Temporary and Permanent Ricardian Confusion: Going Comfortably Numb

Spurred by the heated debates about the need for fiscal stimulus in the US, the issue of Ricardian Equivalence has taken center stage in the economic blogging sphere recently. While it is an impossible task to identify any exact line of events on the net (and possible also irrelevant), this round appears to have been initiated by an article by Justin Yifu Lin (pdf), Chief Economist of the World Bank, who got criticized here by a balanced Antonio Fatás. Fatás notes, among other things, that Lin’s fears that fiscal stimulus could be caught by the “Ricardian trap” (i.e., neutralized by offsetting increased private savings) are unwarranted. While Lin’s endorsement of productive government spending (like public investment) is hard to dispute, it blurs the fact, according to Fatás, that in a depressed economy, spending on perishables can have its own benefits irrespective of the “Ricardian trap”.

That post was then picked up by Nick Rowe and Paul Krugman. Rowe somewhat comes to the rescue of Lin, and has some interesting points about the measurement of government spending, arguing that if spending taking the form of “pure waste,” as considered by Lin, it is equivalent to a transfer payment; i.e., a tax cut. It will therefore have no expansive effects if Ricardian Equivalence holds. Hence, it is indeed important, as Lin argues, that spending is directed towards productive usage. Krugman steps in (step 1 and step 2), and digs out his earlier writings on Ricardian Equivalence, where in one of his many blog posts he writes

“But suppose that the increase in government spending is temporary, not permanent — that it will increase spending by $100 billion per year for only 1 or 2 years, not forever. This clearly implies a lower future tax burden than $100 billion a year forever, and therefore implies a fall in consumer spending of less than $100 billion per year. So the spending program IS expansionary in this case, EVEN IF you have full Ricardian equivalence.

Is that explanation clear enough to get through? Is there anybody out there?”

Yes, I am. And it appears that the thrust of arguments is that a permanent increase in government expenditures is fully crowded out by private consumption rendering it impotent, whereas if the increase is temporary, and counts as GDP, then it is expansive even under Ricardian Equivalence. These claims are just not necessarily correct, as they are based on pure accounting and a simplistic view of private-sector responses to policy changes. Moreover, it generally requires more assumptions to get government spending to be expansive. And most people seem reluctant to mention them. Why, I don’t know. Maybe Fatás has part of the answer when he attributes it to economists’ resistance to be labeled Keynesians; at least I would guess such an explanation fits well for the US (but, luckily, to a lesser extent for Europe I would claim/hope).

So I just want to go through the arguments behind Ricardian Equivalence for the umpteenth time, to make sure that people (and my students) know what it is, and what ramifications it have (or do not have) for the effectiveness and desirability of fiscal policy stabilization. A quick disclaimer from the beginning: I do not believe in Ricardian Equivalence as being a good approximation to reality. Nevertheless, it is clearly formulated theoretical results, which provoke us to think deeper about reality. (I should also note that Fatás last week commented on the debate.)

The Ricardian Equivalence theorem takes as starting point that for a given path of government expenditures it does not matter whether expenditures are financed by deficits or taxes. For the government to satisfy its intertemporal budget constraint (i.e., not running Ponzi games where debt is continuously rolled over and exploding), any change in the path of taxes must be accompanied by an offsetting change in the path of deficits such that the present value of all future deficits and surpluses match the present value of all future government spending (plus the value of initial public debt). So far, this is pure accounting. The big implications come when one combines it with private sector behavior. The path of government net surpluses needed to finance a given path of expenditures is a liability of the private sector. In an idealized world where the private sector is living as long as the government (say, infinity), has access to perfect capital market at the same interest rate as the government, and where taxes are non-distorting, then this liability will not change if government spending does not change. Hence, the strong result: A tax cut has no effect on private sector behavior, as it does not change the present value of the government’s future deficits and surpluses. In words, consumers know that the associated deficit will have to be covered by a tax increase at some point in the future (with interest). They save the tax cut, such that they can be ready to pay the anticipated increases in the future. To repeat, with given government spending the Ricardian Equivalence theorem is clear: Tax and deficit financing is equivalent, and a switch between the two is irrelevant, making consumers in a perfect world be indifferent as well.

When considering changes in the path of government spending, things get much more complicated. Even in the case where Ricardian Equivalence applies (and I assume this throughout as well as I assume that government spending creates GDP), there is now a change in the present value of the government’s total financing needs, and correspondingly a similar change in the liabilities of the private sector. So, from a budget accounting point of view, an increase in government expenditures reduces households’ available life-time resources. The debate is therefore usually focused on whether this reduction is sufficient to fully counteract the increase in government expenditures or not. But then one is inevitably moving away from pure accounting principles and into the realm of private sector behavioral responses towards changes in resources. Hence, it is overtly simplistic to retain a pure accounting approach, and for example state that a permanent increase in government spending leads to a permanent reduction in private consumption of similar size leaving total spending unchanged. This can be consistent with households’ desire to have constant consumption, whereby they indeed each and every period reduce consumption by the same size of the government spending increase. Likewise, it is simplistic to state that a temporary increase in government spending increases total spending. Again it can be consistent with an idea of consumers wanting to keep consumption constant over time: As the drop in the resources is now smaller, their constant per-period consumption falls by less. Hence, total spending will increase (and then drop when the government spending increase is reversed, but that future is rarely spelled out in current times).

