© 2002 The Edwardsville Journal of Sociology, Volume 2                                                                                                                      back to ejs volume 2 contents

http://www.siue.edu/SOCIOLOGY/journal/

This paper examines patterns and trends in racial housing segregation in the St. Louis metropolitan statistical area (MSA), using data from the 2000 Census of Population and Housing (U.S. Census Bureau, 2001). Two common measures of segregation are used in this paper, which focuses primarily on housing segregation between whites and African Americans. The reason for this focus is that, because the St. Louis area does not have the large Hispanic/Latino or Asian American populations that are characteristic of many other metropolitan areas, the overwhelming majority of the St. Louis area population is either white or African American. In 2000, 96.5% of the St. Louis metropolitan area's population identified their race as white only or black/African American only, with another 1.2 percent identifying themselves as being of mixed race (computed from U.S. Census Bureau, 2001). Of the latter, the majority were of combinations that included white, African American, or both. Only 2.2 percent of the population identified themselves as belonging exclusively to races other than white or African American. Approximately 1.5 percent of the area's population in 2000 was Hispanic or Latino, and a significant portion of this total is included in the 2.2 percent who indicated their race as other than white or African American. Most of the rest are included in the white total; according to Census procedures Hispanics are treated as an ethnic grouping and may be of any race. However, more than 96 percent of the St. Louis MSA population in 2000 was either white or African American and is not Hispanic. If those of mixed races including either white or black but not Hispanic are added, at least 96.5 percent of the St. Louis area population in 2000 was white, African American, or a combination including one of these races, and non-Hispanic.

This research as well as past studies of segregation in the St. Louis area is based on two types of indicators of residential segregation. The first, the index of dissimilarity (sometimes called the "segregation index") is a measure of the overall degree of segregation, that is, the overall degree to which any two groups live in separate neighborhoods. The index of dissimilarity can range in value from 0 to 100. A value of 0, meaning no segregation, indicates that, in every neighborhood area (usually a census tract or block), the racial composition is the same as that of the overall area. Thus, for example, since the population of the St. Louis metropolitan area is 18.3 percent African American, a segregation index for whites and African Americans of 0, based on tract data, would mean that every census tract in the metropolitan area would be exactly 18.3 percent African American. An index of 100, meaning complete segregation, would indicate that in every census tract, the population would be either all white with no African Americans, or all African American with no whites. Of course, these are abstractions, and the true level of segregation always falls somewhere between 0 and 100. However, the closer the index is to 0, the less segregated an area is, and the closer the index is to 100, the more segregated the area is. Specifically, the index of dissimilarity can be interpreted as the percentage of either racial group that would have to move to a different neighborhood in order to get the index down to 0 (Taeuber and Taeuber, 1965). The index of dissimilarity is commonly abbreviated as D.

The second type of index, the exposure index, is an asymmetrical index, meaning that separate indices are computed for each group's exposure to the other. For example, in the study of segregation between whites and African Americans, there is a white-exposure-to-African Americans index and an African American-exposure-to-whites index. White exposure to African Americans represents the proportion of African Americans in the average or mean white person's neighborhood (e.g. census tract), and African American exposure to whites represents the percentage of whites in the average African American person's neighborhood. This index is different from the index of dissimilarity not only that it is asymmetrical, but also in that it is influenced by the racial composition of the area under study (Lieberson, and Carter, 1982; Farley, 1984). For example, in an area with a large African American population, the potential for whites to live in areas with African Americans will be greater than in an area with a small African American population. Hence, exposure indices are influenced both by the degree of segregation in an area and by the area's racial composition. While some may see this as a disadvantage, this index also has an advantage: it can tell us more about the actual neighborhood situation of the average white person and the average African American person (Lieberson and Carter, 1982). Exposure indices are commonly abbreviated as p* . Specifically, _wp*_b represents white exposure to blacks or African Americans, while _bp*_w represents black or African American exposure to whites.

