MapCruzin.com
RTK or Left-To-Wonder?
Recommended Reading

Acknowledgements: This research is funded by a grant from the University of California Toxic Substances Research and Teaching Program. We are grateful for Robert Nash Parker's generous help and advice with statistical analysis.

DRAFT: Please do not cite or quote without written permission. Contact Michael R. Meuser, meuser@mapcruzin.com or Dr. Andrew Szasz, szasz@cats.ucsc.edu

---

Introductory Notes: Class, Race and the Unequal Burdens of Environmental Degradation

"Environmental racism" has recently become one of the most important areas of concern to environmental scholars and activists, alike.

Of course, in some ways, concern about the unequal burdens of industrial pollution goes back a long ways. Class effects have been observed for more than a century. One can choose, here, to read Marx and Engels, Dickens, Polanyi, or any good textbook in American labor history.

Industrial pollutants are at their highest concentrations within the plant and on the grounds of the plant, but factory workers are only the canaries for the rest of the population when it comes to the ill effects of exposure to high levels of toxic pollutants. Communities adjacent to the plants are also exposed to high levels of pollutants. Here, too, the class effect is clear. Industries, like many other types of facilities that Popper (1987a, 1987b) has dubbed "LULUs," or "locally unwanted land uses," (halfway houses, prisons, housing projects, hospitals, airports, highways, power plants) tend to be located closer to poorer than to more affluent communities. Most liquid and solid toxic wastes, 80-90% by most estimates, used to be simply dumped on the grounds of the plants where they were generated (Friedland, 1981; Goldfarb, 1979; U.S. House of Representatives, 1975). Toxic discharges into the air or into surface waterways, too, are likely to have their greatest impact on the communities nearest the industrial plants that generate those toxic byproducts of production. Furthermore, until recently, off-site disposal facility siting was determined primarily by economic considerations; that tended to locate them in rural communities or in the poorest, most marginal urban environments (Anderson and Greenberg, 1984; U.S. GAO, 1978).

Given the high correlation, in the United States, between socio- economic status and race, poor communities also tend to have higher proportions of Blacks, Latinos, Native Americans. Knowledge of class inequities should have translated long ago into concern about race inequities, as well. In fact, research on race and environmental degradation is a relatively recent phenomenon. Late as interest was in coming, it is now a topic that draws substantial interest (Bryant and Mohai, 1992; Bullard, 1990; Citizens for a Better Environment, 1989; Commission for Racial Justice, United Church of Christ, 1987; Race, Poverty and the Environment, various issues).

Some researchers have attempted to study both race and class, to try to disaggregate class effect from race effect, study their interaction, assess their relative importance, if possible. Mohai and Bryant's review of the literature (1992) shows that, although some researchers have found race the most important factor, others class, and still others wee unable to say one way or another, the preponderance of the evidence, so far, suggests that, though obviously highly correlated with each other and difficult to disaggregate: (1) race and class have independent effects; and (2) race is the more important factor.

New Data: 1986 Superfund Reauthorization and the Community Right-to- Know

Mohai and Bryant point out, also, that almost all of these studies have been limited to examining only a couple of facets of environmental degradation. Overwhelmingly, existing studies are of ambient air quality or of proximity to toxic waste sites.

Patterns of differential proximity can now be examined with new, more comprehensive data that became available when Congress reauthorized the Superfund law in 1986, shortly after the industrial accident at the Union Carbide plant in Bhopal, India that killed 2000 and injured many others. At the behest of environmental groups, Congress inserted a new provision, Title III, the Community Right to Know, mandating that industry report annual releases of over 300 toxic substances. Although there are problems with this data -- it depends on self-reporting; it falls far short of the approximately 800 hazardous materials on the so- called California list -- the Toxics Release Inventory, or TRI, constitutes a major step forward in providing access to information about the toxic byproducts of industrial activity.

