CO-poverty map. The same is true of the HC-occupation and HC-rent index maps (Figs. 6-8 and 6-9). In all three cases—especially poverty and low rent—a great percentage of the people with the characteristic live in highly polluted areas. The pattern is less clear in the case of Figure 6-10, which relates HC to black residents. Significant areas where high percentages of black people live completely escape the HC pattern. This reflects the movement of middle and upper income blacks away from the central city.

A statistical summary of family income related to carbon monoxide concentration provides more background (Table 6-4). The four characteristics, after all, relate directly to income. Areas where over 15 percent of all households live in poverty will tend to have low median incomes. Also, since the professional-managerial group is the highest paid occupational category, a low proportion of them in an area would mean low median income. The same is true of low rent levels. Finally, since a much larger proportion of black families are poor than nonblacks, high concentrations of blacks, other things being equal, tend to mean low median income.

While around 90 percent of the $12,000 and over tracts fell in the lowest CO range, only 17 percent of the under $7,000 median family income census tracts did so. On the other hand, few of the over $12,000 median family income tracts were located in areas of 5 mg/m³ CO or more, while 36 percent of the under $7,000 tracts were in such zones, indicating substantial pollution. The federal standard, it should be remembered, is 10 mg/m³.

Preliminary mapping of poverty sections and CO levels for five other cities—New York, Chicago, Denver, Los Angeles, and San Francisco-Oakland—show similar results. High CO pollution areas and areas where more than 15 percent of the households are poor go hand in hand.

Jeffrey Zupan, in his book, The Distribution of Air Quality in the New York Region, cites similar evidence. He found that in a 21-county region centered on New York City, high income people (defined as those filing tax returns reporting more than $10,000 per year in 1966) were exposed to significantly less CO and HC than were middle or low income people (less than $3,000 in 1966). Zupan also found approximately equal exposures for middle and low income people. This result differs from our findings, but Zupan's methods of defining income and his way of calculating exposure were quite different, and the two results probably indicate highly technical differences in methods employed more than anything else.

The strong negative correlation between CO and HC pollution and income in the Washington area, and the strong suggestion that the correlation also exists in other cities, is paralleled by research into stationary source pollutants.