Past research, based on computations of the index of dissimilarity, has shown that the St. Louis metropolitan area is more racially segregated than the great majority of metropolitan areas across the United States. This has consistently been the case for the past several censuses. In 1970, for example, the St. Louis metropolitan area, as measured by the index of dissimilarity using census tract data, was the 14^th most segregated out of 237 metropolitan areas (Van Valley, Roof, and Wilcox, 1977). In 1960, its level of segregation measured 18^th out of 137 metropolitan areas (Van Valey, Roof, and Wilcox, 1977). Since then, the St. Louis area has been even closer to the top of the segregation rankings. By 1980, the St. Louis metropolitan area was the 10^th most segregated metropolitan area out of 318 nationwide (Jakubs, personal communication). In 1990, different studies using slightly different measures placed the St. Louis metropolitan area at 11^th or 12^th out of, again, more than 300 metropolitan areas (Farley and Frey, 1994, USA Today, 1991).

As recently as 1970, more than two-thirds of the St. Louis area's white population lived in virtually all-white census tracts, i.e. tracts with less than 1 percent African American populations (Farley, 1984, 1991a). Even in 1980, the majority of the area's white population lived in such tracts (Farley, 1991a).

Using data from the 1980 census, Massey and Denton (1989) identified the St. Louis area as one of just ten areas identified as "hypersegregated," that is, highly segregated on all or four out of five measures that capture different aspects or dimensions of racial housing segregation. Among large, racially heterogeneous cities in 1980, the city of St. Louis was the third most segregated city in the United States. Also in 1980, the index of dissimilarity for the St. Louis metropolitan area was 7 points above the median for 38 large metropolitan areas (Taeuber et al., 1984), and large metropolitan areas like the ones in this study have higher levels of segregation than the average for all metropolitan areas.

So persistent was segregation in St. Louis that, between 1940 and 1980, there was essentially no change in the level of segregation in St. Louis city. As measured by the index of dissimilarity computed with block data, the city's segregation level was 84.6 in 1940 and 83.8 in 1980 (Taeuber and Taeuber, 1965; Farley, 1983). There was a modest decline in the metropolitan area's level of segregation after 1970, but as noted, it remained well above that of most metropolitan areas in 1980.

By 1990, segregation was on the decline both metropolitan-wide and in the city of St. Louis. A practice census done in the city of St. Louis in 1988 in preparation for the 1990 census revealed that, for the first time since before 1940, a significant decline had taken place in the city's level of segregation (Farley, 1991b). The index of dissimilarity, based on tract data, fell by 4 points between 1980 and 1988. It then fell by an additional 1.7 points in just 2 years between 1988 and 1990, indicating an accelerating rate of decline as the decade of the 1980s went on (Farley, 1993). A similar pattern was observed using indices computed with block data. These findings raise a key question for research on trends during the 1990s: has the decline in segregation in the city that began and accelerated in the 1990s continued? Has the acceleration in the rate of decline continued?

Metropolitan-wide, a decline in segregation was also under way in the 1970s, and it, too accelerated during the 1980s (Farley, 1993). For example, based on census tract data, the index of dissimilarity for the metropolitan area declined by 3.6 points between 1970 and 1980, and by 6 points between 1980 and 1990. This again suggests an accelerating rate of decline, but again, the St. Louis area remained one of the most segregated in the nation in 1990. Nationally, the great majority of areas experienced declining segregation between 1970 and 1980, but between 1980 and 1990, the trend was more mixed: many areas continued to experience declines, but others experienced reversals and had increases in segregation (Jakubs, 1986; Massey and Denton, 1987; Farley and Frey, 1994).

A significant reason for the declines in segregation in the St Louis area in the 1980s is that there was a sharp decline in the number of areas from which African Americans were totally or almost totally excluded. As noted above, more than two out of three St. Louis area whites lived in census tracts that were less than 1 percent black in 1970, and the majority continued to do so in 1980. By 1990, however, the percentage of whites living in these essentially all-white census tracts had fallen to just under 36 percent (Farley, 1991a). The number of tracts with African American populations of less than 1 percent fell steadily, from 232 in 1970 to 178 in 1980 and 125 in 1990.

