Everything that needs to be said has already been said.
But since no one was listening, everything must be said again.

Andre Gide

Monday, September 5, 2011

Burtless: It's *not* regulatory and tax uncertainty

Gary Burtless, an economist at the Brookings Institution, seems pretty sure that the relatively depressed levels of U.S. business investment and employment have nothing to do with the alleged uncertainty over future tax and regulatory regimes. Mark Thoma reports the story here.

The main thrust of the argument is contained in the following paragraph:
Then why was uncertainty about taxes and the future burden of the Affordable Care Act holding back business investment right now? If managers thought taxes or regulatory costs might go up in the future, wouldn't it make sense to take advantage of today's low taxes and lower burdens to invest and hire today? According to the "uncertainty" argument, businesses are fearful they might face high taxes and extra health cost in 2016 and 2018. Shouldn't they expand hiring right now and scale back employment when they actually face higher costs (if they ever do)?
Burtless raises some good questions here, but I don't think they are the nail-in-the-coffin he makes them out to be. Why not more investment now, if taxes might go up in 2016 or 2018?

First of all, I'm not sure that those are the only dates businesses have to consider (governments can raise taxes anytime). Second, many large capital projects take a lot more than just a few years to complete. Think about the act of committing a large amount of capital (belonging to you, your friends, your creditors, your shareholders) destined to payoff (if at all) sometime in the distant future. Once committed, this capital is almost completely irreversible and--significantly--it is easily appropriated, since capital cannot run away once it is built). Is it completely crazy to imagine that those contemplating such investments in the current economic climate might want to worry (among other things) the possibility of future changes in tax regime?

Well, maybe my argument does not work so well for employment. As Burtless suggests, why not hire people now and then lay them off if and when taxes rise? One response to this is: How does he know for sure that the future regulatory climate will allow firms to lay off people in this easy manner? If the U.S. is moving to a more European-style economic model (and I'm not saying here whether this is good or bad), then firms may at some point in the future face large penalties for letting workers go.

So what is the problem, according to Mr. Burtless. Predictably, it is this:
The odd thing is, when businesses are asked why they're not expanding, "high taxes" and "heavy regulatory burdens" and "tax uncertainty" don't feature as prominent answers. They mostly say they don't see good prospects for extra sales. But right-wing economists have their talking point, even if they make little sense, and they're sticking to them.
Ah yes, those evil right-wing economists (one can see the halo hovering over his head as he says this).

I've tackled the issue of how firms reply to these business surveys here: Deficient Demand: The Deflated Balloon Hypothesis. Basically the idea is as follows. Consider any shock that leads to a contraction in one sector of the economy. Imagine that sectors are characterized by an interlinking network of demands for intermediate goods and services. A collapse in residential construction can now be expected to reverberate throughout the economy. A decline in the demand for housing also leads to a decline in the demand for all the products that go into making houses. It would not be surprising for someone in the business of producing (say roof shingles) to report that his or her main problem appears to be a "lack of demand" for their product.  But that by itself does not constitute evidence that the macroeconomic problem is a "lack of aggregate demand."

Burtless may very well end up being correct in his assessment. I'm just not sure how he knows for sure that what he says is true.

Updates: September 06, 2011

Regime Uncertainty: The Real (Option) Deal, Craig Pirrong
This is a direct rebuttal to Burtless (h/t Prof J)

Other related links:
Regime Uncertainty: Why the Great Depression Lasted So Long and Why Prosperity Resumed After the War, Robert Higgs
The Great Recession and Government Failure, Gary Becker (h/t Alex Karaivanov)


  1. 2 points:

    1) you're right about the fact that uncertainty will generally lead to lack of investment and hiring: this is just Bernanke's (1983) point about irreversability. If a factor is difficult to get rid of, then uncertainty will tend to reduce demand for this factor since there's an option value to delay. As you argue, even labor has an element of irreversability about it, at least in expectation.

    2) "It would not be surprising for someone in the business of producing (say roof shingles) to report that his or her main problem appears to be a "lack of demand" for their product." You know very well that it's very difficult to generate aggregate fluctuations from sector-specific shocks in macro models. While the "shingle" maker might report low demand from a decline in housing construction, it's a much bigger stretch to get a shoe-maker (or whatevever that's not construction related) to say the same thing in the absence of aggregate shocks. It seems pretty clear that the downturn is not limited to the construction industry (or related sectors) but is fully widespread and "aggregate". So I'm much more willing to buy his point that when business people systematically point to low demand (across industries) as the problem, there most likely is some kind of "aggregate demand" deficiency.

