Posted by Brad @ 6:26 pm on July 31st 2009

Cash for Clunkers: A Stimulus That Worked?

Hard to say. But, the U.S. government offered $1b in $3500 vouchers to anybody trading in an older car and trading up for a newer, more efficient one.

In 5 days, 300,000 people did so and the program already ran through the billion.

I’m not quite sure how that compares to a regular week of auto trade-ins. Clearly, you can’t say the government caused 300,000 people to buy new cars. But, it certainly appears to have had a dual effect of stimulating auto sales (and boy did that industry need the boost), as well as at least theoretically a marginal environmental impact (in that that’s 300,000 people on the road in vehicles with improved fuel efficiency). And, if the government feels the need to pump in money to get people to make better environmental choices, and to pump in money to get people buying again, certainly this program appears to have done both pretty directly. It was, one would have to say, a pretty decent success.

Brad Plumer notes the success, and also notes that if the program is this popular, Congress maybe ought to not only re-up it, but to strengthen the fuel efficiency requirements with it (which makes sense to me):

On the other hand, the program certainly offered a much-needed jolt to the economy, and can provide a huge boon to the ailing auto industry—$1 billion to spur the purchase of 300,000 new vehicles in five days has to rank as one of the more successful stimulus programs to date. (I imagine the program is mostly just moving up purchases that would’ve happened anyway, but in a recession, that’s a good thing!) Still, if the program’s so popular, and everyone’s lining up to trade in their old clunkers, then if Congress decides to re-up, it may as well ratchet up the fuel-economy requirements for new cars and get an even bigger benefit out of this thing.

I have to say, if we are going to drop government money into stimulating the economy, the direct route such as this seems to me to be the least obnoxious way to do it. I certainly prefer it to TARP.

In related news, the Economic Policy Institute released their appraisal of the economy on Day 500 of the recession, and specifically if the American Recovery and Reinvestment Act has had any effect. Their assessment, which concurs with what most people seem to be saying, is that yes, it has, but not much. Adding: “not much”, on a macroeconomic level, does not mean “insignificant”. Their bottom line:

Despite the overall contraction, the fingerprints of the American Recovery and Reinvestment Act could be seen in some aspect of today’s report. Federal government spending grew at an 11% rate in the quarter, adding roughly 0.8% to overall GDP. State and local government spending grew at a 2.4% annual rate, the fastest growth since the middle of 2007. It is clear that the large amount of state aid contained in the ARRA made this growth possible.

Furthermore, real (inflation-adjusted) disposable personal income rose by 3.2% in the quarter, after rising by only 1% in the previous quarter. A large contribution to this increase was made by the Making Work Pay tax credit passed in conjunction with the ARRA, as this was the first full quarter that the credit was in effect. Inflation-adjusted transfer payments (including a one-time payment to Social Security recipients) rose at an annual rate of over 6% in the quarter as well.

This increase in disposable personal incomes did not translate into a sharp boost in consumption spending because the personal savings rate jumped again — rising to 5.2% in the second quarter, up from 4% in the previous quarter. This slippage between personal incomes and consumption spending caused by a rising savings rate makes plain that, instead of focusing on even more tax cuts, it was wise to make sure that much of the ARRA was devoted to direct public investment spending. The public investment spending in the ARRA, while not having a significant impact in the second quarter, will provide an even stronger boost to the economy in quarters to come.

The consensus of macroeconomic forecasters is that ARRA contributed roughly 3% to annualized growth rates in the second quarter. This means that absent its effects, economic performance would have resembled that of the previous three quarters, when the economy contracted at an average annual rate of 4.9%. In short, the recovery act turned this quarter’s economic performance from disastrous to merely bad.


  1. Doug Mataconis tries to give a Liberty Papers critique of the program, and is frankly pretty unconvincing. Of course it didn’t create auto demand, it just tipped a lot of people into doing it now rather than later. But if your intent is stimulus during a recession and the added environmental benefit of not having those 300k people wait a few years, his point there isn’t really a critique of the program, but rather is one of the explicit advantages of it by design.