As this decrease in exclusively-white areas continued, interracial exposure in the St. Louis area increased, but only modestly. White exposure to African Americans - the percentage of African Americans in the average white person's census tract - rose after 1970, but only very modestly. This measure was .04 in 1960 and 1970, .05 in 1980, and .065 in 1990 - far below the 17 percent of the area's population that was African American by 1990. Reflecting the fact that nearly four out of five area residents are white, African American exposure to whites is higher - but not a great deal higher. It changed little between 1960 and 1980: .24 in 1960, .23 in 1970, and .25 in 1980. However, it rose to .306 by 1990 - again far below the 80+ percent of the area's population that was white in 1980 (Farley, 1991a). Thus, both white exposure to African Americans and African American exposure to whites increased, mainly between 1980 and 1990 - but both remain far lower than might be expected based on the region's racial composition.

New findings are beginning to emerge from the 2000 census. Most notably, Logan (2001) has computed indices of dissimilarity based on census tract data for non-Hispanic whites and non-Hispanic blacks in all metropolitan areas in the United States. Despite a decline of about 3 points in its segregation index by this measure, the St. Louis area remains near the top of the list of the most segregated areas in the United States in 2000. According to Logan, the St. Louis MSA was the 8^th most segregated among the 50 largest MSAs in 2000, and among all 331 metropolitan areas in the United States, it was the 13^th most segregated. We now turn to findings from the present study for a detailed analysis of patterns and trends within the St. Louis metropolitan area.

Using the American Factfinder, census tract data were downloaded for each county in the St. Louis MSA (U.S. Census Bureau, 2001). The data include the total population and the number of people in each census-defined racial group, including people who identified themselves as belonging to more than one race, which was permitted for the first time in recent history in the 2000 census. The indices reported in this paper were computed using the 2000 census tract populations of 1) whites, 2) blacks or African Americans, and 3) the sum of individuals who identified themselves as belonging to one or more race for whom one of the races they identified was black or African Americans. The index of dissimilarity was computed for the MSA and for each county with a population that is at least 3 percent black or African American. (Since D for a county is an indicator of segregation within a county, it is only meaningful if the county is racially diverse.) The index was computed for 1) whites and African Americans, 2) whites and persons of multiple races, one of which was black or African American, and 3) African Americans and persons of multiple races, one of which was black or African American. (Because of the small number of people of mixed race, the latter two indices were computed area-wide only.) For the counties with at least 3 percent African American population, D was also computed using block data. Use of block data will typically detect segregation that tract data does not, because unlike tract data, it detects patterns of segregation within census tracts as well as between census tracts. Additionally, exposure indices based on tract data were computed for the metropolitan area and for each county with an African American population of at least 3 percent. These indices included white exposure to African Americans and African American exposure to whites. Finally, the distribution of census tract racial compositions was computed. The trends between 1990 and 2000 in these measures were examined, for the whole metropolitan area and by county.

Consistent with Logan's (2001) findings using a slightly different measure, we find that there was a modest decline in the level of segregation in the St. Louis metropolitan area between 1990 and 2000. Table 1 shows detailed data on the racial composition of each county in the metropolitan area in 2000. Table 2 shows the percentage African American and the index of dissimilarity for whites and African Americans, computed using census tract data, for the metropolitan area and its constituent counties in each census year from 1970 to 2000. The index of dissimilarity for whites and African Americans, based on census tract data, fell about 3 points, from 76.9 in 1990 to 73.8 in 2000. Significantly, this is a smaller decline than occurred between 1980 and 1990, when the index fell by 6 points. Over the period of 30 years between 1970 and 2000, the segregation index fell by a little less than 13 points, with the largest decline occurring between 1980 and 1990. Though significant, this is a very slow rate of decline across this time period. At the rate of decline of the past 30 years, it would take it another 55 years to reach the midpoint between complete segregation and no segregation.

As noted above, in St. Louis City, there was no significant change in racial housing segregation between 1940 and 1980. Since 1980, however, segregation has declined at an accelerating rate, and this trend has continued through the 1990s, with nearly a 10-point decline between 1990 and 2000 alone. The decline in segregation in the city has been about 15 points in the last 20 years - a greater change in 20 years than has occurred in the overall metropolitan area in the last 30 years. Thus, after years of little or now change in the segregation level of the city, its level of segregation is now declining much more rapidly than that of the area as a whole. This trend can be seen in Figure 1.