    - C

  2. I think David's points are pretty much right; uncertainty could wind up looking like deficient demand in surveys. And it is very hard to prove that regulatory and tax uncertainty are not causing important problems.

    But then I wonder what the evidence on the other side is. Does anyone have a quantitative measure of regulatory uncertainty? tax uncertainty? Has anyone done objective work trying to relate such measures to overall economic activity? I think the answer to every one of these questions is "No." (This is my open invitation to all readers to prove me wrong on this point.)

    If that's the case, you have to wonder why the media contains so many reports of quotable people making the same talking points: tax and regulatory uncertainty is depressing the economy. Why would a voting FOMC member, such as Richard Fisher, keep repeating such conclusions without empirical work to back it up?

    More to the point, why would any serious economist take such claims seriously? (Beats me.)

  3. @Anonymous:

    You know very well that it's very difficult to generate aggregate fluctuations from sector-specific shocks in macro models.

    Um, no, anonymous...I do not know this. I do know, however, a great many people like to make that assertion. One of the very first real business cycle papers (Long and Plosser, JPE 1983) stressed intersectoral linkages and sectoral shocks leading to aggregate business cycles. Andreas Hornstein (JME 1997) comes to a very different conclusion from your own; see here: http://ideas.repec.org/a/eee/moneco/v40y1997i3p573-595.html

    I am honestly at a loss to understand why the profession largely ignores intersectoral linkages. I guess it's much easier to just specify a single "aggregate demand shock"...whatever that is.

  4. @Simon:

    More to the point, why would any serious economist take such claims seriously? (Beats me.)

    Which claims should we be taking seriously, and why? Can you provide us with slam dunk evidence of what is depressing investment and employment? I'm all ears.

  5. David:
    1) I'm sure you know that there is a difference between taking something seriously and being unable to rule it out. Can you prove that Sarah Palin is not Elvis' love-child with a space alien? Would you take that claim seriously?
    2) A banking crisis. (Duh!) There's plenty of empirical evidence on their macroeconomic impact. There's lots of theory on the transmission mechanisms (as you may know better than I.)

    P.S. I share your frustration with the way our profession insists on looking at "aggregate" shocks rather than considering intuitive decompositions.

  6. Here is Robert Higgs, who wrote the first paper on the impact of regime uncertainty in lengthening the Great Depression, talking about Burtless' article: http://blog.independent.org/2011/09/05/regime-uncertainty-pirrong-debunks-the-keynesian-debunking/#more-12374

    The need to quantify something before it becomes meaningful is the worst form of scientism. The most important forces that shape our world, like institutions, traditions, etc., are not quantifiable.

    Bob Higgs' work is not mathematical econ. It is objective. And it is important.

  7. "Um, no, anonymous...I do not know this. I do know, however, a great many people like to make that assertion. One of the very first real business cycle papers (Long and Plosser, JPE 1983) stressed intersectoral linkages and sectoral shocks leading to aggregate business cycles. Andreas Hornstein (JME 1997) comes to a very different conclusion from your own; see here: http://ideas.repec.org/a/eee/moneco/v40y1997i3p573-595.html"

    The paper by Hornstein is nice, but shows only how one can get durables and nondurables to move together in response to sector-specific shocks. But this is a long way from establishing that a shock to residential investment in a few geographic areas can generate a significant downturn that affects all areas of economic activity. While intersectoral linkages can go some way in generating comovement (Ruge-Murcia also has done recent work on this), I've yet to see a paper document how regional declines in real estate prices can generate an economy-wide recession. This is precisely why most economists (as far as I can tell) were predicting rather modest effects from a downturn in housing prices prior to 2008.

    - C

  8. @Anonymous:

    But this is a long way from establishing that a shock to residential investment in a few geographic areas can generate a significant downturn that affects all areas of economic activity.

    I agree with you that this class of models remains a long way off (as do most macroeconomic models). As a "newmonetarist," I believe that financial market frictions play an important role in financial crises and subsequent recovery. I think it might be fruitful to combine the sectoral and financial frictions stories to come up with an empirically plausible story.

    Btw, this does not mean I would discourage work that tries to explore the plausibility of "deficient demand" interpretations. My post here was mainly about an economist expressing a very strong opinion about something without having thought very deeply about subject (or so it seemed to me).

