In his March 18 commentary ("Surplus sounds too good to be true — and it is"), William Melton expresses concern that no one is acknowledging that approximately $1 billion of projected government inflation is not factored into Minnesota's forecast budget surplus, making things look a lot rosier than they really are.

Melton is absolutely correct that this concern, which received a lot of public attention and commentary during the days of big budget deficits, has largely vanished like melting snow. That's because money to spend is now available. The sound of crickets shows that the "inflation-inclusion debate" has always been more about revenue raising than responsible planning.

Politics aside, should inflation be included in official budget projections? If so, the state's unofficial measure leaves a lot to be desired.

Contrary to common belief, the current practice of excluding expenditure inflation was historically the norm. According to the Minnesota House's Fiscal Analysis Department, the state only began adding inflation to the expenditure side of general fund budget forecasts in 1991. Importantly, back then the estimate did not adjust many general fund expenditures for inflation because it simply made no economic sense to do so.

But because today's estimate is only an "unofficial" projection, the state does not do this fine tuning. Our current $1 billion "unofficial inflation estimate" is based on applying the Consumer Price Index to all general fund spending across-the-board, making no attempt to exclude areas where inflation adjustments are inappropriate.

False impressions and distortions can also come from overestimating inflation. Accurate, intellectually honest budgeting practices require a careful job of identifying which general fund spending areas truly merit inflationary adjustments.

For example, it doesn't make sense to forecast inflation for one-time appropriations (such as capital investments) or costs that are fixed over time (such as debt service).

Other spending areas already include some sort of inflationary adjustment — mainly in Health and Human Services, which represents around 28 percent of general fund spending. For example, within this area inflation is included in the estimates of managed care rates. The problem arises when a general inflation adjustment is layered on top of adjustments like these already being made for changes in per-unit costs. Any rigorous inflation estimate must account for this, to make sure the forecast does not overstate the "real cost" of state commitments.

Still other items need to be carefully evaluated before being inflated in the forecast. Chief among these are the hundreds of millions of dollars the state provides directly to individuals via property tax refunds and to local governments through general purpose aids. Here, the state isn't providing a direct, tangible service where input cost changes can be evaluated. Instead, it either is trying to mitigate property tax costs for homeowners and renters or it is providing revenues for local governments to support cost structures over which they, and not the state, exercise primary control. Both types of spending involve a very vague relationship with cost trends.

Then there's the issue of government's unique ability to manage inflationary pressures in its largest input: people. While compensation is a relatively small piece of the state budget, billions in state aid funds compensation expenses for schools, counties, cities and other local governments.

Government inflation here is largely a function of collective bargaining agreements, and the issue of compensation inflation illustrates an important difference between the "forecast as a planning tool" and the "forecast as a budgeting tool."

As a planning tool, inflation-adjusted spending amounts could reasonably estimate the cost of keeping the existing government employee base's compensation whole. As a budgeting tool, cost-of-living adjustments would already have been included in the forecast as a "given" and it seems likely negotiations would tend to begin from that base.

Put this all together, and the issue of including inflation in the economic forecast is not as simple as many would make it out to be. Completely ignoring inflation's effects may be a detriment to responsible fiscal planning. But inflating the entire general fund using a projection of CPI, declaring it the "real cost" of current law and having the biennial budget debate framed in a context where every dollar of government spending is effectively indemnified against inflation doesn't serve the cause of acknowledging reality either.

Mark Haveman is executive director of the Minnesota Center for Fiscal Excellence.