are theoretical arguments that corporate taxes have negative efficiency impacts but this is not settled in the least empirically. It is not that we should ramp up corporate taxes to 70% or anything, but there is no reason why recent corporate tax cuts – that have taken Canadian rates well below US ones – should not be repealed. And to the extent that there are efficiency impacts of corporate taxes they must be weighed against the beneficial impacts of any associated public spending. If spent in a pro-growth manner corporate income tax increases may have no efficiency losses whatsoever. But I still think within the range of taxes we have seen in recent years, efficiency impacts will be extremely small.

Most corporations in Canada have to be here to access the Canadian market, or resources they want to exploit. And there are many more other aspects to investment decisions than just CIT rates. Cheap electricity in BC or Quebec overwhelms most differences in CIT rates. These subtleties are almost never captured in empirical studies, so on balance Toby is right to be skeptical of them.

Finally, if you do want to cut CIT rates you must increase top marginal personal income tax rates or else you give windfall gains to the wealthy (I cite Jon Kesselman on this).

I did my own review of this literature a few years ago. Here’s a pertinent passage related to the corporate tax claims above. I still have yet to see any real economic evidence to the contrary:

Another group of studies often cited by advocates of smaller government comes from computer simulations that find that an extra dollar of government revenue actually costs the economy something like $1.38 or more in lost economic activity (see Dahlby 1994). It is important to note that these results are not derived from real-world data. They are quasi-empirical studies that start with a theoretical model where taxes impose large deadweight costs to the economy, then put some real numbers to the model to enable them to simulate what the cost of extra taxation is at the margin. They are “educated fiction” based on the virtual reality of computer models (Lindert 2004).”

So that is the back story. Stephen rebuts by citing an unpublished paper, which starts by listing a series of studies initiated by Dale Jorgenson. But it is precisely those models that are what Peter Lindert calls educated fiction. They fit numbers to a model that assumes corporate taxes have large impacts on efficiency, then guess what they find …

The unpublished study Stephen cites is interesting but deeply flawed. Interesting because it finds that corporate taxes are not significant for investment. But they are significant for FDI, which makes sense given that the study has 85 countries and is thus treating the corporate tax rate in Ghana as an equivalent data point to Canada. Looking more closely at the data, on a simple correlation in Figure 1, they point to a loose linear relationship, but the observations are basically a blob that suggests little. For FDI, Figure 2, the relationship is stronger but heavily influenced by developing country outliers, with most observations tightly clustered together.

Which explains why this study has some of the lowest r-squared values I have ever seen reported in an empircal study (like .03 to .12). In case you missed the econometrics class on r-squared, low values mean poor goodness of fit. So basically no reason to believe their estimates have any traction at all. There is a later one that gets up to an r-squared high of .39, around the level one might actually start to pay attention, but this is for a simple regression that only includes corporate taxes as the determining factor behind investment, and is only based on 16 observations, which makes it a very weak estimation.

There is more to it, but econometrically this study is a dud. Anyway, for Canada, the main source of FDI is the US. But here the authors find no significant relationship for US FDI abroad vis-a-vis foreign tax rates!

For this, I get chastized by Stephen: he is “very, very disappointed” in me for not having done my homework and knowing that this unpublished study was out there. But I’m disappointed in him for coming back with such a weak piece of evidence. The jury is not convinced.

So I go back to my original contention: As far as I have seen there is no conclusive econometric evidence that lower corporate tax rates lead to higher rates of investment, GDP growth or productivity growth — not one unpublished study but many credible ones that come to the same conclusion. There are some cross-country and historical comparisons that small open economies with high levels of GDP per capita and decent productivity growth (ie the Nordics) do tend to rely less on corporate income taxes and more on consumption taxes, so we should take those lessons about tax mix seriously — that you should not have corporate taxes that are way out of line.

So I’m not saying there is zero impact of corporate taxes, but you would be looking at fairly large changes before there was,  and that would vary greatly by sector and by country. Empirically, in the lit reviews I have seen, investment is mostly driven by demand side factors not supply side ones. To imply that lowering corporate rates will lead to increases in investment and thus faster growth is seriously misleading.

So if you are on the fence about increases taxes on, say, Canadian banks, back to the levels they paid a decade ago, you need not fear an economic collapse.


  • Marc, I think the link to your own lit review might be faulty. It links to a paper about the effect that government size has on the economy, not a paper about corporate tax rates.

    As for the working paper Stephen Gordon provided, at least it’s a starting point for econometric measurement. The ball is now in your court to provide an econometric analysis that shows corporate tax rates have no effect.

