Flawed budget analysis from CAP, continued

A few days ago, I criticized a column about Bush Administration budget policy by Scott Lilly of the Center on American Progress. Here's Scott's response, and my follow-up:

Evidently my piece on No Child Left Behind funding failed to make my point as clearly as I had hoped—at least to Mr. Carey.

The question I pose is simple. What has happened to education funding once the White House got Congress to agree to the authorization that Bush so badly wanted? Spellings has attempted to dispel criticism that the Administration forgot about the resource side of the problem once the bill was signed by saying that funding has increased for NCLB programs since 2001. But nearly all of that increase--and in inflation-adjusted terms, more than all of the increase--took place before the act was adopted.

Another way to look at the same question is yearly funding increases for Elementary and Secondary Education for the 6 years before the bill was signed and the six years afterwards.

(Note: having a little trouble with blogger formatting here, in each row the first number is the federal fiscal year, the second is the total Elementary and Secondary Education appropriation, the third number is the dollar amount change from the previous year, and the fourth number is the percentage change from the previous year. - KC)

1996 14,404,928

1997 16,464,731 2,059,803 14%

1998 18,164,490 1,699,759 10%

1999 20,951,877 2,787,387 15%

2000 22,600,399 1,648,522 8%

2001 27,316,893 4,716,494 21%

2002 32,078,434 4,761,541 17%

2003 35,113,253 3,034,819 9%

2004 36,942,478 1,829,225 5%

2005 37,530,257 587,779 2%

2006 36,463,840 (1,066,417) -3%

2007 36,312,228 (151,612) 0%

2008 36,748,577 436,349 1%

The average increase in the years prior was 14%. The average increase since is 2.4%

In real or inflation adjusted dollars the picture is even worse. Real growth tumbled from 11% to a negative .4%. In real dollars we will be spending less on elementary and secondary education if the President’s 08 budget is adopted than was true when the President signed the law. I believed at the time NCLB passed that it was a political ploy to allow the White House to appear to have an education agenda when all they really wanted to do was stop the growth of the education budget. I believe these figures give support to that theory.

Here's the problem: Scott is just using the same flawed analysis in this post that he used in his column. His 14% to 2.4% comparison depends on assigning the large budget increases of FY 2002 to the period "before" NCLB. This doesn't make any sense. As Scott himself explains quite clearly in his column, the FY 2002 budget was negotiated at the same time as NCLB. They were enacted within weeks of one another, as two pieces of a single bargain.

Let me put it this way: If the NCLB had been signed into law a few weeks before the FY 2002 budget was passed, instead of a few weeks after, would that change his analysis? Wouldn't it have to? And doesn't that show that it rests on an essentially semantic distinction?

Moreover, the largest single piece of NCLB -- Title I -- is a forward-funded program. That means that even though the FY 2002 budget was finished in December 2001, schools didn't start to get the money until the beginning of the 2002-2003 school year--nine months after NCLB was enacted. To characterize this as pre-NCLB money makes no sense at all.

Scott might respond that even if you concede this point, the average year-to-year growth rate before NCLB was much larger than the average rate afterwards, which is true. But this also show's why Scott's average annual growth approach to comparing budget policies is, methodologically, the wrong way to look at this issue.

That's because it doesn't take into account the fact that a lot of the big growth in NCLB dollars took place up front, in year one. That significant percentage increase didn't just help schools in the first year, it helped them in the first year and every year afterward.

This makes a big difference. For example, let's say you're a school district that gets $1 million per year from NCLB. You have a choice. You can (A) Get a 25% increase next year, and no increase for the four years after that, or (B) Get no increase for the next four years, and a 25% increase in the year after that. Which would you choose?

According to Scott's analysis, you shouldn't care, because in both cases the average annual growth rate is exactly the same, about 4.5%. But of course, you would care, and you'd choose (A), because with (A) you would get a total of $6.25 million over those five years, while (B) would only give you $5.25 million, a difference of a cool million dollars.

This distinction can be seen in the very numbers Scott has posted above. In the five years from 1996 to 2001, schools received a total of $33.5 billion more than they would have received if ESEA funding had been fixed at 1996 levels for that entire a time, a difference of $6.7 billion per year.

In the seven years from 2001 to 2008 (assuming for the sake of argument that the 2008 proposed number becomes the actual number), schools received $60.0 billion more than they would have if funding had been fixed at 2001 levels for that entire time, a difference of $8.6 billion per year.

So even though the average percentage increases were much bigger from 1996 to 2001, schools actually got more money from 2001 to 2008, because percentages compound, and because NCLB money came sooner rather than later. Even if you adjust for inflation and population growth, the numbers are comparable.

As I said before, I don't think the Bush administration has done nearly enough for NCLB funding. But if you want them to do more, you should start by giving them fair credit for what they've actually done.