In this study, we combine this new source of data about toxic releases with federal census data to examine patterns of differential proximity to industrial toxics at the opposite end of the production cycle (i.e., at the production, rather than waste byproduct graveyard, phase). We use median income as an indicator for social class. The census provides the racial and ethnic composition (% white, % black, % latino, % asian) for each census tract. TRI data is not aggregated by census tract, but it is not difficult -- though certainly tedious and time-consuming -- to take the street addresses listed in the TRI dataset and generate census-tract-level data from it.

Problems With the Data

Before we begin, a note about the limits of our work here.

First, we should note that, even though TRI is a historically unprecedented leap forward in what is known about toxic emissions, TRI emissions are only a small part of the total toxic picture. Studies that use only TRI will leave out, first, the impacts of other phases of the industrial cycle: transportation, offsite storage, offsite disposal (the toxic waste sites for which the race differential is so thoroughly documented by Bullard and the United Church of Christ study). Leaves out consumer (lead paint) and postconsumer (household waste) toxics. Leaves out nonpoint-source pollution -- in the case of Los Angeles, a source of air pollution that likely swamps the point-source sources represented by TRI emissions data. We note here that, in this pilot study, we do not use the whole TRI data set. TRI includes toxic discharges into air, water, onto land, to publicly owned water treatment works, and toxics shipped off-site. Here, we limit ourselves to the air emissions reported in the TRI.

We have already noted problems with TRI data, itself, that not all chemicals known to be toxic are reportable under SARA/Community Right to Know, that the EPA depends on industry self-reporting, with all the attendant uncertainties, incentives to underreport, etc. There are also problems with the way we employ TRI data. We use total pounds, as if all toxics were equally toxic; in fact, there can be significant differences; some materials can be harmful in trace amounts; others may require much greater amounts to do harm. One ought to have a way of factoring in the relative toxicities of the 300 plus chemicals in question -- a weighing of the data by some measure of toxicity -- rather than simply adding pounds of stuff indiscriminantly. Furthermore, we do not model actual dispersal, which would require complex programs that take account of wind direction and speed to model typical plumes; instead, we are making extreme simplifying assumptions that model exposure as spherical, uniform, and, worse still, as extending only to the boundary of the census tract.

There are also serious problems with the 1990 Census: Undercounting, especially poor people of color and illegal immigrants, always a problem with the federal census, was in 1990 more serious a problem than ever. Some states and cities, as well as statistical experts in the Census Bureau, argued that the raw numbers are so off that they should not be used, but, instead, the Bureau should recompute the numbers using known methods to adjust for systemic undercounting. Although these efforts were eventually defeated, the critique is certainly correct and one should used these numbers with caution. Furthermore, the race categories employed by the Census are ambiguous and, to some unknown degree, unreliable (Barringer, 1993).

Finally, there are concerns about using the Census Tract as Unit of Analysis: First, one should note that the demographic characteristics of contiguous tracts are quite dissimilar. We inspected the characteristics of the tracts contiguous to 10 tract with highest emissions and found that their demographic and income profiles differed, at times significantly, from the characteristics of the tract they surrounded. Since pollutants disperse, emissions in tract A will find their way to quite different people living in contiguous tract. Were we to move to a more realistic modeling of exposure, we would have to include this information with the more accurate dispersion plume analysis already suggested above.

Toxic Los Angeles

Keeping in mind all these admitted limitations, we proceed with our analysis. We aggregate TRI emission data across the various substances by simply adding pounds of reported emissions and aggregate emissions by census tract, then seek to correlate emissions with census tracts' demographic and income characteristics.

Originally, we intended to do California as a whole. Comparatively speaking, California does not come close to the amounts of industrial toxics emitted in the States of Texas and Louisiana, but it still ranks as one of the States highest in total annual emissions (U.S. EPA, 1991).

Our first runs of the 1989 TRI California data soon convinced us to focus our attentions initially on Los Angeles County: County 37, which turned out to be Los Angeles, appeared to have half of all the reported toxic industrial emissions in California (see, in back of paper, pp. A1, A2, A3; and see the visual representations A4 - total emissions, and A5 - air emissions).

