Expectations matter. A lot. Stable institutions in a context of trust are necessary for liberal prosperity. In this sense, the work of government is to manage expectations–to maintain a stable strategic framework within which plans can be formed and cooperative action successfully coordinated over the long term. That is why, for instance, I think quickly restoring a legal and administrative framework for functioning financial markets ought to be among the government's chief priorities.
Yet I'm extremely suspicious of what strike me as intellectually contentious, ad hoc interventions into the economy aimed at expectation management. Countercyclical economic mood-control initiatives seem to me inconsistent with the maintenance of a general framework of stable rules — that is, they don't take the importance of expectations seriously enough — while also smacking of illiberal state propaganda. It's hard to draw a principled distinction between framework and ad hoc rules, but I think it's intuitive enough. Steady, predictable expansion of the money supply seems like the good kind of expectation management, while fool-in-the-shower money supply management seems like the bad kind. Likewise, automatic stabilizers, such as the predictable increases in welfare and unemployment transfers during recessions, strike me as greatly preferable to freestyle panic-mode stimulus legislation.
I understand that on some theories, a spell of poor economic performance cannot be expected to end without government assistance, and the failure to open the monetary spigot and induce spending through tax cuts and government spending is to invite disaster. OK. Suppose we're all convinced by those theories. (I'm not, but suppose.) Wouldn't we be naive to ignore the opportunistic nature of democratic politics and to therefore fail to grasp how Keynesian theory plus economic decline will tend to invite a pork bonanza rather than timely, targeted, and temporary stimulus policy? Doesn't taking both politics and the seriousness of recessions seriously suggest rigging policy in advance to include more automatic stabilizers? Why not have it as standing policy to halve payroll taxes, or increase TANF payments a certain percentage, or what have you, whenever the NBER gets around to calling a recession? If you insist on infrastructure, why not plan out, far in advance, some well-considered but non-urgent infrastructure projects that can be switched on immediately upon the onset of a downturn?
Wouldn't this be better expectation management? People about to hit recession would know in advance what's going to happen, instead of being made to wait on the outcome of the unpredictable, volatile and often self-undermining political wrangling of opportunistic panic.
The objections I anticipate to auto-stimulus are (1) it can't work without the element of surprise and (2) that each recession is an unrepeatable precious snowflake that requires special initiatives targeted to its unique circumstances. About (1), why think the political process will reliably deliver the right surprise? About (2) the TARP stuff does seem like that. But so far, I don't see anything in Obama's bill that seems unique to the times. It's just a train wreck of haphazardly considered tax cuts and spending.