…because I’m an economist and a mom–that’s why!

Using Sticker Shock to Clarify the Costs of Deficit Spending

May 13th, 2010 . by economistmom

I attended the National Tax Association’s spring symposium today, and heard a fascinating presentation by Raj Chetty, a “wunderkind” economics professor from Harvard, on using “behavioral economics” (sometimes now referred to as “cognitive economics”) to better understand the effects of tax policy.  Among the many interesting behavioral studies of his that Raj summarized today, one focused on how the “salience” of taxes (how obvious taxes are) matters in terms of how those taxes affect economic behavior.  The hypothesis was that the more visible or obvious taxes are, the bigger the behavioral responses would be.  Raj described how he conducted an experiment involving price tags in a drug store, where for a period of time the price tags on a certain subset of items (hair accessories, actually) in one particular store were elaborated upon, such that the (usual) price before taxes and the gross-of-sales-tax price (referred to as the “total price”) were displayed.  (You can find a photo of these price tags in this slide presentation.) Comparing with the appropriate “controls” (demand for other goods in the same store with ordinary price tags, demand for the same type of (hair accessory) products with ordinary price tags in other stores), Raj and his coauthors found that when the sales taxes and gross prices were spelled out on the price tags (note: taxes were not raised, just clarified), sales of those goods (quantity demanded) fell.  Raj had a funny story about how the store manager did not let him use the special price tags on a larger class of goods, because he had a hunch Raj’s hypothesis would prove right!

It got me thinking on the spot about how this bit of behavioral economics applies to deficit spending.  The reason why deficit financing of government spending and tax cuts proliferates is because it’s (falsely) perceived as “free”–or at least as less painful (less costly) than having to come up with offsetting tax increases or spending cuts.  When in fact, economically at least, exactly the opposite is true.  Even leaving aside the riskiness of high deficits and debt to the stability of our entire economy, there’s at least the objective and easy-to-quantify cost of the compounding interest on the added debt.

So what if every time a deficit-financed spending program or tax cut is proposed, the CBO puts a “price tag” on it that is not just the standard legislative cost of the spending or tax cut, but the gross-of-interest amount, maybe under various assumptions about how long paying down that debt will be put off?  Just for example, if the Bush tax cuts that President Obama wants to extend (all but the top two brackets) are extended and deficit financed (as the President proposes, and remember, these Bush tax cuts are exempt from Obama PAYGO rules), the gross-of-interest “price” would not be “just” $2 trillion over ten years, but maybe over $5 trillion over ten years if you count the compounded daily interest and assume the paying down of principal doesn’t begin for 20 years.  (I got that by assuming a 5 percent annual interest rate, but you can play around with different rates and terms using this handy dandy compound interest calculator here.)

In other words, this would spell out for people–the politicians and ordinary citizens alike–that a deficit-financed tax cut today just means a several-fold tax increase (on our kids) later.  And the later “later” is, the larger the “multiplier” on that future tax increase (or spending cut).

I guess this seems the opposite of the “dynamic scoring” of tax cuts that some conservatives who embrace supply-side economics advocate as a way of reducing the officially-scored costs of tax cuts.  But if deficit-financed tax cuts were to be truly “dynamically scored,” not only would the direct costs of compound interest count against it (which is all I’m here suggesting be added to the “price tag”), but the adverse effects of reduced public and national saving on economic growth would raise, not lower, the costs.

8 Responses to “Using Sticker Shock to Clarify the Costs of Deficit Spending”

  1. comment number 1 by: Greg Ransom


    A deficit-financed sending increase just means a several-fold tax increase later.

    I’m guessing that is the case you are most interested in.

  2. comment number 2 by: SteveinCH


    The problem with what you propose is that no individual program is responsible for debt unless you assume that all other things are equal. As an example on the other side, you and many others have made the point that HCR is “paid for”; but, in point of fact, you could equally argue that the offsets could have been used (except for the penalties on the individual and corporate mandates) to pay down the debt. Thus, the “cost” of HCR using a similar methodology would be the entire cost (net of the penalties) and the additional debt. So, by the same logic, you can’t pay for a program per se since any payments could have been used for an alternative purpose, namely lowering the deficit and future debt.