We then saw that the toxic burden is very unevenly distributed within the County, as well. 217 of these turned out to have some emissions; the other 1435 had none (A6). Further inspection showed that the emissions were distributed very unevenly among the 217 tracts that had some emissions. A7. A8 lists the 50 tracts that had the highest air discharges. The top 10 account for nearly one third of all the air discharges in the county (and, NB, about 18 % of all the air emissions in the whole State of California!).

Los Angeles County is also demographically diverse. Not forgetting the problems of using Census data, the 1990 Census found that 37.30 % of the population is latino, 10.68 % black, 41.01 % white (non-hispanic white). Among the census tracts, the percentage of latinos varied from 0 to 100 %; blacks from 0 to 95%; whites from 0 to 100%. Median incomes within the census tracts varied too, and widely of course, from $ 0 to $ 150,000. Finally, as everywhere in our nation, race and class are intimately connected in Los Angeles. See A9, A10, A11, A12.

The amount of emissions in the County, their nonrandom distribution, the size of the minority population, all suggested that this was a fruitful place to again examine the relationship between race, class and proximity to toxic emissions.

A preliminary look at the data

Our "raw data" are displayed on A13-24. A preliminary inspection of these dramatic graphs points to a clear class effect. The emissions are clustered in tracts that have median incomes from 0 to $ 90,000; the highest emissions are even more severely clustered in tracts with median incomes between $10,000 and about $ 60,000. Race effects are much less clear on these graphs.

Inspection of the Top 50. (Refer here to A8.) Looking at census tracts that had the most intense emissions, the evidence did not seem particularly promising. The top 6:

1. close to average white and latino %s; close to mean income
2. mostly white; above average income
3. mostly black; above average income
4. mostly white; above average income
5. higher than mean white; below average income
6. all white; very low income

Considering, instead, the top 50, the evidence for a race effect begins to take shape. 31 of the top 50 tracts had latino and black (combined) populations that exceeded the County mean. Still, this evidence is hardly dramatic.

Analysis

From here, we divided the analysis into two different tasks: First, we sought to examine differences between the 217 tracts with emissions and the 1435 with none -- analyzing the sheer presence/absence of pollution without regard to quantities. Second, we sought to relate demographic and income data to the quantities of emissions.

Race and the Presence/Absence of Air Emissions

1. Difference of Means Tests

First, a standard measure of the difference between two populations: we compared the mean of % of minority populations of tracts with air emissions with the mean of % minority of tracts without, then tested for the significance of the difference found. The results are on A25, A26, A27. The 217 tracts with emissions have a mean latino population of about 45%; the 1435 without emissions have a mean latino population of 32%, a highly significant difference. On the other hand, this difference is not found when we look at percentages of black population. A27 shows the results when latino and black are combined -- highly significant, again, as we would expect.

2. Comparing distributions

We next took census tracts with vs. without emissions and aggregated them by 10 % increments of the relevant racial compositions. The results are displayed in A28-A39. The findings are best understood by examining A30, A33, A36, A39.

Keep in mind that, for the county as a whole, 217 of 1652, or almost precisely 13 %, of all tracts have some air emissions. A30: The county average is exceeded where latinos constitute more than 30% of the population. A33: The finding for % black is more complex, harder to interpret. The county average is exceeded when tracts have black populations between 10-40 % and between 70-90 %. A36: When latino and black are considered together, the relationship is quite clear -- the county average is exceeded whenever the combined population is higher than 30%. Note, though, that the relationship is not linear, but curved -- it reaches a sustained peak between 50 and 80 %, then decreases rather sharply. A39 shows the corresponding pattern for whites in Los Angeles. Lower than the county average at all points above 60 %. Again, the line is curved, not linear.

Class and the Presence/Absence of Air Releases

We found significant class effects for Toxic LA, as well. The class effect is clear from a difference of means test -- A40. We again aggregated census tracts, this time by $10,000 increments of median income (A41-43). The relationship is strongly non-linear, with little at the lower end and the virtual disappearance of emissions above $ 80,000.

Race, Class and the Presence/Absence of Air Releases

Taken individually, both class and race are related to the presence of toxic air releases, though in complex, non-linear ways. We next attempted to examine race and class together, in order to begin to tease apart their interactions and/or independent effects.