The behavioral responses are thus based on an extremely simplistic consumption-is-all-that-matters private sector. Many entirely standard macro models are a tad more complicated than that, and they can therefore easily give the completely opposite results than what, e.g., Krugman is claiming. Think of the situation where households also face a decision about their labor supply. Then, in a classical model (without any dynamics), a permanent increase in government spending increases total spending and production as labor supply goes up (the first-order effect of reduced consumption spending is an increase in the marginal utility of consumption making work more desirable); see. e.g., Michael Woodford, 2011,” Simple Analytics of the Government Expenditures Multiplier”, American Economic Journal: Macroeconomics 3, pp. 2-4. Then think of a Ramsey-Cass-Koopmans model with endogenous capital formation and fixed labor supply. In this model, a temporary increase in government spending instantaneously reduces consumption and investment, leading to falling output and higher real interest rates. Consumption indeed drops immediately by less than the government spending increase, and lesser the shorter policy change, but it starts growing in anticipation of the future reversal of fiscal policy (In the continuous-time version of the model, consumption will remain approximately unchanged if the government spending increase lasts for an infinitesimal length of time – this case appears analogous to Krugman’s two-period, short-run/long-run model, where a short-run spending increase leaves consumption unaltered.) When the increase in government spending is reversed, investment picks up, and output grows back to the initial level (see David Romer, 2006, Advanced Macroeconomics, p. 73). So it all depends on the private sector responses to the policy change. Budget accounting, on which the equivalence theorem is based and satisfied here, is just not doing the job here when government spending changes. So I hope people in the debate would focus more on specifying what transmission mechanism they have in mind instead of insisting on this merely being a matter of permanent versus temporary changes in government spending. It is clearly not.

The models used to exemplify the confusion are both flex-price supply-side models. Obviously, they are rigged against providing a rationale for expansive fiscal policies against unemployment problems, but in newer sticky-price models, government spending is usually expansive also under Ricardian Equivalence (for the mundane reason that output is demand determined). Actually, most New-Keynesians models focus on Ricardian Equivalence for simplicity. And temporary and permanent fiscal expansions can have positive output effects in these models (the models, however, have some problems replicating that private consumption often increases with government spending). And I would imagine that most economists favoring demand management policies to have some Keynesian sympathies (despite that they may not admit it as Antonio Fatás wrote). Read Pierpaolo Benigno, 2009, New-Keynesian Economics: An AS-AD View, NBER Working Paper, No. 14824 for a simple and clear exposition, which also covers distortionary taxes. Note however, that the flex-price scenarios are not just theoretical playgrounds that silly mathematical economists like to fumble around with. They are the ones policy should aim at replicating as close as possible. One of the main contributions of the New-Keynesian literature is indeed the focus on designing policy to be stabilizing welfare-relevant gaps between, e.g., output and its natural (flex-price) level.

It is therefore very important, as always in economic debates, that people are transparent about what their essential assumptions are. Otherwise things will end in a shouting contest where participants are being, or faking to be, completely numb. It may be comfortable to rest unaffected with one’s old arguments, but economics is not about being right. It is about getting smarter. So, when Paul Krugman ends by crying out “Is There Anybody Out There?”, I feel the urge to extend my “Yes” with:

I need some information first
Just the basic facts
Can you show me where it hurts?”

From “Comfortably Numb”, music by David Gilmore/Roger Waters, lyrics by Roger Waters. © 1979 Pink Floyd Music Pubs. Ltd. All rights reserved.

This entry was posted in Economists, Macroeconomics and tagged , , , , , , , , , , , . Bookmark the permalink.

4 Responses to Temporary and Permanent Ricardian Confusion: Going Comfortably Numb

  1. Pingback: Debate on Ricardian Equivalence (links) | Damoun Ashournia

  2. Pingback: De-convoluting John Cochrane’s 2009 piece on fiscal stimulus – Kantoos Economics

  3. Pingback: A blast to a myth bust: in economie 1 plus 1 nu fac intotdeauna 2 | Lucian Isar

  4. Pingback: A blast to a myth bust: in economie 1 plus 1 nu fac intotdeauna 2 « voxpublica | Platforma de comentarii, bloguri si opinii REALITATEA.NET

Leave a Reply

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


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>