Computer Simulation of Health Care Plan Won't Work
Bremner, John, St Louis Post-Dispatch (MO)
Perhaps America's worst Christmas present in 1993 was given by Maxis Business Simulations. With money from the Markle Foundation, the company produced a computer game called SimHealth, which allows you to design your own national health care system.
Building on computer simulations long used for determining how industrial products will work before they're built, Maxis has created a number of games.
One is SimLife, which allows the player to design genetic mutations. SimAnt does the same for insect colonies. SimCity allows you to try your skills in urban design.
Even tiny changes introduced by the player can create the most amazing and far-reaching effects - in species, ants, or cities. Innocent fun.
But a health care game? That's not so innocent. Consider:
First, players establish their basic priorities. The game allows two: one about the degree of liberty versus equality to build into the system, the other between efficiency and community responsiveness.
These are hardly the only basic choices to be made about health care, or even the most important.
Second, players are given penalties for introducing any inconsistencies into the program. They lose "political chips" for doing so, which translate into lost votes - and ultimately lost elections.
This happens to be the opposite of political reality - no small fault in a simulation.
Third, players are allowed to test any variations they introduce according to both optimistic and pessimistic assumptions about costs.
Thus, increasing certain kinds of health coverage can either raise demand for care by small or staggering amounts, depending on one's choice of assumptions.
As a game, it may be fun, though not particularly realistic. But as a serious policy tool, it's both a snare and a delusion.
Computer simulations may work for testing whether or not an airplane will fly, but they don't do well at all for the messy worlds of economics and politics.
Even standard computer runs by the Office of Management Budget and the Congressional Budget Office on the cost of specific changes in the tax code are discouragingly inaccurate. And they examine only simple numbers.
OMB and CBO wisely make no attempt to quantify the possible range of taxpayer reactions to those changes - never mind trying to calculate the effect of them on the rest of the economy.
They don't try because it can't be done. Policy makers may be the poorer for lacking such information, but they are forced by the limitations of human ability to do without. …