Prescott: “Theory Ahead of Business Cycle Measurement”
Economic theory accurately predicts the business cycle in a peacetime economy.
Business cycles are the result of unpredictable but upward trending technological variation in the economy.
Business cycles are the result of optimizing behavior in the marketplace resulting from technological changes.
10% variations in economic production over a short period of time are not unusual in cyclical economies, even when the change in marginal product of labor during the same period has been minimal.
Investment component of GDP is about 6 times as volatile as consumption.
The path of GDP growth is monotonically increasing over time. Business cycles represent variations from this path.
When uncertainty is added into household utility functions, and utility functions are treated serially, leisure/employment vary in equilibrium.
The elasticity of the labor supply is understated in aggregated data. Causes of understatement and distortion of labor supply elasticity are:
1) “non-convexities” in the “mapping from hours of market activities to units of labor services provided”. Non-convexities are additional costs of working in terms of the sacrifice of leisure. Examples include
1) the hours spent commuting (not leisure, not paid for in wages)
2) Part-time workers with same skills are paid less per hour than full time
2) The way GNP is measured can cause distortions in the modeled elasticity of labor. Example he uses is the fact that some maintenance activities could be counted as capital investments but are counted instead as consumption because of tax laws.
3) Model does not account for changes in the production possibility frontier between consumption and investment goods. Sometimes this can be just a change in the tax laws for the accounting of depreciation.
4) The way imports are measured can cause actual labor elasticity to be less than the model. Example he uses in an oil price increase is the equivalent of a negative technology shock because it results in less output per dollar of input. These effects are not included in the aggregate production function. His conclusion is that theory is good, but measurement techniques currently in use need to improve. When measurement techniques improve, it will be apparent that theory is accurate.