Reason to Be Skeptical About Stimulus

From Econtalk.org

Valerie Ramey of the University of California, San Diego … finds a multiplier between .8 and 1.2. (A multiplier of 1 means that GDP goes up by the amount of spending–there is neither stimulus nor crowding out.) She also discusses a survey looking at a wide range of estimates by others and finds that the estimates range from .5 to 2.0.

If the multiplier is 2.0, then there’s no argument; let’s get cracking on stimulus. If the multiplier is below 1.0, there’s no argument’ let’s avoid stimulus like the plague. If it’s 1.1, 1.2, 1.3? Then I suppose we’re in the range of, “sure it’s an effective policy, just not wildly effective; maybe there’s something better?”

But when we don’t actually know with certainty, then we don’t know if we would actually help or harm the economy with stimulus. I’m not a fan of the precautionary principle myself, but most of my liberal friends have warm feelings for it. So why aren’t they more conflicted about stimulus?

Advertisements

About J@m3z Aitch

J@m3z Aitch is a two-bit college professor who'd rather be canoeing.
This entry was posted in Economical Musings. Bookmark the permalink.

22 Responses to Reason to Be Skeptical About Stimulus

  1. D. C. Sessions says:

    Obviously, there’s a difference between different classes of stimulus. For instance, pulling in maintenance and repair of infrastructure (e.g. bridges) that that have positive economic utility and we’re going to pay for regardless from the future when costs will be higher to now when we’re paying a negative real rate of interest is a whole different thing from paying people to dig holes and fill them in.

    Likewise, presumably, for “infrastructure” like teachers. I don’t know that anyone has attempted to do a cost analysis of sacking a half-million teachers, police, etc. but I’m not bold enough to claim that it’s zero.

  2. James K says:

    DC Sessions:

    The big issue is one of dynamics. Infrastructure (especially maintenance) tends to have good outcomes over long time horizons, but in many cases the effects are too mild in the short run to effect the macroeconomy much. Conversely, firing staff would have an immediate negative effect, that would diminish to zero over time (the better the state of the labour market, the quicker the effect will disappear).

  3. Troublesome Frog says:

    James K:

    There’s still an asymmetry in the claim: If firing staff has an immediate negative effect that diminishes over time, why would we not expect hiring staff at the same margin not to have an immediate positive effect that diminishes over time? If such an asymmetry exists, it would imply some very strange things to me. I would expect at least some continuity here.

    James Hanley:

    I think that one of the reasons we see such variation in multiplier estimates is that it’s hard to nail down the question. To me, the question is not, “What is the average multiplier for government spending?” I’m more interested in, “What is the expected multiplier for these particular circumstances?” The key variables for me are:

    1) Is unemployment very high?
    2) Is the interest rate at or near the zero lower bound?

    Those are the circumstances in which I would expect to see the biggest multipliers. The opportunity cost for labor is quite low and the crowding out effect is negligible. Under normal circumstances, I would expect to see much more crowding out. If you separate out these special cases, you have relatively few data points. If you average them with normal circumstances, the effect would appear to go away.

    DC Sessions makes the additional point that we’re busily cutting investment at exactly the moment when any sensible business would be investing like crazy. Money is cheap and we have a bunch of capital expenditures that we need to make anyway. To me, the simplest question is this: If we put the executives at Goldman Sachs in charge and asked them to decide when to run a surplus or a deficit, what would they do? You’d be hard pressed to argue that they wouldn’t be borrowing money hand over fist at negative real rates to spend it on things that have even the smallest rate of return.

    Have you thought any more about Brad Delong’s post on the velocity of money and interest elasticity?

  4. DR X says:

    I imagine that liberals are more receptive to stimulus because they see government as potentially a force for good or bad, in contrast to today’s conservative movement that tends to see government activity as inevitably bad. But do you really think that the average person, conservative or liberal, has even a rudimentary understanding of the issues? Multiplier effect? How many people who have an opinion actually have a clue about what is meant by multiplier effect, let alone the problems with determining it? The commenters here hardly represent a representative sample of people who have opinions on these matters. But granted that liberals are more inclined to see government activity as a solution to many problems, even informed liberals would be more inclined to see direct intervention as beneficial when they’re compared with people who see intervention as almost always a bad thing.

    On the actual question of stimulus, I think troublesome frog identifies some important issues. One thing that shouldn’t be forgotten in this discussion is that supporting stimulus is just as contingent on broader circumstances as easing or tightening credit.