  9. @Simon:

    Sure, but I am asking you which is your preferred transmission mechanism? Can you point me to the quantitative model that you find most plausible in transforming a 2008 financial shock (whatever that was) into (say) a decade-long period of under-performance?

    As a "newmonetarist," I am very sympathetic to the notion of financial market frictions playing an important role in shaping the nature of a crisis. But I also do not discount the possibility that other factors, like regime uncertainty, may play an important role (even if such things are hard to quantify.)

  10. Seems to me that the "Tax & Regulation" quote gets far more play in the media than the evidence would warrant. With so much slack in the economy, I'd be more inclined to suspect demand as the dominant issue. This might explain the number 1 business concern of slow sales/revenue.

  11. @Anonymous,

    Most people would probably agree that the aggregate demand (for investment goods, in particular) appears to be relatively depressed. The issue is not that; instead, the question is what is responsible for this state of affairs? Many different culprits may be at work and its hard to quantify their relative importance. This, I think, is the source of much disagreement.

  12. David:

    I am more sympathetic to the role of "uncertainty" affecting demand, than to claims by those who have divined, evidently by introspection or similar means, that it is uncertainty caused by recent government actions, and more specifically, tax and regulatory uncertainty. The latter seems so easy to interpret as a convenient political ploy that I'm compelled to ask for evidence. I've still not found it. (I also can't help asking whether past periods of profound tax uncertainty -- e.g. the Reagan Tax reform in the mid-80s -- created similarly weak growth.)

    BTW, doesn't "uncertainty" more generally have problems once we put it into a macro model? I would have thought that once we're talking about decision horizons lasting more than a year or so, our conditional uncertainty doesn't move so much. Aside from dummy variables for recessions and the Great Moderation, do we see important and persistent moves in macroeconomic uncertainty? Or is "uncertainty" just another degree of freedom we're adding to the model?

  13. Uncertainty is really hard to measure directly, but a lot of people consider the implied volatility indices decent proxies and the VXO index is actually very high at the moment. It's shot up pretty dramatically since the summer, and this seems consistent with the recent pause in hiring during August (and the index has been somewhat elevated since the huge spike in 2008 anyway). Also, the large holdings of money and reserves by firms and banks, respectfully, is certainly consistent with a precautionary balance motive related to uncertainty about the future. So, I don't think it's as simple to say uncertainty doesn't matter, as a lot of people are inclined to think. Now, whether this is policy uncertainty or just uncertainty about the future fundamentals of the economy is really difficult to say.

    I do think Gary Burtless is a bit confused, however. If the issue is uncertainty, we aren't talking about fear - we are talking about uncertainty. Fear of the future is different that uncertainty. Dodd-Frank and the Affordable Care Act are thousands of pages long and have hundreds of to-be-written regulations. I don't think it's particularly implausible that a lot of firms since don't know what their future costs will be. Gary seems to be thinking of a world with perfect reversibility of capital and virtually no adjustment costs in capital and labor. In that world, he might have a point. But in our world where capital and labor aren't perfectly reversible and entails adjustment costs, for the right parameters, it is plausible that it may be rational for a firm to wait it out and see what the regulatory costs will be to their firm before engaging in hiring and investment decisions. I also think people ignore the role of firm entry. While firm entry is not standard in macroeconomic models (though Marc Melitz and some co-authors did interesting work on this recently, though they modeled firm entry as a production line - but the idea is similar), I think it's clear that firm entry is extremely important for output and employment growth. If an entrepreneur can time their entry, and they are deeply uncertain about their future costs, if may be rational for them to not enter and wait until they understand regulatory burdens better.

    With all that said, I don't really know if policy uncertainly (which I certainty believe exists in today's climate) is quantitatively important enough to explain the persistence of this slump. I tend to believe that policy uncertainty plays a non-trivial role, but it is less important than other factors (including just general uncertainty about fundamentals) which I don't fully understand. A larger point is that this recession is confusing. There is a lot of moving parts going on and for that reason alone people shouldn't be too quick to throw out conjectures and hypothesis that aren't entirely implausible, even if you don't personally find them convincing. Or, at least you shouldn't do that if you are more interesting in understanding, rather than cheap politicking.

  14. More sources of uncertainty:


  15. "Does anyone have a quantitative measure of regulatory uncertainty? tax uncertainty?"

    the NFIB survey.

    please do let us know the results of your empirical studies "trying to relate such measures to overall economic activity". i, for one, would be interested.