    The only new point he adds is that it’s akin to stimulating an “auto bubble”, and that is something to bear looking at in the future. I don’t think that’s true for this billion dollars, but of course for the next billion, auto dealers may just start reducing the price they pay for trade ins, or increasing sticker prices or whatever, meaning all the government will in the long run do is just raise the cost of auto trade ins by $3500 for no good reason. And that would indeed probably be the long term effect of the program, if it existed into perpetuity.

    I get what he’s saying, and he’s not wrong in the abstract, but in the practical, it seems to me this billion dollars did more or less what it was intended to do, with no unintended consequences that I can see.

    My point isn’t that this justifies government spending, but if you’re going to do that for the express intent of stimulating the economy and adding more of a short-term financial benefit to long-term beneficial thinking, I’d say the Cash for Clunkers thing indicates that at the very least, the direct approach is best. Instead of creating massive new socialized suprastructures (ala TARP), or subsidizing industry that can’t meet consumer needs effectively, why not just plop the money down directly to consumers so they can buy shit? It’s really not very far removed from targeted tax credits, save that it’s explicitly geared to stimulate spending rather than, say, saving. Which, when there’s an economic need in the short term, for an industry on the ropes, and with an added side benefit, seems to me to be a pretty decent way to do it. Beats buying car companies or going hog wild on fuel standards and consumer choice, anyway.

    Comment by Brad — 7/31/2009 @ 7:00 pm

  2. You know what would be even more successful, by these terms? If the government were to just purchase all the cars outright and give them to people. Money spent! Milage improved! Manufacturers rewarded! Everybody wins!

    This is a success by the terms of the stimulus package. A “failure” by the terms of the stimulus package is the government giving out a whole bunch of money that just disappears into a hole somewhere without producing measurable economic activity. It is the measure of the catastrophe we are creating that cash-for-clunkers, in which we can actually SEE the flames from the burning piles of cash, is by that standard a success.

    We occasionally speculate, at this blog, on what judgments future generations will make of current norms. A prediction along those lines: the discussion of the successes and failures of this government approach is going to ASTOUND our debt-crushed grandkids. “Ah, so THAT’S where my paycheck today is going…into a successful 2009 program, where it went up in smoke in a week in order to upgrade my grandpa’s chevy. But hey, it did keep Pontiac from going bankrupt until 2014, so I suppose it was worth it. Well, back to the salt mine.”

    Comment by Rojas — 7/31/2009 @ 7:04 pm

  3. You know what would be even more successful, by these terms? If the government were to just purchase all the cars outright and give them to people.

    No. The success here is not that the cars were sold (as you say, if that were the metric, the government could just buy cars, which they pretty much did with the auto bailouts), but that people bought them. Presumably, the cars cost more than the vouchers. The success of the stimulus isn’t that cars were bought full stop. It was that targeted government intervention in this case tipped the scales towards people, at worst, making that economic decision earlier than they might have otherwise (when the economy, in the macro sense, could most use it). It was…well, stimulating market activity. Not creating it, or supplanting it, but stimulating it.

    So, the government uses one billion dollars, probably (but not necessarily, I’ll grant you) creates two or three or five times that amount in market activity, and does so along a line steered by the government towards more macro interests (in this case, 1. creating spending in a bad economy suffering from a lack of both it and confidence, 2. bolstering an industry that is in trouble, and 3. some marginal environmental effect maybe). Your point is taken on the obnoxiousness of precisely any government spending to an economic end, but I think you’re shadow boxing a bit as well with that one.

    Comment by Brad — 7/31/2009 @ 10:21 pm

  4. Your arguments is that the multiplier effect–the economic activity generated–occurs as a result of the money being spent at the dealership.

    Explain to me again why having the government buy the cars outright wouldn’t do exactly the same thing.

    Same amount of money spent. Same recipient. Same car-switch decision moved up by the same amount of time. Yes, the consumer keeps more of their own money, but that’s irrelevant in terms of your multiplier effect; the dealers can spend the government’s cash as easily as the consumer’s.

