Why Not the Latest Observations?
We use annual data for our computations, therefore, it is necessary to have an annual observation for both the initial year and the desired year. For the indices based on GDP, it is only after the year is over that GDP can be measured. For price indices, the annual observations are usually the average of monthly observations. It would not be valid to compare a monthly observation in the current year with an annual observation in an earlier year.
The annualized growth rate is the average growth rate measured over a year. In other words, it is the hypothetical constant year-to-year growth rate necessary to take the beginning-year value of a series to its ending-year value. To understand the construction of the hypothetical growth rate, see how the growth rates are constructed. Note that, contrary to intuition, this uniform growth rate is not obtained by averaging the year-to-year percentage changes in the observed value of the series. Because most economic variables grow over time, the annualized growth rate is a useful computation for comparing
rates of growth of a given series for various time periods and of different series over the same time period.
This point is illustrated by the following examples of how the growth rate of GDP can vary, depending on the time period of the computation.
- General knowledge of history tells us that growth during the Great Depression of the 1930s was slower than during the "economic boom" of the 1960s. Using the calculator, one can put numbers on the comparison. U.S. real GDP grew at only 0.9 percent per year during from 1931 to 1935, but 4.65 percent during the 1960s.
- Economic historians often ask: did the U.S. economy "take off" after the Civil War? The answer, from the calculator, is negative. The annual average growth of real GDP for 1840 to 1860 (the last antebellum year) is 4.9 percent; while from 1866 (the first postbellum year) to 1886, the growth rate is somewhat less, at 4.6 percent.
Samuel H. Williamson, "Annualized Growth Rate of Various Historical Economic Series" MeasuringWorth.