The spending legislation predicted to pass the Senate contains most of the items on the typical American’s “Why I Hate Congress” list:

  • Passing 11 spending bills in one big lump 2 1/2 months after they were due at the beginning of the fiscal year
  • Punting the Homeland Security appropriations bill to March to keep  some leverage over the implementation of the president’s executive order on immigration
  • Brinksmanship with artificial deadlines and threats of shutdowns
  • Unrelated legislation such as the repeal of some Dodd-Frank swaps rules for banks
  • And a silly, wonky, inside-the-Beltway name like Cromnibus to tie it all together

But when measured at the admittedly low standards of recent history, it wasn’t all that bad.  First, passing 11 complete appropriations bills by Dec. 13 is pretty good. Last year (FY14), no appropriations bills were passed until Jan. 17 for a fiscal year that began on Oct. 1.  In FY13, the first appropriation bills were not passed until March (half way through the FY) and then Congress passed only five full bills while continuing to fund the other seven appropriation areas via continuing resolution (CR).   While the FY12 outcome was similar to this year’s (all bills were complete by Christmas), only the defense appropriation bill passed in FY11 and continuing resolutions funded the rest of the government.   So I’ll give Congress a star for (incremental) improvement.


Graph by Nolan McCarty

A few things from the figure are worth noting.  First, there has indeed been a slight uptick in performance but it comes after a 15-year decline.  The second takeaway, however, is that it has been far worse.  The mid-1980s are actually the low point.

As I have written, these changes in congressional performance are not easily explained.  A simple story about partisan polarization does not work well as Congress is much more polarized now than it was in the mid-1980s.  But ideological conflict across branches and chambers does explain some of the variation.  Fiscal and macroeconomic variables such as growth rates and (lagged) deficits also do not seem to matter much.  Performance began declining in the mid-1990s just as the economy heated up and deficits began shrinking.   The data hint that congressional rules such as statutory pay-as-you-go provisions (PAYGO) improved performance.

But have there been any real consequences of the decline in congressional performance on appropriations?  Here the preliminary evidence seems mixed.  The good news is that there appears to be no connection between the timely completion of appropriation bills and fiscal outcomes such as spending and deficits.   But more troubling is the impact that “governing by CR” has on uncertainty about government policy and performance.  Late appropriation bills not only contribute to uncertainty about overall spending levels but also uncertainty about specific programs.

As an illustration of the possible impact of poor, delayed appropriations on uncertainty, I use a measure of policy uncertainty developed by Scott Baker, Nichols Bloom and Steven Davis that utilizes media coverage of the economy. Their index is based on search results from 10 large newspapers.  An article is treated as an indicator of policy uncertainty when it contains the term “uncertainty” or “uncertain,” the terms “economic” or “economy” and one or more of the following terms: “Congress,” “legislation,” “White House,” `”regulation,” “Federal Reserve” or “deficit.” Because this policy uncertainty index has been shown to correlate negatively with investment and economic performance, a connection between it and appropriation delays could indicate the macroeconomic costs of poor procedural budgetary performance.

Using monthly data, I find a significant negative correlation between this measure of uncertainty and the proportion of completed appropriation bills.  A government operating on CRs generates an uncertainty index that is from eight to 15 points higher than one operating under 100 percent completed appropriation bills.  Using the findings of Baker, Bloom and Davis, I can quantify the economic magnitude of this effect.  Since they report that a 90 point increase in policy uncertainty leads to a 2.3 percent annualized decline in economic growth, a simple extrapolation suggests that the difference between zero and 100 percent completion is between 0.2 percent and 0.4 percent of annualized gross domestic product: a modest but noticeable effect.  It seems reasonable to speculate, however, that the magnitudes are much larger when potential and actual government shutdowns are part of the mix.  Fortunately, the pushes for a shutdown from the left and the right got little traction.

As for the other less savory aspects of the Cromnibus, I will again grade on a curve.  There is nothing terribly new about attaching unrelated provisions to “must pass” legislation. In fact, it seems that the Republican leaders extracted far less from this opportunity than they could have had they waited to pass the spending bills in January when they will have the majority in both chambers.  That said, I am particularly sad to see the Dodd-Frank provisions concerning swaps trading by banks repealed.  I fear we will see this tactic used much more in the future to undermine financial reform.

And the term “Crominbus”  — I like it.  Certainly no worse than “fiscal cliff” or “sequester.”