    I do not understand the economic side of your argument at all. Economically I can find no distinction between this program and any other kind of government expenditure, except for the recipient.

    Comment by Rojas — 8/1/2009 @ 12:20 am

  5. I don’t really need a new car. But I think it would be equally stimulating if the government would just buy me a whole new hi-def home theater and tweaked out audiophile sound system.

    Maybe that is asking a bit much.

    I’ll just be happy if they give me %4500 trade-in credit for my old energy inefficient stereo and video setup against a new system. It would lower my carbon footprint, certainly help stimulate the economy and in particular the beleaguered retail and electronics sector. Circuit City is already gone. Shouldn’t the government do everything it can to save jobs at Best Buy?

    And now that I think about it… I am pretty sure that my wine cellar is filled with wine that was not produced with the most advanced organic, water conserving, locally grown grapes. The domestic vineyards are really hurting these days also. We could solve a lot of problems if the government would give me $4500 to trade-up. I could do some serious stimulating in Napa over the weekend.

    Comment by mw — 8/1/2009 @ 1:42 am

  6. In fairness to Brad, he’s making the argument that this stimulus appears to have achieved its purpose IF you buy into the line that government stimulation of the economy is desirable and sound policy.

    The rest of us seem ill-convinced of that assumption, given the costs involved.

    Comment by Rojas — 8/1/2009 @ 1:57 am

  7. Positive Liberty’s headline is unimprovable.

    Comment by Rojas — 8/1/2009 @ 2:04 am

  8. Yes, that’s the argument I’m making.

    Another point here: you’re taking the stimulus effect purely on the side of the car dealerships. I’m making it more along the lines of the consumer themselves.

    Let’s take the assumption that consumers choosing to spend 3 billions dollars in buying cars is worth twice as much, in terms of the macro-economic effects, if they do it now versus in two or three years. The government is spending 1 billion dollars, then, to spur consumers to spend 3 billion dollars at a time when that spending might account for 6% of economic growth vs. 3% (obviously, just pulling numbers out of my ass). Add to the psychological effects which are murky but very very real (and important), and I think you could make a pretty decent argument that it’s a sound 1 billion dollar investment. Even if you are against all government stimulation of the economy (which is where most commentators here will be coming from), it must surely be at least preferable that that stimulation has a disproportionately positive effect (one billion spent for six billion worth of economic activity), that that stimulation comes not by government fiat but in the most literal sense of causing a desirable economic behavior (i.e. it is ultimately rooted in consumer choice, at least for those buying cars), that it is a direct (A to B) stimulation and one that is at least in the ballpark of being measurable and transparent, and that the side effects, where they exist, be either marginal or even positive (such as the environmental impact here).

    And the other spectrum along which this sort of thing is important is psychological. A lot of political thinkers don’t like to go there because it’s impossible to measure, but that doesn’t mean it’s not actually very, very important. Getting millions of people (far more than will actually receive a government voucher) re-interested in buying new cars, getting some consumer confidence back, getting some markets moving again, is something that, in a capital R Recession situation, is probably easier to underestimate than overestimate. This is, btw, where Doug’s critique in my first comment is most germane, in that it’s very easy to overstep that (as in the case of the housing markets the last decade (probably two). But in a limited and targeted example like this, the psychological stimulation is almost certainly more good than bad for all involved.

    It’s the same reason I can often be swayed towards, say, tax credits that try to monetize otherwise abstract and long term environmental or moral choices (and when I say moral choices, I mean, for instance, tax write-offs for philanthropic gifts), or the government investing and subsidizing pure research in science, etc. etc. It is not very libertarian of me, admittedly, but I do think there are targeted ways that the government can intervene in markets that, while not laissez faire, I can be convinced of the practical wisdom of.

    Here, I am not making the argument that government spending to stimulate the economy is an inherently desirable and sound policy, because I don’t really believe that to be true (though I will cop to not being able to dismiss it outright). I am making the argument that, if the government is going to engage in this kind of thing, I would rather they do it in this sort of way than in most any other.

