The $25 Billion Shell Game
Posted September 26, 2005 | Link

What's a few billion dollars between friends? Nothing, it would seem, if these friends work for a federal agency or are, in fact, federal agencies themselves. The quote in the Stop The Buck! masthead above illustrates just how out of touch these groups are with the people (namely, us) who pay their salaries. The day before 9-11, Donald Rumsfeld admitted that his department could not account for more than $2.3 trillion in defense transactions. Over what period this theft occurred he did not say, but similar warnings from the outgoing Clinton team indicate that this problem is not particular to the current administration.

One of the problems with having a budget as large as the Defense Department's is, of course, that the graft and corruption is higher too. Assuming that all federal departments (and all large institutions in general) waste some percentage of their budget to the ineptitude, greed, or thievery of its executives, we should perhaps set some sort of bar as to when the worrying should begin. Conservatives might complain that this bar should be set at zero and, while I might agree with them on principle, this is effectively a non-starter. Any sober look at our government's past performance will indicate that waste is as much a part of the federal budget as offensiveness is a part of TV's South Park.

So, for the sake of argument, let's assume that the spoilage in any budget should not be more than 5%. I use the term "spoilage" here to mean graft and corruption. Waste is too broad of a term to be of much use and it is highly subjective. In my humble opinion, most federal agencies (except law enforcement) could probably withstand a 25-50% cut in staffing budgets and probably do an equivalent job if tasked to do so. So let's just concentrate on the worst type of waste: the intentional and conscious diversion of taxpayer funds in a secretive or obfuscated way to a purpose other than what was originally (i.e., Congressionally) intended.

Why 5% you might ask? No particular reason other than the fact that a 1% threshold would probably be triggered in every case and a 10% limit actually seems kind of high when the budgets are particularly large. For example, the Defense Department's Fiscal Year 2006 budget estimated that FY 2006 spending on defense (i.e., the period between October 1, 2005 and September 20, 2006) would be approximately $447 billion. Despite the fact that the previously announced shortfall of $2.3 trillion would equate to a $115 billion per year problem over 20 years, it just seems ridiculous to me that we should consider a $45 billion per year ($447 billion x 10%) loss as just part of doing business. Thus, let's stick to the 5% threshold.

But first, a quick lesson on the budget process. In January of every year, the executive branch of our government (the president and his budget advisors) submits a budget to Congress. This budget contains all requested spending for the next fiscal year plus projected spending levels for the four years following the budget year (five years total). After several months of wrangling, the Congress passes all of the appropriations bills and the new fiscal year begins on October 1. The most interesting thing about this process, however, is the fact that the budget also contains complete historical information for every fiscal year in the past. It is in these historical tables that trends can be studied and patterns analyzed.

Okay, so here's where the shell game occurs. I glanced at the national defense budget from FY 1998-2010 (Warning: Excel XLS file; FY 2005 is still in progress; FY's 2006-2010 are projected) and noticed two strange things. The second oddity will be covered in a subsequent article, but the first thing that caught my attention is that defense spending actually seemed to be lower than I expected. This seemed out of sync with a Defense Department that is now spending 50% more than it did just five years ago. So, I did a little checking.

It turns out that the actual spending on national defense in current dollars was in line with my expectations. However, the same numbers in constant (i.e., inflation-adjusted) dollars was in fact much lower than was to be expected. Constant dollars are calculated using special deflators, which are numbers that the government publishes each year to make year-to-year comparisons possible by removing the effects of inflation. The federal budget recommends specific deflators just for defense spending, so the inflation-adjusted constant dollars that one could derive from them are very accurate. The problem is that the government has also been playing games with these numbers.

The problem with such games is that they stand out like a sore thumb. The reason why this is so is because any type of chained price index (what a deflator is) follows, more or less, the ebbs and flows of inflation. After all, that's what a price index is--a measure of the effect of inflation over a period of time. First, look at the Consumer Price Index-Urban (CPI-U) trend in the graph on the left, below. The CPI-U is the most used measure of inflation in our country. As you can see, it's been more or less trending down for a long time with no sudden spikes since FY 1990.