  16. The problem with the uncertainty argument is that it ALWAYS applies. Something that can always be true explains nothing and therefore should be ignored. It also can NEVER be quantitated so it has no use when deciding economic policy.

    One thing that can be quantified is the lack of sales the last few years. No restaurant that has closed or laid people off did so when they had all their tables full every night. No engineering firm laid anyone off when they had backorders. That is what our economy is lacking ..... the volume of sales it had in 2006/7. Its quite simple. All the rest of it is claptrap.

  17. Anonymous: The NFIB survey doesn't have an uncertainty measure, does it? Following your link, I see lots of survey questions about whether it is a good time to expand, etc. But nothing about uncertainty, much less regulatory or policy uncertainty.

  18. @Greg:

    I cannot disagree with you more. While expectations and views on uncertainty are difficult to quantify, it can be done. Indirect evidence is sometimes available through prices (or objects like the VIX).

    You say that the problem is a lack of sales. From national income and product accounting, we know that sales (expenditure) is equivalent to income, which is, in turn, equivalent to production. Consequently, we might equally well say that the problem is a lack of income and production. While this may be true, it is not, in itself, very helpful.

  19. Three basic facts do not lend much credence to regime uncertainty.

    1. Investment in equipment & software is robust.


    2. The tax regime is very favourable.


    3. The decline in private investment in nonresidential structures since the peak in 2008 appears to be following a typical cycle.


  20. Simon wrote: "The NFIB survey doesn't have an uncertainty measure, does it? ...nothing about uncertainty, much less regulatory or policy uncertainty."

    There are two relevant questions, on whether a respondents's greatest concern is 'government regulation' and 'taxes'. Those seem close enough to what you're after.

  21. Of course it's not regulations or taxes. They didn't cause this crisis. In fact, it has never been taxes that are holding business back. Only dogmatic Republicans think so. Compared to the rest of the world US has very low taxes. High tax rates usually results in structural unemployment, but that's not what we are facing.

  22. If it is true that 'while expectations and views on uncertainty are difficult to quantify, it can be done', what emirical evidence suggests that uncertainty over 'future tax and regulatory regimes' sometime in 'the distant future' is at such high a level that it can explain the current predicament we are in?

  23. Anonymous:

    Haha...I guess I should have said "it can be done in principle, and effort should be devoted to the task."

    I do not have any direct evidence. But think about it. Think about the trajectory of government spending and taxation (and implicitly, the debt).

    At the moment, there appears to be a large appetite for US Treasury debt, allowing the US to postpone the taxes necessary to finance a still unknown future expenditure path. What happens if bondholders around the world begin to dump their Treasuries--what happens then to taxes and/or government spending? One could go on and on...surely these type of hypotheticals must be going through the minds of businessmen (I am guessing).

    Here is Barro in a recent WSJ piece:

    What drives investment? Stable expectations of a sound economic environment, including the long-run path of tax rates, regulations and so on. And employment is akin to investment in that hiring decisions take into account the long-run economic climate.

    The lesson is that effective incentives for investment and employment require permanence and transparency. Measures that are transient or uncertain will be ineffective.

    And yet these are precisely the kinds of policies the Obama administration has pursued: temporarily cutting the payroll tax rate, maintaining the marginal income-tax rates from the George W. Bush era while vowing to raise them in the future, holding off on clean-air regulations while promising to implement them later and enacting an ambitious overhaul of Wall Street regulations while leaving lots of rules undefined and ambiguous.

  24. So you 'do not have any direct evidence' and are 'guessing' instead. And on that solid foundation, you (and Barro) claim that it is 'uncertainty about future regulations and taxes' - and not lack of sales - that is the cause of the lack of investment. Okay.

    One question, though: If your theories are correct (never mind that you do not have any empirical evidence to support them) then why didn't the business community become uncertain about future taxes and therefore stop investing back when the debt started rising dramatically already during the first years of the Bush administration?

  25. Oh, and here is El-Erian, CEO of Pimco. The headline is: Policy Makers Will Again Keep Markets on Edge.

    Link: http://www.cnbc.com/id/44572476

  26. Oh, and how about:

    Investors want evidence that the political process in Washington is capable of tackling the towering U.S. deficit and the country's mounting debts, after ratings agency Standard and Poor's cut the U.S. AAA rating in August.

    Link: http://www.cnbc.com/id/44573611

    Summary: I think there is enough evidence around to suggest that policy uncertainty is playing some role (contrary to the claim made by Burtless). Maybe not everything, but possibly something.