Even if stem cells do yield a large, near-term commercial payoff, there's another flaw in the Baker-Deal analysis: It doesn't account for the opportunity costs associated with waiting two decades for royalty payments.

The basic issue is that a dollar today is worth more than a dollar a year from now. This notion—technically known as the "time value of money"—is the reason banks pay interest on savings accounts. At an interest rate of 5 percent, a dollar today is worth $1.05 a year from now. Flip that around, and the "present value" of a dollar expected one year from now is only about 95 cents. These are particularly important calculations if you're looking at investment opportunities, since even fairly large returns are worth considerably less in today's dollars if the payoff doesn't arrive for decades.

You can probably see where this is going. Under the Baker-Deal assumptions, the bulk of royalty income flows in about 20 years from now, but the report doesn't discount those sums to their present value. When Berkeley economist Richard Gilbert applied a standard discount rate of 5 percent to the Baker-Deal figures—a conservative figure derived from the interest rate on 10-year Treasury bonds—he found the present value of expected royalty income is about two-thirds less than the report indicates.

The Baker-Deal analysis has proven overly optimistic in other respects, too. Gilbert also modeled the California program's actual royalty policy, which of course didn't exist at the time Baker and Deal did their study. The economist found that payments to inventors and other limits in the policy reduced the state's expected royalty income even further—to just 0.6 percent of the overall research budget in the worst case.

When I talked to him, Baker argued that the effect of discounting future royalties would be largely balanced out by other factors he and Deal also left out of the report, such as a potential increase in overall health-care usage. Of course, if people use medical services more frequently in the future, overall health-care costs are unlikely to go down, which might explain why the report didn't model that possibility in the first place.