Can’t Get No Satisfaction

August 1, 2009

Mary Katharine Ham, writing for the Weekly Standard, compares the Obama administration’s Cash for Clunkers program to “the KFC grilled-chicken giveaway.” And that’s by far not the most troubling thing about her post. Rather, the low point – indeed, the wide canyon – comes when Ham criticizes Cash for Clunkers for its success. The government actually got people to buy new cars? that are more fuel efficient? The horror!

Many Obama-supporters on Twitter* today have argued that the program is only a failure insomuch as it is a great success. You see, the Obama administration simply revealed the tremendous demand for $4,500 hand-outs fuel-efficient cars, and should be congratulated for that. The utter lack of competence, planning, or understanding of incentives is not an indication of the federal government’s unsuitability to mucking around in the private sector, but a reason to invite more mucking.

Here’s the deal. Conservatives have long argued that in this recession, no amount of government spending will have a lasting effect until consumer spending and production rebound. This critique has all but ruled out any government program as successfully stimulating consumer spending and the private production of goods and services. For example, take this Forbes piece from last February:

If the Bush spending plan can’t productively stimulate the economy, what government economic plan can? None. Production does not need stimulation from the government; it needs liberation from the government . What a productive, dynamic economy requires of a government is that it restrict itself to protecting property rights from force and fraud, and refrain from interfering in free production and trade.

In a time when consumer spending is beginning to seem more like playing financial Russian Roulette than fulfilling a civic duty, this critique is implausible. If consumers are holding off on consuming, no sane producer would fail to roll back production. You can’t force demand for something by merely making it – witness the unsold inventories of retail goods that have piled up as this recession dragged on.

But you can force consumer demand by subsidizing it. And that’s what we’re seeing with the Cash for Clunkers program. When people are paid to buy new cars, they will buy new cars. There is no way to get around this – unlike the Arizona incentive for alternative-fuel vehicles Ham mentions, which did not stipulate that owners actually use alternative fuels. Every check cut under the Cash for Clunkers program stands for a new car purchased: no exceptions. And as demand for new cars increases, the production of cars – particularly their production in the U.S. by American workers – will be bolstered by this government action. Under the paradigm of conservative economics, we are witnessing nothing short of the impossible.

* Ham cites Twitter as a source not once, twice, or even thrice, but four times in a short post. In Twitter veritas.


10 Responses to “Can’t Get No Satisfaction”

  1. Sarah Says:

    Okay, I think you’re mistaken here.

    The “conservative” argument about consumer demand (actually not all conservatives believe it; Martin Feldstein disagrees) is based on an economic model that originates with Milton Friedman. The idea is that people smooth consumption throughout their lives; that is, they spend pretty much the same amount over time, borrowing in bad times and saving in good times. You can actually prove that (assuming the relevant functions are well-behaved) this is the smartest way to spend money; it makes you better off, in terms of lifetime utility, than splurging every time you get a paycheck. So many economists, and the political wonks who follow them, assume that people are “smart” in this way.

    Now, under this assumption, a one-time government transfer can’t make people spend more. They’ll just spend as they always do, and save any extra left over. Even an in-kind transfer like this, that subsidizes a certain kind of car, shouldn’t increase total consumption; people will just spend less on everything else. The Cash for Clunkers program may be excellent at getting people to buy new cars, but it doesn’t follow that it increases consumer demand on the whole.

    Now, I have seen some evidence that people don’t smooth their consumption perfectly; I don’t think the Friedman model entirely represents how people behave. This is a very difficult and far from resolved problem, and it’s unfortunate that people stick to their guns so ideologically on what is actually an empirical question.

    This discussion also leaves aside the question of whether we SHOULD spend more than we do now on fuel-efficient cars, and whether the government can be involved in that decision. I tend to think that attempts to cherry-pick one industry or another to subsidize won’t lead to the best possible outcome. Even if we agree that we ought to point the economy in a greener direction, no government official knows for certain which technologies represent the best, most efficient, way to decrease carbon emissions.

  2. grandmute Says:

    They’ll just spend as they always do, and save any extra left over. Even an in-kind transfer like this, that subsidizes a certain kind of car, shouldn’t increase total consumption; people will just spend less on everything else. The Cash for Clunkers program may be excellent at getting people to buy new cars, but it doesn’t follow that it increases consumer demand on the whole.

