Exactly. It is a Lilien type of dispersion index. This isn’t actually how we calculate it, but it is essentially how the industry’s share of employment changes relative to total employment. So it is the growth in that industry’s employment relative to total employment. What we do is that we know that over the business cycle each industry has a typical pattern—durables employment goes down a lot during recessions whereas, say, health care doesn’t go down that much. Those types of changes aren’t typically associated with sectoral reallocation. That is, we don’t really think of those things that are typical over the cycle as being associated with an increase in the NAIRU. We want to take those out, so we take out the typical cyclical movements, such as manufacturing always goes down in a recession and finance is acyclical.