    Our perspective on the impact of a program will be entirely dependent on our view of the counterfactual. That makes doing the type of analysis you describe quite complex.

    Having said that, I would be completely supportive of an approach that added in deficit impacts, assuming of course that the only offsets a new program could count would be offsets that can only exist as a consequence of that program.

    As to “A deficit-finance spending increase just means a several-fold tax increase later.”, that statement would only be true if it included “or an even larger cut in services” before the word later.

  3. comment number 3 by: npm

    Diane - I want to preface this by saying that I’ve been reading and enjoying and almost completely agreeing with the blog since the beginning. But in this case, I think your argument is wrong.

    Cost-benefit analysis 101 says the costs of any policy decision should be put in present value terms for clarity and ease of comparison. Therefore, not only should we continue to exclude interest costs, but we should discount (using a nominal interest rate) the projected noninterest spending flows. (And in principle, we should also project the infinite horizon stream of outlays before discounting, but that’s not feasible.)

    Another way of thinking about it: you write that “a deficit-financed tax cut today just means a several-fold tax increase (on our kids) later.” In what terms? In present value terms - assuming no macro feedback - the tax increase is exactly equal to the additional spending.

    There are plenty of arguments to be made for fiscal responsibility - we simply shouldn’t spend now and transfer the responsibility forward; there are macro feedbacks from capitol crowdout; to minimize deadweight loss, we shouldn’t have high variance in tax rates; etc. - but the simple fact that there is interest is not one of them.

  4. comment number 4 by: VAT Brat

    Diane — Why endorse such a Rube Goldberg mechanism for educating the voters about deficit spending?

    The best way to enhance someone’s interest in an issue is to give them the power to affect its outcome. A much simpler and effective approach is to give voters the ultimate authority for raising the public debt limit.

    A Debt Limit Referendum (see ) would require Congress to get the voters to ratify any request to increase the public debt limit before it could take effect.

    During the general election when candidates make their arguments to the voters about whether or not the public debt limit should be increased, the pros and cons of deficits will be explained very thoroughly.

  5. comment number 5 by: SteveinCH


    I understand your excitement about your cause, but posting the link if every post you make on every thread is unlikely to help the cause.


  6. comment number 6 by: economistmom

    npm: Thanks–I appreciate your comment, both your preface and the immediate matter of contention. (You are not the first smart person to point out my neglecting the present-value equivalence of the compound interest, btw.) Of course, you are correct that in present-value terms whatever spending or tax cut we finance today is exactly offset (and no more) by the lower spending and higher taxes later. But it’s precisely because it’s our generation that benefits from the extra spending today, but our kids’ generation that will pay for it (with interest) in the future, that makes our generation unfairly and very subjectively (irrationally) over-discount/under-value that later cost–i.e., even beyond the market interest rate. At the NTA conference, panel moderator Joel Slemrod (of U. of Michigan) actually referred to my point as asking how we could get people to follow “Ricardian” accounting–a reference to the “Ricardian equivalence” theory where individuals account for the welfare of future generations in their own decisions today. I’m simply saying I think we need a way for people to see more clearly the burdens they’re placing on their children when they support deficit-financed spending today–that even if the tradeoff is “fair” from an objective, market-based sense, people aren’t necessarily seeing it that way, and their subjective, nearsighted biases turn what could have been a properly-considered tradeoff into an irresponsible, ill-conceived one.

  7. comment number 7 by: Nathan Andover

    If it could be figured out it would be a nice way to present the cost of government programs.

    In a similar way, most people think they are buying a house for $200,000, but they are really buying a house for $370,000 over the life of the mortgage.

    I wonder how the purchase of homes would be different if the price plus mortgage costs were listed for all homebuyers before they sign on the bottom line.

  8. comment number 8 by: SteveinCH

    Not sure this is totally on topic, but the following is an interesting (and quite frightening) read.