    You slam the paper for being unpublished, but it is pretty new, and is by authors at Harvard and the World Bank, so it has some credibility.

    I agree with your assessment of Figure 1, but Figure 2 looks pretty convincing to me.

    I do agree with your point that it seems strange to include analysis for developing countries such as Ghana and believe the findings are still applicable to Canada. It would have been nice to see if their results held regardless of country wealth.

  • Thanks, David. In part my post was a reaction to Stephen’s arrogant and condescending tone, which I felt was inappropriate given what evidence there actually is, and what he offered up by way of rebuttal.

    My paper was about taxes and economic performance, and I only spent a bit of time on corporate taxes. I think most econometric analyses that try to explain these big relationships ultimately degenerate into competing claims about methodology and data sets, and thus come out inconclusive.

    Like my post said, I’m not arguing the effect of corporate taxes is zero, but probably only detrimental if above world (or even just US) rates by a decent margin, and the straight evidence from Toby’s post is that lowering rates ever more below US rates has had no appreciable effect. Any decent analysis should also take into account electricity prices, size of domestic market, etc that also determine international investment patterns, rather than only focus on corporate taxes as if that is the only thing that matters.

    I conclude that the impact of corporate taxes is small enough that, within the range of recent history, we need not be concerned about corporate tax increases back to 2000 levels, especially in the purely domestic sector, which comprises most of the services in the Canadian economy.

    And if the proceeds are used to support pro-growth spending, it’s all good.

  • David why is the onus on Marc because Stephen produced a junk study? And since when did simply having a name brand behind junk make shine like a diamond? If argument to authority is all that now counts I will wait for you to get your Swedish bank prize or do a Ph.D. at Harvard before I bother reading anything you have to write.

    The issue is really straight forward. The question is do corporate tax decreases increase investment and growth in Canada. What happens in Ghana is of no import. If this shoddy study is all one needs to believe that corporate tax = growth then it is kind of proof that one is engaging in confirmation bias.

  • I see your point Travis. But in economics, we’re interested in analyzing incentives and mechanisms. Corporate tax rates in Ghana matter if the mechanism and incentives work similarly in Canada. It’s unclear whether or not this is the case, but it’s a good starting point.

    I agree that “brand names” can produce bad studies; I was just trying to argue that we shouldn’t discount the study solely on the grounds that it hasn’t yet been published, because it does some have some credibility behind it at first glance.

    Marc, I think your case is quite reasonable. It think you have a decent argument that corporate tax rate increases, at the margin, might not be that detrimental. I’ll differ with you ideologically on whether raising taxes so that government spending can be increased is a good idea. But if your goal is raising government revenue, I think you’ve got a case that corporate tax rates might be an OK way to do it.

  • We do not need to start with Ghana we have all the time series data we need for Canada.

    Dave you have to ask yourself, with so many Canadian economists kicking around who have access to the CDN data going back at least forty years, why we do not have a raft of empirical studies showing the growth miracle that is corporate tax cuts. You would think given how many (liberal) economists believe in the CIT cuts = extra growth that it would be one of the most studied and empirically well documented truths. That it is not tells us something.

  • I don’t find the paper terribly convincing relative to the debate at hand. The claim is that decreasing corporate income tax rates in Canada has not led to an increase in private fixed capital investment. A cross-sectional analysis, however interesting and well-constructed, doesn’t really address that claim. As Travis notes, one would need a time-series analysis of Canadian data — or a cross-sectional time-series analysis allowing for varying coefficients across time and countries — to adequately rebut the claim. This study does not provide that.

    Five things in the paper do stand out for me, though, to the extent that they may apply to our present case:

    1. The statutory (as distinct from the effective) corporate income tax rate has no effect on investment. If we assume the paper’s direct applicability, that implies that all the real work is done by what is or is not tax deductible.

    2. Investment levels are higher as the top marginal personal income tax rate increases. So we need to increase taxes on the rich.

    3. There’s also a direct relation of investment to currency in circulation. So we need to loosen up on the money supply.

    4. Figure 2 is more convincing than Figure 1, though both depend heavily on a comparatively small proportion of the cases. But it also suggests a non-linear relationship that is basically flat from about 20% and rather close to flat from about 14%. Far from rebutting Toby and Marc’s argument, that actually supports it — even FDI is not very responsive to variation in corporate income taxes in the range in which our present debate is taking place.

    5. Of course, if we assume that declining corporate income taxes actually has increased private fixed capital investment relative to what it would otherwise have been, yet investment is still negligible, that suggests that Canada faces a much bigger problem. And this paper certainly provides no succour to the most common arguments about what else to do (i.e., labour market “flexibility”, “cutting red tape”, “trade liberalization” — cf. tables 5d, h, i, l).