    As for this particular estimate of the multiplier effect, I’d look at the mean estimate as a guide rather than selecting estimates closer to the tails. To reject stimulus based on the low-end estimate seems as suspect as endorsing stimulus based on a high-end estimate.

    One other matter, I think in our specific situation, infrastructure investment isn’t just helpful in the long run. The construction industry has been especially hard hit and infrastructure investment probably taps most directly into use of currently underutilized capacity.

    In another blog, someone made much of the idea that this would make sense, except that people in the construction industry aren’t, by and large, suited for the kinds of infrastructure projects government would spend on, to which I say nonsense. Civil engineering and construction is my family’s business. My brothers have a substantial business and they’ll build just about anything you ask them to build, from shopping malls, to office towers, hospitals, stadium sky boxes, schools, restaurants or hotels. In connection with these projects, subs they’ve hired have done road building, sewage and drainage work and other kinds of work that would be typical of municipal investment in infrastructure.

    They’ve weathered the current storm, albeit with substantial decrease in their revenue, because they’re wealthy, because they’ve have only about 35 people on their payroll and they do considerably more than construction. They’ve been in the business long enough to anticipate and plan for big slumps. But they’ve probably stopped using the services of a couple of thousand employees of subcontractors in the past 3 years. In fact, they’ve built a business on subs precisely because it affords them the greatest flexibility in changing conditions.

    There are shitoads of underemployed iron workers, heavy construction laborers, electrical contractors, plumbing contractors, paving contractors, painters, landscape workers, bricklayers, HVAC, haulers and a host of people that would be ready to go for work on a wide range of infrastructure projects–the list occupations and unused manufacturing capacity is very long. And I know for a fact that people in that business will relocate, they’ll live in cheap apartments, trailers, motels or whatever if they get work.

    So if you compare the effects of stimulus in projects that more directly employs these people, versus say stimulus with the government buying from an industry that hasn’t seen a steep decline in employment, I think the difference in multiplier would be appreciable. In a sense the neglect of infrastructure in recent years might be considered fortuitous given that this is exactly where we have so much underused capacity currently.

  5. James K says:

    Troublesome Frog:

    If firing staff has an immediate negative effect that diminishes over time, why would we not expect hiring staff at the same margin not to have an immediate positive effect that diminishes over time?

    Sure, the effect should be symmetrical, or nearly so.

  6. James Hanley says:

    I’m more interested in, “What is the expected multiplier for these particular circumstances?” The key variables for me are:
    1) Is unemployment very high?
    2) Is the interest rate at or near the zero lower bound?

    I’m not persuaded that those are the variables that determine the multiplier. Given that I don’t actually know, I can’t make any serious claim that they’re not, but likewise you haven’t actually demonstrated that they are. I’m inclined to think the type of project spent on is more meaningful. Keep in mind that we expect private spending to be efficient because people spend on what they want. We expect government spending to be less efficient because much will be spent on what they think is wanted by others.

    For instance, pulling in maintenance and repair of infrastructure (e.g. bridges) that that have positive economic utility and we’re going to pay for regardless from the future when costs will be higher to now when we’re paying a negative real rate of interest
    Infrastructure spending is always good (making the assumption we’re not talking about bridges to nowhere), and obviously it makes sense to do it when you can do it more cheaply. I think it’s a mistake to put that in terms of stimulus, though, because it obscures the actual purpose of that spending and makes it more of a political football than it ought to be.

    I’m glad that here in Michigan our governor is talking about getting serious about ending Michigan’s decades-old tradition of deferred maintenance on roads and bridges, and glad that he’s flouting the current anti-tax-at-all-costs idiocy of his party colleagues to advocate a secure funding mechanism for it. He’s not pitching it as stimulus, but investment.

  7. D. C. Sessions says:

    I think it’s a mistake to put that in terms of stimulus, though, because it obscures the actual purpose of that spending and makes it more of a political football than it ought to be.

    You can subtract the net present value of the project from the cost, and obviously if it’s a good investment it’s a good investment. Mind, that’s not an argument that works with the “austerity at all costs” crowd: they’re the ones who will insist that when Americans are tightening their belts and deferring dental repairs [1], the government should let bridges fall down.

    However, even projects that are marginally less then fully justified can be worthwhile if the multiplier is greater than zero but not high enough to justify the exercise as make work. First, because putting people to work reduces other government expenses (unemployment insurance, food stamps, Medicaid, etc.) Secondly because the project’s net present value is legitimately subtracted from the cost but the multiplier applies to the full domestic tab, less the savings on other expenses.