    I disagree that financial incentives cannot be used to induce people to spend more – in fact, the entire craft of marketing stands behind my dissent. No one really plans out their consumption over their entire expected lifetime, and it is doubtful that people even behave as if they were doing so.

    Consumer demand, at the micro level, is extremely malleable. Think of it this way: why do you find bales of coupons in the Sunday paper, rather than the amount of cash retailers expect you to save by using the coupons? If people consumed rationally, they would behave the same whether they were offered a discount or whether they received “free money.”

    Of course, people do not behave that way. They are much more likely to put away actual “free money” than they are to use a discount and also reduce the rest of their spending accordingly. That’s why it makes sense for retailers to offer coupons, rebates, discounts, and so forth – all these instruments subtly and not-so-subtly encourage people to spend more than they had intended or anticipated.

    Now, back to the national scale. In a recession, it is indisputable that consumer demand will falter. The government could cut checks to taxpayers and call it a day (e.g. the Bush tax cuts) – but, as you’ve noted, this offers no guarantee that people will actually go out and spend that much more money. Cash for Clunkers gets around this difficulty by not only increasing consumers’ spending power, but actually persuading them to use it. Psychologically, it takes advantage of several powerful impulses: 1. Getting something for nothing, 2. Getting something brand new, 3. Getting a discount, and more recently, 4. Getting something which may run out.

    Given this bundle of incentives (not to mention $3500-$4500) consumers who take advantage of the program probably do not coolly and rationally throttle back the rest of their spending.

  3. Sarah Says:

    Well, that sounds reasonable, and I believe that’s partly true. There are two extremes (a windfall grant will not increase consumer spending at all, or a grant will increase consumer spending by the full amount of the grant) and “both” economic positions can be correct in the sense that the grant boosts spending, but not enough.

    consumer spending did not rise in the last quarter, because the savings rate went up. That’s either an argument for the ineffectiveness of grants for increasing spending, or an argument that not enough grants were given, depending on your political stripes.

    What I was saying is that we don’t have any evidence that the car grants worked at *what they were designed to do,* which was boost consumer spending. The mere fact that they sold a lot of cars is not proof of success.

  4. grandmute Says:

    I don’t think the Friedman model entirely represents how people behave. This is a very difficult and far from resolved problem, and it’s unfortunate that people stick to their guns so ideologically on what is actually an empirical question.

    This merits its own reply: the admission that people just might not be perfectly rational actors is a great understatement. Oh, individuals might be capable of some decision processes that we might like to describe as rational, but the fantasy of homo economicus is just that – a fantasy.

    It is not a matter of ideology to reject an economics based in complex manipulations of the same unproven statement equating human behavior with rationality. The empirical question of whether the axiom of perfect rationality is true has long been resolved in the negative. The real empirical question now is: to what degree is the behavior of actual people in real situations (quasi-)rational? It is certainly not entirely rational, probably not completely irrational, but undulating in between these theoretical extremes.

  5. grandmute Says:

    Correct me if I’m wrong, but are we not just two weeks into the program? Boosting consumer spending might take a bit longer than that, whatever we choose to believe about the feasibility of this project.

  6. Sarah Says:

    Oops, didn’t realize that it was so new. In that case, for the incentive reasons you mentioned, it seems like it would be better at increasing spending than tax cuts.

    I’m going to defend the usual practice of economics for a second here.

    Okay, people are not Homo economicus. That’s established. There have been lab experiments that demonstrate various psychological biases. Historical consumption patterns show that people do, in fact, spend more when they have more disposable income (which implies that they don’t smooth consumption perfectly.)

    What’s not known nearly as well is how to model behavior once we account for irrationality. What you’re doing now is making an economic hypothesis: you’re saying that giving an in-kind discount will induce people to spend more total, because of “several powerful impulses: 1. Getting something for nothing, 2. Getting something brand new, 3. Getting a discount, and more recently, 4. Getting something which may run out.” But if we wanted to be quantitative about this, we’d need functions that described the effects of these impulses. We’d need data on just how much people depart from rationality. We’d need economists to start talking to marketers and psychologists. We’d need to build a whole new science. I think we actually need that whole new science, and it’s already underway (especially in behavioral finance.)

