One of my favorite sports commentators, Tony Kornheiser, often screams out to his co-host, Mike Wilbon, on their TV show "Pardon the Interruption": "Stop it, just stop it." He says this when someone says something that is over the top or unbelievable.

I often feel the same when I hear politicians talking about the supposed return on investment (ROI) the state or nation will receive from proposed public expenditures.

This new trend of reframing government programs as public investments is intended to put a businesslike face on the delivery of a wide range of public services.

My point isn't to question that some public investments are worthwhile and do, in fact, create future savings or facilitate economic activities that are beneficial. What I am concerned with are the economic projections often used to justify the claim.

Nowhere is this more of a concern than in debates about pre-K funding. One of the major problems is the time lapse between the educational boost a 4-year-old will receive by expanding his vocabulary, learning his colors and increasing his social skills and any public benefit or savings that will influence future government costs.

When it is stated that providing 4-year-olds opportunities to attend high-quality preschool will result, 10 to 15 years later, in the state receiving a 7-to-1 return on its money, my head starts to explode and I want to channel Tony Kornheiser.

When you think about the number of assumed facts you'd have to get right for that projection to even be in the ballpark, it's unbelievable. Here is a partial list: How fast will the economy and tax revenues grow over the next 15 years? Will our state and national unemployment rates increase or decrease (this fact alone will have significant impact on wage rates for every segment of the population)? What will the state and nation do regarding increasing or decreasing a wide range of benefits targeting the poor? Will our nation change its current policy regarding incarceration (which of course will have a significant impact on state and federal budgets)?

Perhaps most important, increases or decreases in the rate of out-of-wedlock births will affect levels of educational achievement, lifelong earnings and government expenditures. So, what will happen to that rate?

Finally, projections of state ROI from pre-K or any other social or educational funding are often based on the assumption of a high level of program implementation, requiring strong leadership, quality staff, supportive communities and stable funding down the road. Yet one of the things we know is that pockets of program excellence in education across the country are hard to maintain and replicate. The Harlem Children's Zone is an example of a good program that would be hard to replicate. It is run by a charismatic leader and receives major funding from some of the world's largest hedge funds. Those conditions are not likely to be found in many communities across the country.

There simply is no economic model that can take all of these factors into consideration and project out what an ROI would be for a state 10 to 15 years in the future. Even straight-line projections based on current budgets are rendered meaningless, because they are subject to huge fluctuations based on small changes in baseline assumptions that are impossible to predict. Even the Federal Reserve, with its big brains and even bigger computers, doesn't have a clear idea what the economy will look like next year, let alone 15 years from now. It routinely misses large swings in the economy.

This entire problem exists to a greater or lesser extent for projections of financial benefits by proponents of spending on sports facilities, higher education, transportation and various job-training programs.

The simple fact is that we only have a general sense of what state or national budget will look like in the future or what factors will cause them to grow or shrink. All we have are informed hunches that are often right, but sometimes wrong.

This is not to say that these hunches should be ignored. To the contrary, despite my caution regarding projecting government cost savings from "investments," I remain a cautious supporter of increased funding for pre-K, higher education and transportation. I just don't want to elevate our educated guesses into facts. That only undermines the public's confidence in government. After all, we would be hard-pressed to demonstrate a dollar amount saved from many past expenditures.

The next time a friend tells you that we need to fund a specific government program because it will provide a specific return on our investment, tell him: "Please, just stop it."

Peter Bell is a senior fellow at the Center of the American Experiment.