    Finally, and less easily quantified, it’s better to have people working than drawing the dole. Aside from the moral hazard aspects and other intangibles, it maintains valuable human investments in skill and work habits.

    As above, this is a marginal argument. Manifestly not enough to justify any random project, but important enough to be factored into a cost/benefit calculation.

    [1] A co-worker told me that in 18th century America, the #1 cause of adult death was tooth infections. I haven’t checked it, but wouldn’t be surprised given the prevalence of dental-infection caused endocarditis and other bacteremias. News item recently of a gent who died of dental bacteremia because, following an ER visit he could afford either the antibiotics or the pain meds, but not both — and guessed wrong.

  8. Troublesome Frog says:

    I’m not persuaded that those are the variables that determine the multiplier.

    I’m surprised to hear that, because that’s pretty standard fare in this model. I am pretty convinced that the efficiency of the purchase has relatively little to do with it. Remember, we’re trying to stimulate aggregate demand without being swamped by the loss in aggregate demand that comes with government grabbing scarce resources. A few thought experiments to demonstrate what I’m saying:

    1) Imagine a dollar disappears from my pocket and shows up in your pocket by magic. What happens to aggregate demand? Is what you would do with the dollar less “efficient” than what I would do with it? If the change in wealth is large enough that it busts up my well-laid plans (say, I’m half way through building a factory and now I can’t finish it), that could be a problem. Otherwise, the difference in aggregate demand should be determined by our relative marginal propensities to consume. I argue that a magical, frictionless transfer of money from low MPC people to high MPC people would be stimulative on average.

    2) Imagine you were going to spend a dollar wisely, but you flip a coin to determine what you buy with it instead. What effect does this have on aggregate demand? My answer is that the first-order effect is zero. The second order effect is determined by the MPC of the two vendors who each had a shot at your dollar. The efficiency of the purchase wouldn’t enter into it.

    So if transferring money from high MPC people to low MPC people stimulates demand, what are the costs (aside from the political costs of taking money from people)? There’s the hit to “efficiency” caused by the government reallocating funds, but where does that come from? Opportunity cost.

    On the loanable funds side, the opportunity cost would be any loanable funds that the government borrows that now can’t be used for private investment. The mechanism for this would be increasing interest rates. This is where Brad DeLong’s earlier post comes in. We’re in a state where that mechanism is extremely weak, so it’s pretty safe to assume that we’re not crowding out private loanable funds.

    On the taxation side, a tax looks roughly like (1). The inefficiency is not in the dollar going from one place to another but rather the fact that the dollar reflect actual resource allocation when the rubber really hits the road. Yes, the private sector is generally better at allocating resources, but that effect is almost certainly weaker now: If resources were allocated the way we wanted them to be, we wouldn’t have 9+ percent unemployment. We have a huge pile of resources that are sitting around doing nothing. The opportunity cost of an unemployed person is quite low. If we were really operating at high efficiency, I don’t think we’d be spilling ink on this question.

    I think it’s a mistake to put that in terms of stimulus, though, because it obscures the actual purpose of that spending and makes it more of a political football than it ought to be.

    Really? That seems like a terrible way to make policy. In my experience, when somebody lists N reasons to do something and makes a convincing case that the action a good idea, the only people who break the reasons apart and point out that each one is individually insufficient are people who aren’t interested in making good policy. I think it’s hard to argue with:

    1) We need a bridge because ours fell into the sea.
    2) We can borrow the money at or near negative real interest rates, so it makes sense to borrow now.
    3) We have a lot of unemployed construction workers so building now would not steal construction resources from efficient private sector endeavors.
    4) Taking the “free” money in (2) would increase aggregate demand partially because we are demanding a bridge and partially because we’re moving money from low MPC people (savers) to high MPC people (the unemployed).

    Is any single reason good enough? Maybe not. Put together, they seem like an amazingly good case for building the bridge right now.

  9. James Hanley says:

    I argue that a magical, frictionless transfer of money from low MPC people to high MPC people would be stimulative on average.

    Yeah, but I’m still not persuaded that the assumption that people just stop spending makes anywhere near as much sense as the assumption that people don’t have money to spend. Stimulus doesn’t actually help the latter; it just shifts the spending around. I’ve found Scott Sumner pretty persuasive that the problem is in fact too little money, and that the zero lower bound issue is dramatically overblown and lacking any real empirical basis.