    What I often see from liberals who are not economists, though, is an argument like this:
    1. People aren’t rational, so such-and-such economic claim is based on false premises.
    2. Instead, here’s what the real result of my policy will be.

    Between 1. and 2. there’s a great unknown. The classical economic claim in 1. is usually the solution to a system of differential equations. Add in behavioral factors, and you’d need a new set of equations if you want to make predictions with the same rigor, and you have no idea whether these more complicated equations are at all tractable. What I’m saying is, if an economist makes a rigorous claim based on (very) oversimplified assumptions, then even if you correctly critique the assumptions, you don’t have an equally justifiable competing hypothesis. Claiming, intuitively, “Human behavior is like this so the results will be these” is trying to do economics while skipping the economics.

    Now, there is a rigorous framework for explaining how people behave if they consume according to their current disposable income. It’s called Keynesianism. And it *is* equipped to make predictions. And there are great economists who subscribe to it. (Any belief which requires you to believe that Paul Krugman is an idiot is probably a false belief.)

    My objection is to arguments that take the “people are irrational” shortcut in the cases where we don’t have a model to replace the classical one. I don’t think we know very much about how your “powerful impulses” work; simply knowing that they’re there tells us no more than that classical models are probably inadequate.

  7. grandmute Says:

    Ah, “rigor” – the bugaboo that economists always try to foist upon their fellow social scientists. As they might understand it, the way to truth is paved with equation after rigorous equation, and God help the poor soul who essays a prediction about the state of human affairs without expressing it neatly in a formula or a graph.

    I am not advocating the polar opposite – that any prediction be entertained, no matter how implausible. But predictions can be reasonable and well-grounded in known truths without requiring the formal “rigor” of economics papers.

    Your defense of economics-as-usual is hubristic in its claim that absent “a new set of equations, [one cannot] make predictions with the same rigor” as the classical model. Rigor is not an end in itself. If the classical model cannot accommodate reality, the solution may not necessarily entail tweaking the model, but may require, for the time being, giving up our dreams of accurately, quantitatively predicting human behavior.

  8. Sarah Says:

    Well, I’m an (aspiring) mathematician, not an economist, so I’ll turn purple and squirt smoke out my ears if you say rigor isn’t an end in itself.

    All joking aside, I think you’re right that not very much is known about human behavior. I still think that policies need to be judged in quantitative terms. In the case of a stimulus package, “Is this a good bill?” is (almost) equivalent to “Will this help end the recession?” By its nature, that’s a quantitative question, and it needs to be answered immediately — we don’t have the luxury of giving up on predicting human behavior, and we’d like our predictions to be reasonably accurate! Give up on models, and we don’t have a CBO or an OMB to give us cost projections, we don’t have a CEA to estimate the macro effects of a policy — we have nothing but competing arguments. Sound and fury. I don’t want to paint economists as all-knowing; I’m just saying it’s of great practical importance to know at least a little.

    In the case of the car discount, it shouldn’t be impossible to quantify the kinds of effects you’re talking about. This isn’t the first time an administration has tried to stimulate demand in a recession. We can estimate the “coupon effect” by comparing the effect on consumer demand of past in-kind grants versus cash transfers. We can then think of the effect of Cash for Clunkers as the effect of a cash transfer plus the coupon effect.

    Unless somebody does even a rough estimate, all you’re saying is that the coupon effect is non-negative. Great, but we care how big it is — in particular, we care whether the positive effect on the economy is large enough to justify the cost in taxes (presumably we’d like the project to be a good deal.)

  9. grandmute Says:

    I agree that policies should be judged quantitatively based on performance. Today, this is fairly straightforward when the policy has already been enacted, preferably in a natural experiment situation. It also helps if enough time has elapsed for good data collection.

    But we must be extremely humble in making the quantitative predictions that we do. When the data are all in, the “sound and fury” give way to accurate analysis. On the other hand, looking forward, while we can and should attempt to estimate quantitative policy effects, we should also be candid about the limitations of economic methods.

    In that sense, I would argue that we should not, within reason, give much thought to cost-benefit estimates for urgent policies. To wit, if something looks like a good policy for an economy in recession, and it seems to have some positive effect, then we cannot afford to not enact this policy on the fear that, by our rigorous guesstimate, the positive effect is outweighed by the cost in taxes.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: