Nine days ago Mr. Rex Nutting of MarketWatch wrote a provocative column titled “Obama spending binge never happened.” Here is the key quote:

Over Obama’s four budget years, federal spending is on track to rise from $3.52 trillion to $3.58 trillion, an annualized increase of just 0.4%.

Referencing and relying on this article (rather than on the hundreds of talented OMB career staff who sit nearby), White House Press Secretary Jay Carney told reporters:

I simply make the point, as an editor might say, to check it out; do not buy into the BS that you hear about spending and fiscal constraint with regard to this administration. I think doing so is a sign of sloth and laziness.

President Obama then followed suit in a campaign speech in Iowa one week ago:

But what my opponent didn’t tell you was that federal spending since I took office has risen at the slowest pace of any President in almost 60 years.

President Obama and Mr. Carney are both relying on Mr. Nutting’s key quote above. Let’s examine the quote as Mr. Carney suggests we should do.

Problem #1: The President argues that his fiscal stimulus law, enacted in February 2009, had a big positive effect on the growth rate of the economy. We are now asked to believe that President Obama’s policies did not significantly increase spending but did significantly increase economic growth. This is, to say the least, an intellectually inconsistent argument. The whole Keynesian fiscal stimulus argument is premised on a significant increase in government spending.

Problem #2: Mr. Nutting assumes that since a President serves a four year term he should be measured for four budget years. But since budget years begin in October and Presidential terms begin in January, the fairest and most accurate way to measure the budget effects of a one-term President is to look at five budget years, not four.

Mr. Nutting mistakenly assumes that FY 2009 spending must “belong” to either President Bush or President Obama. As I explained a while back, the two Presidents should share responsibility/blame for this transition year, since each influenced the spending within it. A form of FY 2009 spending should count in both of their records. The same is true for any budget year that spans a Presidential transition.

The easiest way to see the flaw in Mr. Nutting’s four year budget window is that his 0.4% average annual growth rate assumes the Obama Presidency began on October 1, 2009.  That can’t be right.

Problem #3:   Beginning measurement eight months into a Presidency would be bad under any circumstances, but in this case it’s critical. The year Mr. Nutting excludes is the financial crisis year, when government spending spiked because of bailout costs for the banks, AIG, Fannie & Freddie, and the auto manufacturers, as well as the first year effect of President Obama’s stimulus law.

Mr. Nutting therefore measures the growth rate not just from the wrong date, but relative to a spending level that was extraordinarily high because of one-time events.  Any spending growth rate, over any timeframe, measured relative to the $3.52 trillion starting point of FY 2009 will look good because that starting point is so unusually high.  Mr. Nutting cherry-picked his start date to make President Obama’s record look good.

The start/end point gimmick is not a new tactic. Rep. Debbie Wasserman Schultz did something similar with job creation two years ago.

Problem #4: Mr. Nutting mistakenly uses $3.58 trillion to represent the President’s budget for FY 13.  This is CBO’s projected baseline spending (i.e., current law spending), not their estimate of what President Obama has proposed.  The correct number for the President’s FY 13 proposed spending is $3.72 trillion. (See Table 2 here.)

So in the key Nutting quote upon which the President relied:

  • “Obama’s four budget years” should be five;
  • $3.52 trillion skips all spending increases in the first eight months of the Obama Administration, including the early implementation of the stimulus law;
  • $3.58 trillion is a simple factual error;
  • and the calculated 0.4% average annual growth rate depends on all three of the above.

If you instead do this calculation the right way and measure the average annual growth rate from FY 2008 to CBO scoring of the President’s budget proposal for FY 2013, you get an average annual growth rate of federal spending of 4.5%.  That’s a nominal growth rate, so the real growth rate will be in the 2s.

As I describe, below you should be careful even using my number because growth rates are an incomplete and therefore inaccurate measure of spending.

Problems 1-4 above reveal the core fiscal policy lesson, which as best I can tell no one else has publicly revealed:  don’t rely only on growth rates.

Problem #5:  It is a mistake to judge a fiscal policy only by looking at growth rates.  At a minimum you need to look at both levels and growth rates.  The best thing to do is to examine average levels over time.  The spending levels reveal more useful information, and it’s nearly impossible to gimmick spending levels as the Nutting article gimmicks spending growth rates.

The second most important gimmick in the Nutting article is his choice of FY 2009 as his starting point for measurement.  The most important gimmick is his choice of an average annual growth rate as the right metric for spending.  You calculate an average annual growth rate by picking a starting point and an ending point, drawing a line between them, and then figuring out the slope of the line.

  • This works fine if the path you’re measuring is a smooth line or even a smooth curve.  The more irregular the path, the more your choice of endpoints affects the measurement of the slope.  Mr. Nutting took advantage of that arithmetic fact here by choosing his start point to make President Obama’s spending growth look surprisingly low, but even someone not trying to spin you would be relying too heavily on judgment calls about the appropriate start and end points.  The 4-vs-5 years debate matters a lot if you’re measuring the slope of the line, and much less if you’re instead measuring levels.
  • The average annual growth rate metric also ignores all the intervening years and therefore loses a lot of potentially useful information.  Measuring average levels over time includes this information.
  • As a policy matter we should care about how much government is spending (the level), and whether based on our values we think that amount is too much, too little, or just right. If properly calculated, growth rates can be a useful shorthand to allow us to evaluate how levels are changing over time, but growth rates are useless if we don’t understand the levels.  A 5% growth rate from a historically high starting spending level is a very different animal than the same growth rate from a much lower starting level.


I will conclude with some facts.

  • The historic average is federal spending of 20% of GDP, +/- 0.2 percentage points depending on when you start your measuring window.  That’s pre-2008 crisis, so I’m ending the measurement with FY 2008.
  • Federal spending averaged 19.6% of GDP for President Bush before the crisis year of FY 2009.  If you include FY 2009 in President Bush’s average, including the Obama stimulus and the appropriations laws President Obama signed, Bush’s average is 20.1%.
  • The highest level since the end of World War II, pre-financial crisis, was 23.5% of GDP in 1983.
  • In FY 2009, the financial crisis year that spanned the Bush and Obama presidencies, federal spending was 25.2% of GDP.
  • If President Obama’s FY 13 budget is enacted as he proposed it, during the first term of an Obama presidency spending will average 24.1% of GDP. If for some reason you want to exclude FY 2009 as Mr. Nutting did, your average is still 23.8% of GDP for a one-term Obama presidency.

Federal government spending in the first year of the Obama Administration was extremely high, and much of that was either put in place before President Obama took office or was out of his control.  Nevertheless his policies have maintained an extraordinarily high level of spending through his first term, and he proposes to continue to do so if elected to a second term.

Federal spending has averaged 20% of GDP for decades.  President Obama is presiding over a much bigger government, at 24% of GDP.  If his latest budget were enacted in full and he were elected to a second term, the average over his tenure would be 23.4% of GDP.  This means that, relative to the economy, federal government spending would be 20 percent larger than the historic average during a one-term Obama presidency and 17 percent larger than average during a two-term Obama presidency.

That is a spending binge.

(photo credit: Purple Slog)