    The mechanism for this would be increasing interest rates. This is where Brad DeLong’s earlier post comes in. We’re in a state where that mechanism is extremely weak, so it’s pretty safe to assume that we’re not crowding out private loanable funds.
    Safe? That’s not what Sumner argues. He’s arguing that low interest rates are disguising a tight money policy. And he’s arguing for targeting NGDP growth. If he’s right, DeLong’s all wrong. If DeLong’s right, Sumner’s all wrong. Either way, I think it’s a stretch to say there are any safe assumptions here.

    As to your first analogy, it doesn’t necessarily work, because in the case of transferring it to government, they don’t necessarily spend it on anything that a private citizen would really prefer to spend it on (whereas if your dollar magically transfers to me, it’s still being spent on something a real live consumer really wants). I mean, at some point you have to agree that government spending lacks the efficiency of private spending or you will have to argue that a truly socialist economy would do as well as a free market economy.

    Your second analogy sounds persuasive, but is it really extensible? If we shifted from people making rational decisions about what to spend money on to people flipping coins to decide what to spend money on, do you really think things would work out as well?

    In my experience, when somebody lists N reasons to do something and makes a convincing case that the action a good idea, the only people who break the reasons apart and point out that each one is individually insufficient are people who aren’t interested in making good policy.

    I get what you’re saying, but I think you’re missing something. A) Some people will oppose it, so they’ll be looking for some argument to hang their hat on. B) Lots of people won’t know whether to oppose it or not, but will be moved by rhetoric. So, for example, adding as one of your justifications, “and it’s a good socialist project” could be a killer, no matter how true. “Stimulus” isn’t that scary a word, but it’s been given those overtones in the current climate. Even absent that, it can be awfully easy for opponents to move the middle by turning “essential investment in infrastructure” into “wasteful government stimulus.” For example, here in Michigan supporters of building a new bridge to Canada have tried to bolster their case by arguing that it will create construction jobs, to which their opponents have responded, “temporary jobs created by government spending.” Yeah, it’s a stupid response, but what’s at issue is not its stupidity but it’s effectiveness. Yeah, it may be a terrible way to make policy, but that is in fact how we do make policy in a democracy.

  10. Michael Heath says:

    James Hanley reports from his cite:

    Valerie Ramey of the University of California, San Diego … finds a multiplier between .8 and 1.2.

    His blog post title, Reason to Be Skeptical About Stimulus

    The whole quote is as follows:

    Valerie Ramey of the University of California, San Diego talks with EconTalk host Russ Roberts about the effect of government spending on output and employment. Ramey’s own work exploits the exogenous nature of wartime spending. She finds a multiplier between .8 and 1.2.

    It appears to me that the topic discussed was either government spending in general, or perhaps government wartime spending as well. I don’t see a reference to targeted stimulus spending on GDP in a recession. However I didn’t listen to an hour long talk. James, is your source this abstract or is the abstract incorrect in describing the interview where Dr. Ramey’s topic was fiscal policy effects on GDP in a recession?

    James concludes:

    So why aren’t they more conflicted about stimulus?

    Here’s a handful of major reasons:
    1) The country I’ve observed most devout in practicingKeynesian fiscal policy prescriptions was China, both in terms of running surpluses during the up cycle and aggressive stimulus when the economy started contracting. Their stimulus was incredibly successful at getting GDP growth back.

    2) We knew the U.S. stimulus would be less than optimal prior to its implementation for several reasons. Here’s a handful:
    a) We were late by at least a year. President Bush’s reaction was knowingly sub-optimal and insufficient.
    b) The amount was insignificant and slowly deployed.
    c) Over the years conservatives have been very successful at the state, county, and local levels passing policies which forces states to implement contractionary policies when state revenues contract. Their contractionary efforts countered whatever the feds did.
    d) Because our country is comprised of economic illiterates coupled to the political climate, especially what’s going on within the conservative movement and GOP, the Democrats allocated 37% of the stimulus plan to tax cuts which were less than or very close to 1.0.
    e) The initial federal stimulative efforts were mostly directed at counter-acting contractionary state policies.

    3) There are other studies which did focus on this recession which shows the federal stimulus yielded results in line with spending. Findings which held-up when broken down by the various components of the plan and their predicted stimulative effect. See the CBO reports, Blinder, and Zandi.