In addition to the difference between the city and the rest of the metropolitan area, the trends in segregation have also varied in other parts of the metropolitan area. While the African American population has continued to suburbanize since 1990, some suburban areas have become less segregated while others have not. There are three suburban counties, in addition to the city of St. Louis, that are sufficiently diverse to meaningfully measure the degree of segregation within the county. In two of these counties - Madison and St. Clair Counties, Illinois, and the city of St. Louis, segregation declined between 1990 and 2000, as can be seen in Table 1. However, in St. Louis County, segregation actually increased between 1990 and 2000, so that today, whites and African Americans are more segregated from one another in St. Louis County than in the city of St. Louis or either Madison or St. Clair County, IL. The different trends in segregation among these four geographic areas are shown graphically in Figure 2. Twenty years ago, St. Louis County had the lowest segregation index of these four areas; today it has the highest segregation index. This is very significant, because with a population of just over 1 million, St. Louis County has a population that exceeds the combined poulation of St. Louis city, St. Clair County, and Madison County (about 860,000 combined).

In contrast, the decline in segregation has been particularly notable in St. Clair County, where the index has fallen from 89.6 in 1980 to just 65.6 in 2000. This is particularly significant because St. Clair County is a very diverse county - 28.8 percent of its population is African American. Thus, St. Clair County is both a diverse county and one that is rapidly becoming less segregated.

The reasons for these different trends in different parts of the metropolitan area are not entirely clear. However, some possibilities may lie in differing trends in housing discrimination in different parts of the area, which is a plausible possibility for several reasons. First, it is likely that real estate agents working in the Illinois part of the metropolitan area are better-trained in fair-housing law than those working in Missouri. This is because fair-housing training is a required part of the continuing education that real-estate agents must obtain in Illinois, but not in Missouri. There is strong evidence that steering is occurring in St. Louis County. Two series of discrimination tests conducted by the Metropolitan St. Louis Equal Housing Opportunity Council (EHOC) and the city of Florissant found extensive evidence of steering, with black and white home-seekers being told about housing in different parts of north St. Louis county. According to the St. Louis Post-Dispatch (Parish, 2001), these tests resulted in the payment of a substantial settlement to EHOC and to the city of Florissant. On the other hand, a series of discrimination tests I conducted in Madison County during a sabbatical leave in 1997 found little evidence of steering. In addition, some of the real estate agents who were tested made statements that reflected their fair housing training, such as "I sell houses, not neighborhoods."

It might also be noted that intense enforcement efforts were under way throughout the 1990s in St. Clair County, particularly Belleville. The city was under court supervision for discrimination as a result of a U.S. Justice Department lawsuit, after the police chief admitted to racial profiling on a national television show, and after a near total lack of diversity in the city's employees was documented by the Justice Department and by local media. Another fair-housing lawsuit was filed by the Justice Department against the city of Fairview Heights, which was settled in 2001. Also, a business was placed off-limits to military personnel by Scott Air Force Base after it was discovered that the business had engaged in racial discrimination. As illustrated by these cases, St. Clair County was "under the microscope" with respect to discrimination throughout much of the 1990s. The decline in segregation in St. Clair County may in part be a result of this increased attention and enforcement effort.

Table 3 shows the distribution of census tract racial compositions in 2000. Both whites and African Americans are less likely than in the past to live in areas composed almost entirely of their own racial group. However, the trends in this regard are somewhat different for whites than for African Americans. Among whites, the percentage living in census tracts less than 1 percent African American has fallen from nearly 70% in 1970 to about 31 percent in 2000. However, most of this decline had already occurred by 1990, when 36% of whites lived in tracts with less than 1 percent African American population. Among African Americans, the percentage living in tracts that are 95 percent or more African American is at a record low of under 28%. This percentage has fluctuated in the past, and was over 46% as recently as 1990. Today, unlike the past, most people in St. Louis do not live in neighborhoods that are almost entirely composed of others of their own race. On the other hand, most do not live in neighborhoods with large numbers of people of another race, either. 70% of whites live in areas that are less than 5 percent African American, and almost half of African Americans live in neighborhoods that are more than 80 percent African American.

