An alternative (or even complementary measure) to strict abuse detection is graduated benefit reduction. Graduated benefit reduction might prevent the moral ha

An alternative (or even complementary measure) to strict abuse detection is graduated benefit reduction. Graduated benefit reduction might prevent the moral hazard, but it does so at the expense of introducing all sorts of assumptions about the availability of work, and the likelihood of getting an offer and becoming reemployed again quickly. In fact, it may have a non-linear detrimental effect, because people who end up unemployed long-term will come to depend more on UI as more and more of their savings dry up. It stops being effective just about the time they need it to be effective. The UI system inherently sucks because of this; either people will game it, or it will be eroded too severely to be of any real use during really tough economic times. As an aside, to help my own thinking about this, how much does the average employer contribute per employee to UI benefits? This is really open-ended on purpose, I'm just kind of looking for ballpark numbers, from as many contexts as people are willing to offer them. :-)