    4) The U.K.’s austerity reaction appears to have failed relative to our insufficient attempt at a stimulus. Cite: http://goo.gl/pRkPg [Disclaimer: I just found this result yesterday. I find it interesting though not convincing since I prefer seeing new findings debated for awhile prior to developing a conclusion of my own. I present it here anyway because I also I find the U.K. experiment interesting and analogous to the U.S. I.e., Can new-found fiscal discipline implemented in the down cycle restore aggregate demand by way of increased confidence by investors and consumers?]

  11. Matty says:

    While I am deeply worried about the austerity programme, mainly for reasons that aren’t strictly economic, I suspect there are enough other differences between the UK and US economies to make such a comparison difficult. Just to pick the first example that comes to hand we have a lot of trade with the euro area making British business more exposed to the drop in demand as those economies struggle.

  12. James Hanley says:

    I don’t see a reference to targeted stimulus spending on GDP in a recession. However I didn’t listen to an hour long talk. James, is your source this abstract or is the abstract incorrect in describing the interview where Dr. Ramey’s topic was fiscal policy effects on GDP in a recession?

    Eh, did you not notice this? She also discusses a survey looking at a wide range of estimates by others and finds that the estimates range from .5 to 2.0. That’s not just about war time spending. Anyway, you’d need to explain why wartime spending is less stimulative. Hell, because so much stuff gets blown up in wartime and we have to keep replacing it, maybe it’s more stimulative because it requires more spending. And “targeted” is a nice political word, but it assumes government is good at targeting. Color me doubtful, given how Congress manages its budgeting decisions.

    We knew the U.S. stimulus would be less than optimal prior to its implementation for several reasons. Here’s a handful:
    a) We were late by at least a year. President Bush’s reaction was knowingly sub-optimal and insufficient.
    b) The amount was insignificant and slowly deployed.

    Ugh, that’s so self-justifying for failure, but it doesn’t demonstrate a damn thing. It assumes bigger, timelier stimulus would have worked just fine, then “proves” the claim by pointing out how smaller later stimulus didn’t work. The logical error in that is just egregious.

    See the CBO reports, Blinder, and Zandi
    And see the critiques of those reports, rather than just cherry-picking them.

    You don’t explain what’s good about China’s stimulus. I’m willing to listen, but preemptively I’m going to note that there are criticisms of it, too. Such as “investing” in unneeded infrastructure and perhaps as much as 20% will have to be written off. And in fact their stimulus may have actually been monetary rather than fiscal.

    Look, I remain uncertain about this. I’m not going to crow that fiscal stimulus can’t work and that only monetary stimulus can (although I admittedly find Sumner persuasive). But I find this “well obviously it works and no sensible person could think otherwise” type comments to be somewhat lacking in seriousness, especially given the paucity of empirical evidence for stimulus.

  13. Troublesome Frog says:

    James Hanley:
    Thanks for spending the time on this. I’ve always enjoyed discussing these things with you. Also, sorry for the long post. I’m trying to edit, but there’s a lot in here.

    Yeah, but I’m still not persuaded that the assumption that people just stop spending makes anywhere near as much sense as the assumption that people don’t have money to spend.

    I find it persuasive because I’m one of those people. My wife and I are both employed and we have cash/low-yield “safe” assets that amount to years of average household income, but we’re not doing anything with it because of uncertainty over our relatively short time horizon. Of course, the two statements amount to the same thing. We have stopped spending and chosen to hold money. When a lot of people do this, it amounts to having too little money relative to money demand. I don’t see Keynes, Friedman, DeLong or Sumner disagreeing on this. When everybody decides to grab money simultaneously, you get a recession unless the money supply expands quickly enough to accommodate.

    Safe? That’s not what Sumner argues. He’s arguing that low interest rates are disguising a tight money policy. And he’s arguing for targeting NGDP growth. If he’s right, DeLong’s all wrong. If DeLong’s right, Sumner’s all wrong. Either way, I think it’s a stretch to say there are any safe assumptions here.

    Maybe I’m misreading one or both of them, but I think that they’re both right about what is happening and they agree that an increase in money and money-like assets (or an increase in inflation) is the answer. This is why I keep going back to DeLong’s old post, which basically shows the effect of government deficits in terms of monetary policy. The zero lower bound isn’t magical. It’s just that it makes the traditional trade of money for short-term government debt a nearly net-zero operation. For every dollar the Fed gives us, it takes out an equal amount of low-yield, zero-risk debt that looks a lot like a dollar of cash. This changes nobody’s behavior at the margin. The Fed either needs to buy something else that doesn’t look like money, or (per DeLong) the Treasury needs to create more short-term debt.