The distribution of census tract racial compositions is mapped for the overall metropolitan area in Figure 3, for St. Louis city and St. Louis County in Figure 4, and for Madison and St. Clair Counties in Figure 5. In these figures (as shown on the legend), the bright green color represents tracts that are 10-24.99 percent African American. Since the area as a whole is about 18 percent African American, this is the range where we would expect to find most tracts were it not for the fact that housing is highly segregated in the area. As shown in Table 3, only 62 out of the metropolitan area's 525 census tracts actually are within this range. And as Figure 3, Figure 4, and Figure 5 show these tracts tend to be found in certain areas: south St. Louis city, the outer part of northwestern St. Louis County, northern St. Clair County including much of Belleville and Fairview Heights and the area around Scott Air Force Base, and a handful of tracts in Madison County, mostly around Edwardsville, Madison/Venice, and Alton. These are about the only places in the metropolitan area where the racial mix of the census tract comes anywhere close to the racial mix of the area as a whole.

Interracial exposure indices for each census year from 1960 through 2000 are shown for the metropolitan area in Table 4, and are broken down by county in Table 5. These exposure indices confirm that the residential neighborhood of the average white and of the average African American is more diverse than in the past. Since 1970, the percentage of African Americans in the average white person's census tract has risen from about 4 percent to about 8 percent. Similarly, the percentage of whites in the average African American person's census tract has risen from about 23 percent to about 34 percent. For both whites and African Americans, the change has mainly occurred in the 1980s and 1990s, and occurred at about the same rate during both of these two decades. These indices vary among the counties in ways that reflect both segregation patterns and racial composition. In St. Louis County, which reflects the area-wide racial composition but is highly segregated, the average African American lives in a census tract that is more than 40 percent white. Whites, on the other hand, are very isolated in St. Louis County: the average white lives in an area that is less than 10 percent African American, even though African Americans are 19 percent of the county's population. In Madison County, which is less segregated but also has a smaller African American population (just over 7 percent), these patterns are even more pronounced: the average African American lives in a neighborhood that is about two-thirds white, while the average white person lives in a neighborhood that is less than 6 percent black.

The situation is quite different in St. Louis city and St. Clair County. Like Madison County, these areas are less segregated than the metropolitan area as a whole. But unlike Madison County, their African American populations, at 29 percent and 51 percent respectively, are far above the area-wide percentage of about 18 percent. Given the large African American population and the lower levels of segregation of these areas, one would expect whites to have greater neighborhood exposure to African Americans, and that is the case. The average white person in St. Clair County lives in a census tract that is 13 percent African American, and the average white person in St. Louis City lives in a tract that is nearly one-fourth African American. These numbers are both higher than has been found in the past, and higher than some national research suggests whites are usually willing to accept (see, for example, Massey and Denton, 1993). In this regard, the numbers for these areas are encouraging. Average African American exposure to whites in these areas is also rising, and in 2000, the average African American in St. Clair County lived in a neighborhood just over 20 percent white, and the average African American lived in a tract just over 30 percent white. (It should be noted, though, that these are averages, which may be affected by the skewness of the distribution. In a forthcoming paper [Farley, forthcoming], I show that results using median rather than mean exposure are somewhat less encouraging.)

Further light can shed on the segregation pattern by comparing trends and patterns with segregation indices computed at the tract level and the block level. Block level data may detect patterns of segregation missed by tract data, because they detect patterns of segregation that may exist within tracts. A comparison of segregation indices computed using block and tract data is presented, by county, in Table 6 for each census from 1970 through 2000. Two key findings are evident in these data. First, the county trends between 1990 and 2000 at the block level are similar to those at the tract level: in both cases, St. Louis city, Madison County, and St. Clair County show substantial declines; St. Louis County does not. While at the block level St. Louis County shows a very small decline rather than the slight increase in segregation shown at the tract level, the overall conclusion is the same: there is no meaningful change in the level of segregation there between 1990 and 2000, while in the other three counties, there is a significant decline in segregation. Over the longer run, St. Louis County stands out even more clearly as having a different trend than the other three areas. Since 1980, there has been little decline in segregation there: most decline there occurred prior to 1980. In the other three areas, by contrast, segregation has declined at an accelerating pace each decade since 1970.