    Comment by Brad — 8/1/2009 @ 2:18 am

  9. Oh, and one more thing. Lest people think the purpose of the post above was to defend efforts to stimulate the economy, my first thought when I was writing the post above wasn’t that per se, but rather was contrasting something simple, transparent, direct, and more or less effective like this Cash for Clunkers thing, to TARP, the auto bailouts, the Fed policy, and the like.

    Ironically actually, Rojas’ suggestion of the government just buying all the cars isn’t far from what they actually tried to do (for cars, for bank assets, for mortgages, etc.). And, ironically, I think on a dollar for dollar basis that proved far LESS effective (and certainly a helluva lot more effective and prone to outright graft) than this little Cash for Clunkers thing.

    Comment by Brad — 8/1/2009 @ 2:26 am

  10. I think I’m pretty much with Brad here. That is, if the government is going to spend $x billion dollars in an attempt to increase AD, I’d at least prefer they did so as efficiently as possible.
    I’d say the big difference between the ‘Rojas’ plan (government buys 100% of car) and the ‘Clunker’ plan, is not that one is better at ‘stimulating’ (as Rojas says, they’re equally good), but that the latter is more efficient.
    I’d say the flaw in Rojas’s ‘4’, is the following:

    “Explain to me again why having the government buy the cars outright wouldn’t do exactly the same thing…. Same recipient. Same car-switch decision moved up by the same amount of time.”

    I don’t think this is correct. If the government offered me $3000 for my ’97 Ford Contour on the requirement that I buy a newer car, I wouldn’t accept. On the other hand, if they offered me a 2010 Ford Taurus in exchange, I would obviously jump at the offer.
    Similarly with mw, if the government offered me some sort of wine discount, it wouldn’t change my behavior at all, but if they gave me free wine, I’d take it (and then give it to my parents or something).
    The Cash for Clunkers seems to be working on the margins to influence decisions, which is good, given you accept the idea that government should be attempting to increase AD.

    Comment by Redland Jack — 8/1/2009 @ 12:26 pm

  11. FYI

    The House voted 316-109 to approve the transfer of $2 billion in emergency funding from the $787 billion economic-stimulus plan to the “Cash for Clunkers” program, ensuring it has sufficient funds to continue. The Senate won’t consider an extension until next week. At a news conference Friday afternoon, President Obama praised lawmakers for moving quickly to rescue the popular program, which he said has succeeded beyond expectations.

    The move follows a scramble Thursday after news emerged that the initial $1 billion in funding may have been close to exhausted after just one week. The National Auto Dealers Association said it was given “specific assurances” by the Obama administration that all deals secured on Friday will be honored, though the NADA’s chairman reiterated concerns that some dealers might not be reimbursed for rebates extended to customer.

    Comment by Brad — 8/1/2009 @ 4:18 pm

  12. By the way, one new libertarian argument here that I’d be remiss to not pass on: Daniel McAdams pointing out how this program hurts the poor.

    Comment by Brad — 8/1/2009 @ 4:18 pm

  13. This program makes *some* economic sense if you assume the following:

    a) you would pay unemployed people something anyway via their unemployment benefits. You would be getting zero productivity for those dollars.

    b) Some substantial portion of the dollars spent go to domestic jobs.

    c) using less gas is a good thing overall.

    If you make those assumptions, the government

    1) takes the same money that would be used to pay people to be idle and it uses it to get new cars produced instead,

    2) it gets customers into new cars that will last longer than the previous ‘old’ cars; and

    3) reduces the gas purchased for a given number of miles driven by having more fuel efficient cars .

    So, while I don’t like government spending in general, I prefer this program to a lot of the other government spending such as stimulus programs, bail outs, wars and foreign aid that we currently waste dollars on.

    It would make more sense to limit the program to domestically made cars (which would include certain Honda, Nissan and Toyota vehicles, but that might cause GTO problems…)

    Comment by daveg — 8/2/2009 @ 10:50 pm

RSS feed for comments on this post.

Leave a comment

You must be logged in to post a comment.