    I also don’t see major disagreement that NGDP targetting would have an effect. The question is how to make it happen. I can’t see the Fed causing much of anything by buying zero-rate debt unless the Treasury is creating more debt to replace it. The Fed could buy up all of the longer term treasuries, or maybe buy all of GM’s debt, or buy and retire all of the mortgages in the country. It could hire everybody to make pie for $1000 an hour. Somewhere before they bought all of the world’s assets, we would certainly get our inflation. It just gets the Fed into the business of doing things it hasn’t really done in the past. Having the Treasury create debt for the Fed to buy is something we all know a lot about.

    As to your first analogy, it doesn’t necessarily work, because in the case of transferring it to government, they don’t necessarily spend it on anything that a private citizen would really prefer to spend it on (whereas if your dollar magically transfers to me, it’s still being spent on something a real live consumer really wants).

    The question was why I think that unemployment and a zero interest rate are more important than the efficiency of the purchase. The transfer can increase aggregate demand by moving buying power from one MPC to another. The government taxing me and spending that dollar on “nothing” by just giving it to you can have a stimulative effect if you have a high MPC. So immediately we see that it’s not the act of taking a dollar and moving it elsewhere that introduces the inefficiency. So, to the source of the efficiency:

    If, instead of taking a dollar from me and giving it to you for nothing, it takes a dollar from me and says, “I’ll give you a dollar if you make me a pie,” we may get some inefficiency because your time is better spent using your advanced degree doing what you were trained for. The inefficiency comes from using a resource for something other than it would otherwise be used for. That’s why high unemployment and other underutilized resources (like auto assembly lines and construction equipment) makes stimulus more likely to be effective: the opportunity cost is, on average, lower than it is at full employment. If you were unemployed, being asked to make a pie wouldn’t be all that inefficient, and we’d get a pie out of the deal.

    Your second analogy sounds persuasive, but is it really extensible? If we shifted from people making rational decisions about what to spend money on to people flipping coins to decide what to spend money on, do you really think things would work out as well?

    The point wasn’t that it was good policy. The point was that the efficiency of your spending choice is less relevant than the MPC of the recipient of the money when the question is the multiplier for your dollar spent.

    As for the politics of it, I live in a world of spherical horses and frictionless axles. I’m most interested in what the best policy than in the one most likely to be politically possible. Things being what they are, it seems that very few sensible things are politically possible.

    Also, I am not a real economist. This is all vague memories of a BS in econ and readings of Keynes along with two hours of commuting think-time per day. Dr. Troublesome Frog is not a real doctor. Do not try his diet or let him touch your private areas.

  14. D. C. Sessions says:

    Ugh, that’s so self-justifying for failure, but it doesn’t demonstrate a damn thing. It assumes bigger, timelier stimulus would have worked just fine, then “proves” the claim by pointing out how smaller later stimulus didn’t work.

    Bear in mind that several economists (Romer and Krugman, to name two) called the outcome in advance. Their quantitative predictions were quite accurate given that the econometrics they had turned out to be quite a bit less grim than turned out to be the case.

  15. Michael Heath says:

    James response to me:

    Eh, did you not notice this? She also discusses a survey looking at a wide range of estimates by others and finds that the estimates range from .5 to 2.0. That’s not just about war time spending. Anyway, you’d need to explain why wartime spending is less stimulative. Hell, because so much stuff gets blown up in wartime and we have to keep replacing it, maybe it’s more stimulative because it requires more spending. And “targeted” is a nice political word, but it assumes government is good at targeting. Color me doubtful, given how Congress manages its budgeting decisions.

    You misunderstood by point about targeting. My context is specific to your blog post title, which is why I quoted it, not government spending in general which appears to me is what your citation is about. I used the term ‘target’ to refer to a specific government spending initiative serving as a reaction to a recession. My point had absolutely nothing to do with politics and everything to do with time and motivation – which was the quality of fiscal policy initiatives to defend against economic contraction.

    We already know government in general is not an efficient stimulator relative to the private sector. But we also know that it is more efficient in some markets and may be the only actor willing to play during some times. So the question in regards to your blog post title isn’t how stimulative government spending is in general or in a war, but instead how stimulative government spending is when the private sector retreats and whether it’s a response which effectively re-engages the private sector. I also threw in the U.K. performance since even if stimulative spending was effective, that alone isn’t an argument to deficit spend in a recession if a superior alternative existed.