The second important finding is that, today, segregation as measured at the block level is less variable than segregation measured at the tract level. In all four areas, block-level segregation indices for 2000 were between about 74 and 77. Since the block level data present a more "close-in" measure of segregation, measuring it block-by-block within tract, it is possible that this consistency in the level of segregation may reflect some upper limit on the degree of racial diversity that whites will accept within their immediate neighborhoods. If so, this could further explain why there has been less segregation decline in St. Louis County than elsewhere: at the block level, St. Louis County was already near this level of segregation two decades, while the other three areas were substantially more segregated.

The trend in segregation, however, suggests that this may not be the case: despite their approach to this level of segregation, the rate of decline in all three of the other areas accelerated between 1990 and 2000. In effect, the levels of segregation at the block level are converging, but the rates of change are diverging: slowing to essentially zero in St. Louis County; accelerating in St. Louis city, Madison County, and St. Clair County. For this reason, the next census, in 2020, will be very telling. If the convergence in levels is the key process - i.e. if St. Louis County has reached a limit of the amount of integration that whites will accept - then the decline in segregation in the whole metropolitan area may slow to a snail's pace. But on the other hand, if the trend of an accelerating decline in segregation in St. Louis city, Madison County, and St. Clair County is the key process, we will might see something quite different in the next census. In that case, we could see a significantly less segregated city and Metro-East, with St. Louis County being left behind as a last bastion of racial segregation. One piece of data suggesting this rather more encouraging scenario is the fact that, at both the block and tract level, segregation in the city and Metro-East has already fallen below the level of segregation in St. Louis city. This suggests that the level there may not at all be a limiting factor in the degree of neighborhood diversity acceptable to whites, despite the convergence between 1990 and 2000. Of course, specific events in any of the above geographic areas could add other possibilities.

Finally, as expected, our results confirm that persons of mixed African American ancestry are less segregated from either the white or African American population than are whites and African Americans from one another. In the St. Louis area, 32,353 individuals, or 1.2 percent of the area's population, identified themselves as belonging to more than one race. Not surprisingly in an area with an above-average degree of segregation, this percentage is lower than the national percentage. In fact, it is just half of the national percentage of 2.4 percent who identified themselves as belonging to more than one race. Of those in the St. Louis area who did indicate more than one race, 12,235 individuals, or 37.8 percent of those who indicated more than one race, identified themselves as a combination of African American and some other race. We were interested in the extent to which these individuals experienced segregation from the white and African American populations. We anticipated that they would be less segregated from either group than the level of segregation between whites and African Americans who identify as only one race. We found that this was the case. The segregation index between whites and people who were a combination of African American and any other racial identity(ies) is 47.5 in the metropolitan area as a whole, based on census tract data. The segregation index between African Americans and people who were a combination of African American and any other racial identity(ies) is 39.2. Both of these indices are well below the white-African American index for the metropolitan area, 73.8.

These findings show that the overall level of racial segregation in the St. Louis area continues to decline at a very slow rate. The slowness of the decline is illustrated by the fact, noted above, that at the average rate of decline over the past 30 years, it would take more than half a century for the area's level of segregation to reach even the midpoint between total segregation and no segregation. Similarly, it would also take about that long for the level of segregation in the St. Louis area to reach what now is the average level of segregation for all metropolitan areas in the United States. Hence, while the St. Louis area is making small steps toward becoming a less segregated area, they are very small steps indeed.

A new feature emerging in the decade between 1990 and 2000 is the divergent trend in St. Louis County from the trends in other racially diverse parts of the metropolitan area. Racial segregation in St. Louis County, as measured with the index of dissimilarity computed with census tract data, actually increased between 1990 and 2000. In contrast, St. Louis City, Madison County, and St. Clair County all experienced declines of at least 6.5 points in their segregation indices. A separate analysis of census block data showed a similar divergence in trend: using block data, no significant change in the level of segregation was found in St. Louis County, while the level of segregation in all three of the other areas fell by at least 7.5 points. As a result of these changes, St. Louis County has been transformed, in the short time of a decade or two, from being the least internally-segregated of these areas to being the most segregated.

