Those are the major ones. In any given year we may have reductions associated with the sunsetting of programs, for example, that are only financed for a finite period. If those programs end, we would potentially reduce our FTE count as a result of that.
Also, through fairly complicated formulae, we attribute to what we call our enabling services, such as finance and human resources and so on--administrative overhead--numbers of FTEs that are associated with major program initiatives. There can be some reduction in those as well, but the major impact is the 3,000 that we've referred to.