    A part of my earlier response on why I concede Keynesian fiscal policies have and continue to work:

    We knew the U.S. stimulus would be less than optimal prior to its implementation for several reasons. Here’s a handful:
    a) We were late by at least a year. President Bush’s reaction was knowingly sub-optimal and insufficient.
    b) The amount was insignificant and slowly deployed. [more were presented]

    James response to this part which he quoted:

    Ugh, that’s so self-justifying for failure, but it doesn’t demonstrate a damn thing. It assumes bigger, timelier stimulus would have worked just fine, then “proves” the claim by pointing out how smaller later stimulus didn’t work. The logical error in that is just egregious.

    It would be if our stimulus had failed and I avoided this fact, but I didn’t I pointed out it did in fact work. Re my noting which you ignore here when responding to these two points, the CBO, Blinder, and Zandi found the stimulative plan worked consistent with the predicted stimulative effects of its spending and tax cuts components in terms of a corresponding rise in both GDP and job growth. So you may disagree with my premise but I certainly didn’t commit a logical error, not even close let alone egregious. These are facts where I think they matter a hell of a lot when we’re confronted with another slow-down within the context of my directly responding to the last question of your post on why I continue to defend Keynesian economic policy.

    Me earlier:

    See the CBO reports, Blinder, and Zandi

    James response:

    And see the critiques of those reports, rather than just cherry-picking them.

    I read all three (CBO in Aug-2010) and don’t even quote them here so how am I cherry-picking them? Instead I responded with my understanding of their summated conclusions, no bits reported here by me at all. Strange reply on your part James.

    I haven’t seen any credible critiques of these reports. So perhaps I am insufficiently informed on the quality of their findings. I didn’t find them provocative but instead report what was predicted which is consistent with other efforts made in the past, not just in this country but others as well. E.g., the stimulative effect of the ’09 tax cuts were consistent with the results when Carter tried it as did W. Bush early in his first term.

    I am aware of some critiques regarding China’s policies, some which I obviously find compelling given their inflation rate. However I don’t find their supposed policy errors all that important relative to their results and relative to their predicament at that time. I’d be shocked* if any government did a perfect job executing any complex policy and therefore don’t judge government performance from that perspective. It’s been a long time since I took econ so maybe my memory’s foggy but it seems to me it’s a hell of lot better fighting inflation after a recovery than failing to extract oneself from a down-turn.

    *Actually I am shocked at the good quality of execution of the last U.S. stimulus. The policy itself was at best mediocre but the transparency, direction of funds as intended, actual vs. budgeted costs, and completion of projects started has been generally successful.

  16. James Hanley says:

    I read all three (CBO in Aug-2010) and don’t even quote them here so how am I cherry-picking them?

    Because you ignored the criticisms of those. If the critics are right, then, no, the stimulus policy did not work. My point is that the evidence is still out on whether it worked or not. Pointing to one study that says it worked is not sufficient when there are other studies claiming to report flaws in the first study.

  17. James Hanley says:

    T-Frog,

    My apologies, but I’m a bit under the weather, and my head’s not clear enough to think about your post right now, given that it’s fairly deep.

  18. James Hanley says:

    P.S. Here’s how I see the logic of Krugman’s not-enough-stimulus argument. It’s like telling a sick patient that he needs to take 10 doses of an experimental drug. He only takes 5 and is still ill. Then the researcher says, “see, he didn’t take enough–that proves the drug would have worked if he’d taken all 10 doses.”

  19. Dr X says:

    Here’s how I see the logic of Krugman’s not-enough-stimulus argument. It’s like telling a sick patient that he needs to take 10 doses of an experimental drug. He only takes 5 and is still ill. Then the researcher says, “see, he didn’t take enough–that proves the drug would have worked if he’d taken all 10 doses.”

    Or

    I told you, in advance, that this would be inadequate, so don’t point to the results and say: “Ha! Stimulus doesn’t work.”

    I’m not saying that you’re saying that, James, but many people are.

    If the physician explains in advance that you must stay on the antibiotic for 10 days, and you return 10 days later and say: Doctor, I took the antibiotic for 3 days and the infection didn’t clear up; antibiotics don’t work, the doctor would be right to respond, of course the antibiotic didn’t work. You didn’t follow the prescription.