  • Um, I didn’t expect anyone to believe that one study would be definitive. I called your attention to the long, long – and yet far from exhaustive – list of studies that found the same thing.

  • And if we’re going to get into the econometric nitty-gritty, then as a Bayesian, I would note that if you have dogmatic prior beliefs, then there’s really not much in the way of empirical evidence that could ever persuade you to change your mind.

    Any reading of the literature shows that the vast weight of evidence demonstrates the negative effects of corporate income taxes. At this point, refusing to join that consensus says more about you than it does about the data.

  • C’mon, Stephen. At least direct us to the literature surveys that overwhelmingly support your point.

    For surveys, I would point to Peter Lindert’s Growing Public book(s), an IMF survey by Gerson, and a review of the econometric literature by Sala-i-Martin in support of my position (full references in my CCPA paper), which is hardly dogmatic.

  • Huh? I’m not arguing for smaller government, nor does that link say anything about corporate income taxes.

    And you will no doubt recall from “Growing Public” that low corporate tax rates are part of the recipe for a prosperous social democracy.

  • C’mon, Stephen. At least direct us to the literature surveys that overwhelmingly support your point.

    In thought I did. Remember this paragraph?

    “Starting with Jorgenson (1963) and Hall and Jorgenson (1967), many public finance economists have addressed this topic. A small selection of important studies includes Summers (1981), Feldstein, Dicks-Mireaux and Poterba (1983), Auerbach (1983), King and Fullerton (1984), Slemrod (1990), Auerbach and Hassett (1992), Hines and Rice (1994), Cummins, Hassett, and Hubbard (1996), Devereux, Griffith, and Klemm (2002), and Desai, Foley, and Hines (2004b). Auerbach (2002), Gordon and Hines (2002), Hasset and Hubbard (2002), and Hines (2005) survey aspects of this literature. Generally speaking, this research finds adverse effects of corporate income taxes on investment, although studies offer different estimates of magnitudes.”

    That’s 10 important articles (among many) plus four literature surveys. What more do you want?

  • Marc, have you really missed the point that badly? Do you really think that an argument for lower corporate taxes is the same thing as an argument for a smaller government?

  • Stephen, nothing you have said contradicts the basic observation that in the Canadian context we have not seen an increase in efficiency, investment or growth in relation to reduced corporate income taxes.

    All you seem to be able to do is say “but the literature says …” when in fact the literature is much more subtle and nuanced than you care to admit.

    I still think the corporate tax cuts of the past decade in Canada could be reversed with minimal impact on the economy and could even be pro-growth with good spending of the proceeds.

    I suppose we’ll have to agree to disagree. In the meantime, I’m sick of having you continue to misrepresent what I have said, while replying in a disrespectful and patronizing manner with your comments.

    Good night. Go Canada Go!

  • Well, I’m sorry if I misrepresented you, but I think it’s an honest mistake. You countered my point about corporate taxes with an article that doesn’t mention corporate taxes. I did the best I could.

  • It might also be worthwhile to take a look at this paper


    when considering the appropriate level of corporate taxes.

  • Honest mistake… more like pathological. Stop playing the church mouse when you are properly called out for what can only be characterized as an adolescent goes to grad school manner of communication. You banned me from your blog for much less. Quite why PEF tolerates you is beyond me. Nietzsche once wrote that the impotence of Christians’ love for God was evidenced by the fact that they refused to burn him (FN). I guess you do not rank that high. QED.

  • Stephen – here’s an earlier post of mine re a Department of Finance study.

    It shows that there are more effective ways to support corporate investment than across the board cuts to the general rate. That would be the argument I would make – redirect the cost of a rate cut to targeted programs and selective tax credits and the like, rather than to general revenues. I suspect you are right that taxation of capital income has an impact on the level of real investment but my sense is that it is pretty modest. Also that it will not affect the real rate of investment in very profitable sectors, like resources. In an ideal world the CIT is not a good way to tax and we should levy tax instead on profits paid out to owners, but that is difficult to do in a context of high foreign ownership.

    As for the rest of you, be civil. Debate is good.

  • Its a bit hard to be civil when the debaters on the other side don’t understand what that means.

  • “As for the rest of you, be civil. Debate is good.”

    I am disappointed in you not because you have lied to me but because I can no longer believe in you. It is to weep.

  • As for the rest of you, be civil. Debate is good.


Leave a Reply

Your email address will not be published. Required fields are marked *