1. Two-way distribution

First, we followed the same procedure as above. We aggregated the census tracts into increments, this time by both % race and income increments, then counted tracts and tract with emissions within each cell. This time, we defined larger segments, four % race segments defined by quartiles and six income segments, $0-20,000, ... 80,000- 100,000, and one final segment for tracts with greater than $100,000 median incomes.

The results are displayed in A44, 45, 46, 46a, 46b. A46 shows a number of interesting things: Absolutely, emissions cluster along a band that stretches from poorest, highest minority to moderate income white tracts. The highest concentrations are to be found at $20-40K, higher than average % minority. Relatively: Absolute numbers do not take account of the fact that some cells have many more tracts in them. It is more appropriate to look at the relationship between the number of tracts with air emissions and the total number of tracts within each cell. There are several ways to do that. A46a gives the magnitude of the difference between the observed number of tracts with air and the number of tracts expected to have air (number of total tracts multiplied by the county-wide average of 13 %). A46b is an analogue of the last columns in A34 and A41. Again, one compares the ratios within each cell to the county-wide average. That average is exceeded in five cells. Note how the ratio increases with %race in the three income strips with significant emissions. Note, too, the cell at $40-60K, highest %minority: although few in number, moderately well off minority communities have the highest likelihood of having toxics-emitting plants sited in them.

Finally, we note that the departure from expected frequencies is statistically significant (A46, again -- the chi square test has to be interpreted with some caution here).

2. Regression

We then did dummy variable regressions on the data divided into these same six income categories and four race categories. For income, the wealthiest category was used as the reference; for race, the most white quartile. See A47-49. Dummy variable regression on income alone shows one category, $20-40,000, significantly different from the highest. On race, all quartiles are significantly different than the whitest quartile. Most interestingly, when all the dummy variables are run together in the same regression, no income variable is significant; all race variables are significant. This suggests that, although class appears to be related to presence/absence, class is, in this case, largely a proxy for race.

(NB: This finding was identical to the result we got with a somewhat different regression procedure in which we took a stratified sample of zero emission tracts, combined that sample with all 217 of the air emission tracts, and regressed presence/absence of air emissions, coded as 1 or 0, on median income, median income squared (because we know from previous inspection that the relationships is definitely curvilinear); then on %race and %race squared; then on all four variables. Again, both income and race are significant when used separately as the independent variable. When both are used, %race stays significant, income ceases to reach statistical significance, though r squared is still greater when both race and class are included in the regression.

Quantities of Emissions

All of the above deals only with the presence or absence of emissions. As such, it sets equal the census tract with 10,000 pounds of emission and the tract with more than 4 million pounds. We are currently analyzing the relationship between the magnitude, not merely the sheer presence, of toxic air emissions and demographic and income characteristics.

The raw distributions are displayed in A50-A54. These scatterplots appear to tell a very different story than the one told above. The magnitude of the emissions seems to decrease with %minority. For income, the curve from $10-80K is still quite apparent, but, with the exception of the one extraordinary outlier at about $33K, the curve now peaks toward the higher end of that range.

In order to understand what these scatterplots are saying, we have first pursued a two-way aggregation analysis, along the lines of the analysis we did earlier with presence/absence (A44-46). A55 displays the total pounds of emissions in each of the cells. The class effect is still clear; the race effect much less so. We modified the data in A55 in two different ways in order to get a better understanding of the numbers. A55b divides the pounds in each cell by the number of tracts with air emissions in that cell. This is a measure of the magnitude of emissions when they actually occur. A55b could be used in conjunction with A46b: A46b gives the likelihood that a particular type of community will have emissions; A55b gives the average magnitude of emissions, if they are indeed present.

We are still in the process of interpreting the magnitude distribution data.

Discussion

Studies of differential exposure to pollution have found both race and class effects. Almost all of those studies have used ambient air pollution or proximity to toxic waste dumps as the indicator for pollution. In this study, we used newly-available data on industrial plant emissions to study these same issues.

We think, first, that we have shown that TRI data provide an important new window on questions of environmental equity.