    This doesn’t make Krugman correct in his prediction about the effect of a larger stimulus, but is he supposed to pretend that his recommendation was followed and therefore he’s been proven wrong, as many opponents of stimulus have actually claimed? I don’t think so. It’s not his job to affirm their misrepresentations of hypothesis. He said the stimulus needed to be much larger and that’s still his position. Somehow his being right insofar as his hypothesis has been tested is being turned by many opponents of stimulus into Krugman being wrong.

  20. Kevin Donoghue says:

    “If the multiplier is below 1.0, there’s no argument’ let’s avoid stimulus like the plague.”

    You need to do some unpacking there, I think. If the export multiplier was 0.7, would you feel inclined to put curbs on exports? I presume not. Now you may have good reasons to like exporters and dislike government, but your argument is incomplete unless you spell out those reasons. 1.0 is just another number, it’s no more special than 1.23. Karl Smith has a sensible respone to Russ Roberts here:

    http://modeledbehavior.com/2011/10/12/keynesianism-multipliers-and-stimulus/

  21. Sorry for the extensive quote, but it’s straight on-target.

    Karl Smith:

    Keynesianism, Multipliers and Stimulus
    Wednesday ~ October 12th, 2011 in Economics | by Karl Smith
    Lots of folks are piling on Russ Roberts but I want to focus on one thing he says here

    I recently interviewed Valerie Ramey on the multiplier. In her work, the multiplier ranges from .8 to 1.2. A multiplier of 1 means there is no stimulus from the spending–GDP rises by the amount of the increase in G but by no more. Private spending doesn’t grow. When you include the taxes (either today or in the future to pay back debt from the increase in spending) that finance the government spending, it’s particularly costly.

    I think this is substantively wrong. A multiplier of zero means there is no stimulus.

    It might be easier to dispense with the term stimulus to see why.

    A multiplier of zero implies complete crowding out of the private sector by government. The government spends more and the private sector offsets this exactly by spending less.

    A multiplier between zero and one means that there is some crowding out by the private sector. The government spends more, the private sector spends less but not enough to offset government spending.

    A multiplier of 1 would suggest no crowding out. The government spends more and private spending is unaffected on net.

    A multiplier above one would suggest crowding-in. The government spends more and this increases private as well as public spending.

    So now suppose the multiplier was one and invariant to the level of spending. Then the government could choose whatever level of unemployment it wanted without reducing the production of private goods.

    That would mean that for example the government could have lowered the unemployment rate from in 2009 by two or three percentage points had it chosen a large enough expenditure [Added: with no loss of private production. This is vastly strong claim than simply we could assist in easing the pain]

    As for the total cost, it depends in large part on whether you place any value on those expenditures. So if we are buying roads and schools, you may not believe that are as useful as private construction spending might be, but are they 100% useless? If they are even only 50% useless then you have come out ahead.

    How the borrowed funds affect the long run economy is complex and depends on what the monetary response is, in large part. However, note that the government does not ever have to pay the bonds back and in general does not even have to tax the population in order to service them.

    That’s because when the recovery hits the interest rate on t-bills is less that than nominal growth rate.

    So even rolling in the interest payments continually will result in a declining debt-to-GDP ratio. Eventually the debt will just shrink away.

    A multiplier above one suggests that when the government spends more we get more government goods AND more private sector goods. There is crowding-in.

    Here it makes sense to do spending even if the project you are spending on is worthless. This is digging ditches just to fill them back up again territory.

    Now that having been said I think people worry too much about the multiplier.

    I heard lots of people suggesting that we should do direct spending over tax cuts because the multiplier is bigger, but this is not the right way to think about it.

    You have to consider the logistics of actually spending this money and the effect that its going to have on the economy.

    If you have some project sitting on the shelf that you wanted to do anyway and it has lots of return then you should do that. Trying to assemble projects is going to be more difficult and the side-benefits are going to be less. Plus if the only reason you are doing this project is to lower unemployment then you have to consider the distorting effects it has on the economy.

    One great thing about tax cuts is that they are really easy to implement and we don’t have to worry about trying to direct resources to some end.

    People can direct resources to whatever end they want. We also get the benefit of being able to do a broad based stimulus that directly benefits lots of people at once.

  22. Oops – didn’t get all the way to the bottom of the previous comments to see that Kevin had linked the exact same Karl smith post.

Comments are closed.