Federal outlays in the late 1990s

Patrick Ruffini’s blog is one of the more thought-provoking political blogs out there. I often disagree with Patrick, but his writing demonstrates a literacy and his research a numeracy that I appreciate and learn from, and I often forward his entries to friends.

I was surprised, then, to read these assertions in one of Patrick’s entries for May 11:

The late Clinton years showed the tendency of economic trends to overwhelm virtually anything in their path. As a result, we had effortless surpluses that flowed as easily as black gold does into Saudi royal bank accounts. But instead of reinvesting the surplus back into the private economy in the form of tax relief to individuals, President If-You-Don’t-Spend-It-Right squandered it on lavish domestic spending.

Looking back through history, the biggest absolute changes in the size and scope of government don’t occur low-spending-growth, low-revenue growth periods (like now, minus the war). They occur in periods like the late ’90s when nobody’s looking and nobody cares about double-digit spending increases, and in periods of wartime. Much to the chagrin of budget hawks, we have just had two such periods run up against one another.

As we’ll see below, Patrick has a point when he says that the “late Clinton years” saw relatively liberal spending, but only when compared to the early Clinton years, which were a time of remarkable and laudable fiscal restraint by the federal government. The increase in federal on-budget outlays during the relatively liberal “late Clinton years” was similar to the increases seen during the Reagan and Bush 41 administrations.

Looking at the Clinton years as a whole, there was not a single year when the federal government’s on-budget outlays had a “double-digit spending increase” — the largest increase was between FY1999 and FY2000, when the outlays increased by 5.6% (and that figure is unadjusted for inflation).

One way of looking at the spending pattern of the ’90s: The on-budget outlays in Reagan’s last budget (FY1989) accounted for 17.3% of the national GDP and his largest budget (FY1983) accounted for 19.2%; the on-budget outlays in Bush 41’s last budget (FY1993) accounted for 17.4% and his largest budget (FY1991) accounted for 18.3%; the on-budget outlays in Clinton’s last budget (FY2001) accounted for 15.1% and his largest budget (FY1994) accounted for 17.0%. (The Congress played a major role in all of these budgets as well, of course.)

Another way of looking at it: Below is a table showing how much of the incremental increase in the national GDP was used for an incremental increase in federal on-budget outlays in each fiscal year of the Clinton administration (remember as you read these that every one of Reagan’s and Bush 41’s 12 budgets used at least 17% of the national GDP for on-budget outlays):

GDP delta    Outlay delta   % taken

FY1994       $386.2b       $39.6b       10.3%
FY1995        379.4b        44.6b       11.8%
FY1996        370.6b        32.5b        8.8%
FY1997        490.6b        31.0b        6.3%
FY1998        478.7b        45.4b        9.4%
FY1999        473.8b        45.1b        9.5%
FY2000        581.1b        76.9b       13.2%
FY2001        302.7b        59.0b       19.4%

In other words, only in the final year of the Clinton administration did the percentage of new GDP taken for new on-budget federal outlays exceed the percentage of total GDP taken for federal on-budget outlays in each year between FY1980 and FY1993.

I would argue, then, that nothing like “the biggest absolute changes in the size and scope of government” (if we’re measuring “size” and “scope” in terms of money) occurred during the Clinton years. As a budget hawk myself, I was surprised and impressed by how effectively the federal budget deficit was wiped out by the Clinton administration and the Republican Congress.

(All numbers used in this post are from the Historical Tables of the Budget for Fiscal Year 2004, a PDF version of which is available at the website of the Office of Management and Budget.)

%d bloggers like this: