If We’re Going to Take Budget Forecasts Seriously, Then This is a Good Way To Present Them

March 15th, 2014 at 4:52 pm

The figure below, from one of the administration’s recent budget reports, provides a range of uncertainty around their central forecast of the deficit as a share of GDP.  Using the errors from past forecast, we can get a rough sense of the confidence interval around our current estimates.

By 2019, the deficit has a 90% chance of being within a range of +4.6% of GDP and -9.1% of GDP.  And this is just the five-year range.  The 10-year range fans out much more than what you see below.

I’m not saying we shouldn’t make such forecasts.  Within a few years of the forecast, they provide useful guidance of the budget path under a number of assumptions (each of which have their own range of uncertainty).  I’m saying we should be a) very mindful that these are estimates that grow increasingly uncertain over time and thus b) very careful about asserting certain policy courses based on them.


Source: OMB, Analytical Perspectives

Print Friendly, PDF & Email

5 comments in reply to "If We’re Going to Take Budget Forecasts Seriously, Then This is a Good Way To Present Them"

  1. Larry Signor says:

    Does anyone want to buy a bridge? How about a lottery ticket? The real truth about economic forecasting (witchery?) is that if you wake up and your wife still loves you, it’s a good day to hunt bear. Beyond that, it’s all black magic to most people.

  2. David Merkel (@AlephBlog) says:

    They estimated standard deviations off past forecast errors, and assume that future forecast errors will follow the same level of volatility.

    Two problems: this misses the fact that most forecasts on the budget deficit have been too optimistic. Shift all the percentiles down. Second, forecast errors tend not to be normally distributed — the tails are fat because of the boom/bust cycle.

    If this is an improvement, it is a small one.

  3. David Merkel (@AlephBlog) says:

    One more note: current politicians will wring their hands but not do much over the deficit. They would *never* let the budget get into a surplus. Even Clinton didn’t — Social Security provided the surpluses.

    • Robert Bostick says:

      Well, they’ve screamed for cuts (smaller deficits) and now there are fewer net financial assets in the private sector and that’s what’s preventing a robust recovery. Public sector deficits are equal, to the penny, to non-public sector surpluses. Cut the former the same happens to the latter. Move to a balanced budget and we’re screwed. Sectoral balances don’t lie.

  4. John says:

    I am having trouble reading the diagram. The title is “Range of Uncertainty for the Budget Deficit.” Year 2014: the range of uncertainty might be between -3% and -6.5%. The problem is that uncertainties can never be negative — they’re always positive by definition.

    Maybe it means that the deficit will most likely be between -3% and -6.5% of the GDP. In other words, we’re virtually certain to be running surpluses. (Negative budget deficit means budget surplus.)

    One further point: surpluses for a currency creator are evil: they draw money from the economy. Net savings by everyone else is necessarily equal to the amount of money put into the economy. If an economy consists of two persons, and one person manages to save $5, then the other person had to lose (or unsave) $5 if the total amount of money is fixed. Furthermore, money supply must grow with economic growth to maintain stable prices. (Money velocity has only limited ability to increase to account for the growth.)

    Another point is that for a currency creator, money in their own currency isn’t real money. They take it in and then recycle it, reuse it, scrap it, or burn it as they choose. (A better metaphor is “coupon” or “ticket” for them.) The only reason for taxation is to give the “money” a genuine value, in that one must turn in so many coupons.