Print Friendly and PDF

The Bear's Lair: Discounting political numbers

April 28, 2014

For the last couple of decades, U.S. budgets have been prepared on a 10-year basis.
The Congressional Budget Office scores the 10-year deficit, while the sustainability of Social Security and other entitlement programs is calculated on the basis of their deficit projected 75 years ahead. Politicians have learned to game this system, inventing new programs that spend money up-front, with supposed vast savings later in the decade—which of course never materialize. The reality is that money a decade or two ahead is worth far less than money today because the electoral process can intervene and cause it to disappear. Ordinary business discount rates are hopelessly unrealistic, especially at current low interest rates. I therefore propose herein a new formula, a political discount rate, for public-sector budgeting.

There are two factors involved when looking at a projection of public sector income and expenses. One is the cost of money, which for the public sector is effectively the Treasury bond rate, perhaps that for 10 years when a 10-year projection is being considered (or somewhere just under 3% per annum today). The other is the uncertainty in political projections, because presidential elections every four years and, to a lesser extent, Congressional elections every two years make future cash flows very uncertain.

The nature of elected government means long-term plans for policy or spending are futile. There is simply too much uncertainty about the world beyond the next election for such plans to be meaningful. For example, if Mitt Romney had won in 2012 (his defeat feels inevitable now, but didn't then) then the Obamacare legislation would have been effectively repealed, one way or another. Equally, if Al Gore had beaten George W. Bush in 2000, then we probably would not have had the Iraq war (even though we would almost certainly still have had the 9/11 attacks), and the Republican who would probably have been elected in 2008 might have corrected any overspending follies that a Gore presidency would have brought us. Not only the budget trajectory, but the world's entire economic trajectory, would today be very different.

One solution to this would be to have one of the two chambers of Congress elected for very long terms – say 20 years, on a staggered basis. That would ensure that the chamber elected in such a way took long-term costs and benefit into account and (if it had to opine on budgets) would also ensure that gimmicky front-loaded proposals were quashed. The British system had such a Second Chamber, in the hereditary House of Lords, which being hereditary could take a proper interest in the world of 2100 or even 2200, but alas its hereditary nature was abolished by the destructive Tony Blair in 1998.

However, short of changing the entire political system, we can at least budget in such a way as to reflect its realities. As a first approximation, we can assume that the chance of political change at each Presidential election is 50%--the election is essentially a coin-flip. Then one dollar due in 2018 is not only worth 11% less than one dollar due today, it is worth 55.5% less (44.5 cents, not 89 cents) because it has only a 50% chance of existing, while in the other 50% case it can be abolished by the other party. Similarly one dollar due in 2022 is worth not 79 cents today, but 19.5 cents, because it is subject to two election cycles that may make it disappear through a complete change in policies.

You can do this in a more sophisticated way if you like, to take advantage of Congressional elections and the reality that a President once elected has about a 75% chance of being re-elected. Thus, rather than an 8-year cycle involving one 50% step-down, the 8-year cycle could consist of an initial 10% step-down for the first midterm election in which incumbents generally do fairly well, then a 25% step-down in the next Presidential election in which the incumbent runs for re-election, then a 26% step-down for the second midterm election, in which the incumbent President has a much greater chance of losing and nullifying his policies for the last two "lame-duck" years of his second term. At the end of eight years, you have another 50% write-down and the cycle continues.

Thus if you're setting a 10-year Congressional budget today, the first year, to September 2015, would be scored at 100%, then the second and third (2016 and 2017) would have weightings of 72% and 70%, respectively, reflecting the upcoming midterm elections. (Elections by and large have little effect on the fiscal year in progress when the election occurs, but fully affect the following one.) Then 2018 and 2019 would have weightings of 34% and 33%, reflecting the effect of the 2016 Presidential election, at which there will be no incumbent. Fiscal years 2020 and 2021 would have weightings of 29% and 28%, reflecting the 2018 midterms. And fiscal years 2022 and 2023 would have weightings just above and just below 20%, reflecting the 2020 Presidential election, at which an incumbent will be standing for re-election. And so on, as far into the future as you like.

This would have a number of interesting effects on the budgeting process. For a start, all the gigantic, multi-trillion 75-year deficits in Social Security and Medicare programs would disappear. The current estimate is that the 75-year deficit in the trust funds, discounted in the normal way, is something like $87 trillion, with the annual deficit in 2087 being something like $4 trillion, in today's dollars, on a GDP of $66 trillion in that year (compared with $17 trillion now.)

Yippee! Political discounting makes that deficit look much manageable. The present value of 2087's $4 trillion deficit, discounted politically, is $1.8 million, a much less frightening number, indeed infinitesimal in budget terms! What's more since the discounting factor falls below 0.1% in 2050, the entire accumulated deficit over the next 75 years would be a small fraction of the current national debt rather than a substantial multiple of it. Phew!

Of course that long-term number hasn't gone away altogether. What's more, every four years when you have an election, the politically discounted deficits in the Social Security and the Medicare system double and look frighteningly large compared to their previous level. But that's great; it will make the new administration do something serious about them, and stop them growing to the point at which they swallow the U.S. economy seven times over.

The effect would be even more significant on the usual 10-year budget horizon. Budget gimmicks that increase spending early on in the 10-year cycle, to be funded by supposed restraint in the latter years of the cycle, now wouldn't work. For example, under the Congressional Budget Office's calculations, the President's 2015 budget, instead of reducing the deficit by $1.05 trillion over a 10-year view compared to the baseline, reduces it by only $168 billion. But that entire reduction is achieved through $353 billion of tax increases. On the other hand the budget put forward by Rep. Paul Ryan (R-Wis.) avoids these gimmicks and would still reduce the 10-year deficit by $1.6 trillion instead of its projected $5.3 trillion. (To be fair, we don't yet have a Congressional Budget Office scoring of the Ryan budget for a proper comparison.)

The same principle can be applied in other countries. When David Cameron's Coalition government was elected in Britain, his promise to balance the budget by 2016 initially had credibility because it concerned actions all of which would take place within the lifetime of the 2010-15 Parliament. However, as time has worn on and the U.K. budget has failed to move towards balance, the promises of good behavior to come become less-and-less credible because there is clearly a substantial chance that Cameron's government won't be around after May 2015 to introduce spending cuts due in 2017 or so.

Voters know that political promises are mostly pie-in-the-sky. So let's have a budgeting system that reflects that and doesn't give excessive credit for promises to cut spending and balance the deficit eight or 10 years from now. Politicians should be graded on what they do this side of the next election and thrown out if they fail to perform.


Commentary Archive