To reprise, briefly, what we have done:

• Initial visual inspection of the Los Angeles TRI data, displayed together with race and income data from the 1990 census, suggested a strong class effect. A race effect was less apparent.
• Focusing on the simple presence/absence of toxic air discharges, we found support for both race and class effects, when taken separately. We noted that in both cases the relationships found were not linear. Everything in the literature leads us to expect that the poorer the neighborhood, the higher the amounts of toxics released in that neighborhood. At least in the case of this addition to the total burden of pollution, communities of working people of modest means are more affected than the poorest communities. That does not negate the likelihood that pollution stemming from other phases of the production/consumption cycle --waste disposal, for example -- are more likely to impact the lowest socio-economic strata. As always, though, the wealthy do very well in benefit/ cost terms, bearing the least burden while reaping the greatest rewards from economic activity.
• Further analysis suggested that, although initially class seemed the clearest effect, in fact race is statistically more important when both are considered together.

Our initial analysis of TRI air emissions for Los Angeles county seems largely to confirm what other scholars have found. In the second phase of our work, we are beginning to look at quantities of emissions. Our first look at this suggests more complex patterns, at first blush more difficult to interpret in terms of what is already in the literature on environmental racism.

Further Research

We believe, first, that TRI emission data can be used to examine the relationship between race, class and toxics in other locations.

Second, several simplifications in our model that make our study necessarily exploratory and tentative; it could be repeated, and improved, with certain additions and modifications. Instead of using "total pounds of materials emitted," as if all the releases are in some sense equivalent or uniform in their potential for harming environment and public health, it would improve the analysis if one would multiply discharge amounts with some sort of weighing factor that accounts for different materials' varying toxicities. Also, as noted earlier, we tacitly assumed uniform dispersion from the point source. It would be much better were we able to model potential exposure better by taking into account prevailing wind direction, the various dispersion rates of the different pollutants, and so on.

Third, we view statistical analysis as only a preliminary to historical analysis. Even when patterns of differential proximity are confirmed, we are not yet be in a position to know the cause or causes for those patterns. To sketch the extremes of what is possible: Poor, working people and people of color move into industrial zones because housing is cheaper and that is all they can afford. Polluting facilities are sited near communities of the working poor and communities of racial and ethnic minorities because those communities are politically and economically least able to resist them. Both arguments can be found in the literature. Both probably occur, and occur in some interactive combination.

We believe that the right way to resolve the causal question is to do two parallel histories, one of a geographic area's industrial development, the other of that area's shifting demographic patterns of habitation. With two parallel sequences, it would be possible to look at the sequences and synchronies of the two histories and develop a more dynamic understanding of their historical interrelationship.

Political Implications

When we make a map, a graphic or a table like A6, A20, or A46, we are doing what we might call "industrial/social geography." We are located industrial activity in physical and social spaces. Doing so allows us to study, as we have done here, the costs, the burdens, as opposed to the benefits (also quite real) of industrial activity.

These graphs and figures bring home again understanding of what we increasingly are all too aware of: you cannot separate jobs from environment. Oh, one can separate them. The two have in fact been kept separate historically -- with the results we see in these data: to a large degree industrial geography is pollution geography. Some believe that this separation should be continued, if not intensified: Reagan- era notions of "enterprise zones," the idea that one could encourage the reindustrialization of poor, minority communities by providing the incentive of relaxing environmental and labor standards... One does not have to have a terribly fertile imagination to see what kinds of TRI graphics those policies are likely to produce.

The national economy is mired in low-growth, no-growth mode. People, especially the people already most exposed to industrial pollutants, as shown again by this study, are increasingly desperate for jobs. Just look at Los Angeles, the topic of our pilot study: reeling from military budget cuts, recession and the deeper, structural reorganization of the national economy. Even the professional classes are feeling the effects; for the working poor, for people of color, we scarcely need periodic explosions of rage to be reminded of their circumstances.

The pressure for new jobs can only get more intense. And new jobs, I think, must mean reindustrialization -- one can get by only so long with more fast food franchises, more health care jobs, more financial services and more tourism.

