Why Didn’t ARRA Include Triggers?

One of the biggest political and substantive challenges the Obama administration has faced is that the scale of the economic collapse in Q4 2008 and Q1 2009 was much larger than the initial statistical estimates made by the government indicated. That led the White House to overpromise in terms of the likely trajectory of recovery under their stimulus. On one level, this isn’t Team Obama’s fault. On another level, though, you wonder why the team didn’t do more to account for the existence of data fuzziness.


In particular, they were faced with what they thought was an output gap of about $1.6 trillion and what they thought was a political landscape that didn’t allow for stimulus bigger than $800 billion or so. So as a matter of both policy and substance, why not try to write a smaller stimulus ($700 billion? $600 billion even?) but that built in large triggers for additional stimulus (payroll tax cuts, state government grants) in case bad data came in? I said the other day that to the best of my knowledge trigger stimulus was never considered, but Nick Bunker showed me this memo on stimulus policy options that says otherwise:



— stimulus could be divided into rounds with 2009 actions immediate and subsequent rounds triggered if economic indicators show prolonged recession.


— offsets could take effect after economy improves to avoid drag during recession—could apply to all stimulus or only later rounds.

Argument for triggers:

— puts in place stimulus large enough to address severe recession and restore confidence.

— avoids need for time consuming process if additional stimulus is needed later.

— prevents need to spend political capital on multiple rounds of stimulus

— might be more acceptable to Blue Dogs than larger package all at once

Arguments against triggers

— difficult to define triggers that will operate smoothly

— politically difficult to build consensus for actions that may or may not occur

— strong pressure to put items in first round since it is the one certain to take effect

One point I would make about this is that even when thinking about triggers they only thought about forward-looking data. The possibility of downward revisions to the data about the present and the recent past genuinely doesn’t seem to have been considered even though downward revisions are common during recessions. The other point is that this seems like a really good argument for triggers! All four of those arguments for triggers are totally sound, and putting in place a stimulus large enough to address a severe recession is a really important policy goal. The arguments against triggers, by contrast, don’t make a ton of sense. I don’t even understand what the “difficult to build consensus” bullet points means. And while defining workable triggers would, indeed, be hard it’s not like what ARRA actually did was easy. Writing a giant bill to try to spend a huge sum of money in a non-wasteful way across dozens of agencies is also hard. Life is hard.