Today is Tuesday. Suppose you want to save some money on Wednesday for something that you can only purchase on Thursday. But, you have a temptation problem. On Wednesday, you cannot resist spending money on frivolous things. Once Thursday rolls around, you have a bunch of frivolous things, and if you caved into Wednesday's temptation, you lack the resources to purchase what you really wanted to buy at the outset.
In economics, this is what we mean by facing a commitment problem. One solution to the commitment problem is to contract privately with a supplier of commitment devices. For example, Tuesday I can commit to bear an explicit cost Wednesday for caving into temptation.
Sometimes, however, private commitment is difficult to attain, a fact that has led some economists and policymakers to favor governmental nudges -- i.e., paternalistic policies that impose costs on individuals for caving in to certain temptations. Setting aside the issue of how to decide which temptations ought to be curtailed centrally, if individuals are susceptible to temptation, governmental nudges can improve well being. On this basis, a benevolent government would use nudges to improve overall welfare.
On the other hand, we don't live in a world with a benevolent government. Policies are the outcome of a political process. According to some new research, this fact changes whether we ought to use nudges.* The takeaway point: If individuals are susceptible to short-term temptation, politicians will indulge this behavioral flaw to be elected. Depending on the setting, the resulting policy can be much worse than if the government were constrained from using nudges.
There are some details to be worked out, but the core idea is an interesting critique of the use of behavioral economics in public policy.
*A paper by Alberto Bisin. Alessandro Lizzeri and Leerat Yariv.