It is exactly at these moments, when people just want to work, that we must continually make the connection between jobs and environment. If we don't, we might well provide people with the former while denying them the latter.

Specifically, it seems to me that the following are some of the things that we must advocate:

1. Strong and uniform regulation: Especially uniform -- contrary to the notion of the enterprise zone -- so that the exit option as a way of escaping regulation is closed and so that industry can't get concessions by playing one community (one state, one country -- e.g. NAFTA) off against another;
2. Source reduction;
3. Community Right to Participate: in plant design and plant operation. Also, financial and technical support (along the lines of Superfund TAG grants) that enable communities to actual do what the laws give them the right to do in terms of participation.

References

Anderson, Richard F. and Michael R. Greenberg 1984 "Siting Hazardous Waste Management Facility: Theory versus Reality," pp. 170-186 in Majumdar and Miller, eds., Hazardous and Toxic Wastes: Technology, Management and Health Effects.

Barringer, Felicity 1993 "So Who Are We? Ethnic Pride Confounds the Census," The New York Times, / /1993.

Bryant, Bunyan and Paul Mohai, eds. 1992 Race and the Incidence of Environmental Hazards: A Time for Discourse, Boulder, CO: Westview.

Bullard, Robert D. 1990 Dumping on Dixie, Boulder, CO: Westview.

Citizens for a Better Environment 1989 Richmond at Risk: Community Demographics and Toxic Hazards from Industrial Polluters. San Francisco: Citizens for a Better Environment.

Commission for Racial Justice, United Church of Christ 1987 Toxic Wastes and Race in the U.S.: A National Report on the Racial and Socio- Economic Characteristics of Communities with Hazardous Waste Sites.

Engels, Friedrich 1845 The Condition of the Working-Class in England, Moscow: Progress Publishers.

Friedland, Steven I. 1981 "The New Hazardous Waste Management System: Regulation of Wastes or Wasted Regulation?" Harvard Environmental Law Review 5(1981):89-129.

Goldfarb, William 1979 "The Hazards of Our Hazardous Waste Policy," Natural Resources Journal 19(1979):249-60.

Marx, Karl 1967 Capital: A Critique of Political Economy, Volume 1. New York: International Publishers.

Mohai, Paul and Bunyan Bryant 1992 "Environmental Racism: Reviewing the Evidence," pp. 163-176 in Bryant and Mohai, 1992.

Polanyi, Karl 1944 The Great Transformation: the Political and Economic Origins of Our Time, Boston: Beacon Press. Popper, Frank J. 1987a "The Environmentalists and the LULU," pp. 1-13 in Robert W. Lake, ed., Resolving Locational Conflict.

1987b"LP/HC and LULUs: The Political Uses of Risk Analysis in Land-Use Planning," pp. 275-287 in Robert W. Lake, ed., Resolving Locational Conflict.

U.S. Environmental Protection Agency 1991 Toxics in the Community: National and Local Perspectives. The 1989 Toxics Release Inventory National Report, 560/4-91-014, September. Washington, DC: Govt Printing Office.

U.S. General Accounting Office 1978 "Waste Disposal Practices -- A Threat to Health and the Nation's Water Supply," CED-78-120, Washington, D.C.: Government Printing Office.

1983 "Siting of Hazardous Waste Landfills and Their Correlation with Racial and Economic Status of Surrounding Communities," GAO/RCED-83-168, Washington, D.C.: Government Printing Office.

U.S. House of Representatives 1975 Waste Control Act of 1975. Hearings held by the Subcommittee on Transportation and Commerce, Committee on Interstate and Foreign Commerce. April 8-11,14-17. 94th Congress, 1st Session, Serial #94-28.

Urban Habitat Program, Earth Island Institute "Race, Poverty and the Environment," various issues.

