Quiz question: About 60% of Americans identify as White (non-Hispanic), followed by 19% Hispanic, 13% Black, and 6% Asian. Is this distribution fairly consistent by age?

Okay, if you've previously seen the graph, which you undoubtedly have, you already know the answer is no. See how many of your friends answer "yes" when you ask them. It is simpler to think about the population of the United States if we assume that the distribution is the same by age. The trouble arises when this oversimplification leads to a misperception of issues.

For instance, older people typically earn more money than younger people. Assume that the only age groups in our population are those of 25 to 29 and 60 to 64. Assume that those aged 25 to 29 earn $20 per hour, whereas those aged 60 to 64 earn $40. What is the typical hourly wage by race/ethnicity?

For Asians, the computation would be (22,155,597x0.07x20+ 21,056,253x0.05x40)/(22,155,597x0.07+ 21,056,253x0.05). In other terms, multiply the number of Asians in each group (i.e. 22,155,597x0.07 for the 25-29 group) by the corresponding hourly wage. Add these two values together and divide the total number of Asians in the two groups. The findings are shown in table 1 below.

These are significant differences in mean hourly wage in our example population, and they are all related to age distribution differences**. **In other words, every measure that is influenced by age, such as pay, crime, or marriage rates, must be adjusted to account for the variance in race/ethnicity distribution by age. If not, the results will very certainly be misleading, even if not on intention by the person doing the reporting.

You've been briefed.

Crime studies suggest it is highly skewed across the age distribution.