While the trends in segregation have been very different in St. Louis County and other parts of the metropolitan level, the levels of segregation have converged, particularly at the "closer-in" block level. At this level, St. Louis city and County and Madison and St. Clair Counties all are about three-quarters toward the segregated end of the scale when the index of dissimilarity is computed at the block level. The accelerating rates of decline in segregation in the city and in Illinois suggest that this convergence may be temporary, but the next census in 2010 will be telling in that regard. For the present, however, the St. Louis area remains one of the most segregated in the United States.

Table 2.  Percent Black and Black-White Housing Segregation Indices (Based on Census Tract Data), St. Louis MSA and Selected Counties, 1970-2000.

                                                   Percent Black			Black-White Housing Segregation Index (based on census tract data)
                                 ------------------------------------------------               ------------------------------------------------------------------
                                  1970          1980          1990         2000                  1970         1980          1990          2000
                                 ------------------------------------------------               -------------------------------------------------
St. Louis Metropolitan 
   Statistical Area               15.7          17.3          17.3          18.3                 86.5         82.9          76.9         73.8

St. Louis City                    41.0          45.5          47.5          51.2                 83.8         83.8          78.1         68.7
St. Louis County                   4.8          11.2          14.0          19.0                ------        72.4          68.4         69.5	    
Madison County                     5.2           5.7           6.5           7.3                ------        73.8          64.6         58.1
St. Clair County                  22.3          27.6          27.1          28.8                ------        89.6          79.6         65.6
St. Charles County                 1.3           2.3           2.7           2.7
Franklin County                    0.9           0.9           0.9           0.9
Jefferson County                   0.6           0.7           0.7           0.7
Jersey County                                    0.5           0.5           0.5
Clinton County                     0.9           3.0           3.9           3.9
Monroe County                                    0.1           0.1           0.1

Counties with no segregation index reported are not sufficiently diverse for segregation indices to be meaningful.

Source:  Data for 1970-1990 are from John E. Farley, 1993, "Housing Segregation in the St. Louis Metropolitan 
Area, 1980-1990: Comparing Trends at the Block and Census Tract Level."  Journal of Urban Affairs 15, 6 
(November): 515-527.  Data for 2000 are computed from the 2000 Census Redistricting Data. 

NOTE:  The data for 2000 are based on those people responding to the Census who reported one race only.  
In the St. Louis metropolitan area, 98.76% of the population in the 2000 Census reported one 
race only.  The 2000 segregation indices were computed by Professor John Farley of Southern Illinois 
University at Edwardsville.

Table 4.  Mean Interracial Exposure Indices, St. Louis Metropolitan Area, 1960-2000

              White Exposure                Black Exposure 			
              to Blacks (wp*b)              to Whites (bp*w)			
            --------------------          --------------------			
2000                .079                        .336				
1990                .065                        .306			
1980                 .05                        .25				
1970                 .04                        .23				
1960                 .04                        .24

Table 5.  Mean Exposure Indices by County, St. Louis Metropolitan Area, 2000

                 White Exposure            Black Exposure 			
                 to Blacks (wp*b)          to Whites (bp*w)			       
               --------------------      --------------------			
St. Louis Metro Area   .079		          .336			
St. Clair Co.          .130		          .308			           
Madison Co.            .055		          .675			
St. Louis City         .235		          .201			
St. Louis Co.          .099		          .401			

Table 6.  Trend in Index of Dissimilarity, Block and Tract Level, 1970-2000

Area/County                   2000	     1990       1988        1980       1970
                              ----	     ----       ----        ----       ----
St. Louis MSA       Block     ----     	     83.7       ----        88.0       91.3
                    Tract     73.8	     76.9       ----        82.9       86.5
St. Louis City      Block     75.0	     84.5       86.1        90.3       90.1
                    Tract     68.7	     78.1       79.8        83.8       83.8
St. Louis County    Block     77.0	     77.6       ----        79.9       87.0
                    Tract     69.5	     68.4       ----        72.4       ----
Madison County      Block     76.4	     83.9       ----        87.2       95.9
                    Tract     58.1	     64.6       ----        73.8       ----
St. Clair County    Block     74.3	     84.1       ----        91.5       92.1
                    Tract     65.6	     79.6       ----        89.6       ----

*Index of dissimilarity is reported only for counties with at least 3.0 percent black population in 2000.