List of Tables and Figures Toxic Los Angeles

A1, A2 California toxic releases, by county

A3 Los Angeles toxic releases, as % of California

A4 Map of California -- Total Toxic Releases, by county

A5 Map of California -- Air Emissions, by county

A6 Map of Los Angeles: Presence/Absence of Releases, by census tract

A7 Map of Los Angeles: Releases by census tract

A8 LA: Census tracts with 50 highest air emissions, with income and race characteristics

A9 Demographics of LA census tracts: median income x %latino

A10 Demographics of LA census tracts: median income x %black

A11 Demographics of LA census tracts: median income x %(latino + black)

A12 Demographics of LA census tracts: median income x %white

A preliminary look at the data

A13 Income x %latino, all tracts

A14 Income x %latino, tracts with air emissions

A15 Income x %latino, 50 tracts with highest air emissions

A16 Income x %black, all tracts

A17 Income x %black, tracts with air emissions

A18 Income x %black, 50 tracts with highest air emissions

A19 Income x %(latino + black), all tracts

A20 Income x %(latino + black), tracts with air emissions

A21 Income x %(latino + black), 50 tracts with highest air emissions

A22 Income x %white, all tracts

A23 Income x %white, tracts with air emissions

A24 Income x %white, 50 tracts with highest air emissions

Race and the Presence/Absence of Air Emissions

A25 Difference of Means Test for %latino

A26 Difference of Means Test for %black

A27 Difference of Means Test for %(latino + black)

A28 %latino: frequencies of all tracts, tracts with and without air emissions; ratio of tracts with air emissions/all tracts

A29 %latino: ratio of tracts with/without air emissions; chi square test

A30 %latino: graph of ratio of tracts with air emissions/all tracts

A31 %black: frequencies of all tracts, tracts with and without air emissions; ratio of tracts with air emissions/all tracts

A32 %black: ratio of tracts with/without air emissions; chi square test

A33 %latino: graph of ratio of tracts with air emissions/all tracts

A34 %(latino+black): frequencies of all tracts, tracts with and without air emissions; ratio of tracts with air emissions/all tracts

A35 %(latino+black): ratio of tracts with/without air emissions; chi square test

A36 %(latino+black): graph of ratio of tracts with air emissions/all tracts

A37 %white: frequencies of all tracts, tracts with and without air emissions; ratio of tracts with air emissions/all tracts

A38 %white: ratio of tracts with/without air emissions; chi square test

A39 %white: graph of ratio of tracts with air emissions/all tracts

Class and the Presence/Absence of Air Emissions

A40 Difference of Means test for Median income

A41 Median income: frequencies of all tracts, tracts with and without air emissions; ratio of tracts with air emissions/all tracts

A42 Median income: ratio of tracts with/without air emissions; chi square test

A43 Median income: graph of ratio of tracts with air emissions/all tracts

Class, Race and the Presence/Absence of Air Emissions

A44 All tracts

A45 Tracts with air emissions

A46 Tracts with air emissions, observed and expected; chi square

A46a Difference between observed and expected
A46b Ratio of tracts with air / all tracts

A47 Dummy variable regression -- class

A48 Dummy variable regression -- race

A49 Dummy variable regression -- race and class

Class, Race and Quantities of Emissions

A50 Pounds of air emissions per census tract, by %latino

A51 Pounds of air emissions per census tract, by %black

A52 Pounds of air emissions per census tract, by %(latino + black)

A53 Pounds of air emissions per census tract, by %white

A54 Pounds of air emissions, all census tracts, by median income

A55 Total pounds of air emissions, by demographic and income characteristics of tracts

A55a Pounds divided by total number of tracts
A55b Pounds divided by number of tracts with emissions

Privacy, Fair Use and Disclaimer - Advertise on MapCruzin.com

Home | Free GIS | Downloads | Parks & Public Lands | Books | Environmental Justice | News Archives
Free GIS Tutorial | Consulting | TRI 2004 MAPS | Recycle Reuse Business | Toxics Explorer
North Coast GIS | Contact/About Us | Redwood Ecotours | Global Positioning | EnviroRisk Map Network
Climate Collapse | Free GIS Tutorial | What is GIS? | Right to Know | Reimagination | Health & GIS | Shrubbed | Search

---

Website maintained by Michael R. Meuser

Copyright © 1996-2009 MapCruzin.com
All Rights Reserved
Visit us for Free GIS Maps and Resources