Consumer Price Index-Urban (CPI-U) Notice how the trend is going down over time and how there is no spike in FY 2003	Defense Deflator vs. The Chained Price Index Notice that the slopes of both lines are the same and that the CPI shows no spikes.

Now look at the graph above and to the right. In red, the Defense Department's deflator follows, as would be expected, the trend in the Chained Price Index (CPI; used as the universal deflator for aggregate federal spending). However, it experiences two relatively large spikes in FY 2003 and 2007 that are both unexpected and unexplainable. If we look more specifically at the actual Defense Department's budget, we can see how unusual those deflator spikes actually are.

The graph above clearly illustrates, albeit in a crowded way, how the deflators for defense spending changed in every federal budget from FY 2000 to this year's FY 2006 budget. Like the CPI-U trend noted earlier, they do not fluctuate much from year to year. However, in the FY 2005 budget, the FY 2006 deflator was suddenly projected to be much higher than any previous year. When the FY 2006 budget arrived earlier this year, the FY 2006 deflator moved back down to previous levels, but the spike moved to FY 2007 instead and went up even higher. In addition, the FY 2006 budget actually went back in time and retroactively changed the deflator from a past year (FY 2003) and made it increase almost twice as much as the prior budgets said it did!

Now, all of this is hardly as exciting as Hurricane Katrina or the start of the 2005-06 NFL season, but bear with me a bit more. Our government, which frequently uses inflation-adjusted dollars when making budget comparisons from year to year, just went forward and back in time to tweak two different deflators and increase them by two to three times the rate that historical precedent would suggest. What was going on?

Well, the first things that came to mind wound up being the only things that came to mind, namely, the costs associated with the War on Terror and Operation Iraqi Freedom (OIF), the rising costs of oil, and the extra profits being incurred by leading Iraqi/Afghani reconstruction contracts. However, these were all quickly ruled out: most of the OIF and reconstruction contracts were appropriated in addition to the national defense budget; the extra profits going to companies like Halliburton/KBR, despite their heft (and the Army being overcharged as well), are not enough to drive inflation across the entire DOD; and the rising price of oil did not have an inflationary effect in any other branch of government. In short, there was no obvious reason for the disconnect.

So, what was the impact of this tweaking? Well, this part is fairly easy to estimate. Depending on whether you take a long-term average of defense deflators (e.g., FY 1986-2010) or a local average around the suspect years (FY 2002/04 and FY 2006/08), you come up with a constant dollar amount that suggests our national defense budget is off by about $22-25 billion, just from FY 2003 to FY 2007. In other words, despite the fact that our government reported raw, current dollar numbers that are ostensibly correct, it then artificially raised the effect of inflation--and only for defense spending--and caused the constant dollar totals for defense to appear approximately $22-25 billion less than they should have been.

This problem, I'm afraid, doesn't have a happy answer. Although it's nice to know that the actual/current dollar amounts appear to be reported correctly, I'm still not sure why the recent past and imminent future have been so drastically changed--and only for defense. What is the value of artificially lowering inflation-adjusted defense spending numbers? If the CIA, or the intelligence community in general, was hiding funds in the DOD budget as it historically does across many departments, then they could just take it out of the current dollar accounts and leave the deflators alone in order to avoid suspicion. Likewise, any other obvious shenanigans would probably not involve such a high value discrepancy, since most secret operations operate on much less than $25 billion.

I suspect that these changes have a lot to do with remaking the DOD into a new type of entity going forward--much different than anything we've seen to date and certainly a world apart from the Cold War monolith that has operated, more or less intact, since the Korean War. Remember the first oddity that I saw when looking through the budget? We'll try and tie this all together in an upcoming article.