TO:		Marc Edmonds, US EPA

FROM:	Will Smith, US EPA

DATE:		November 1, 2006

RE:		EJ analysis of the Phase II and alternative Phase II rules

This is an Environmental Justice (EJ) comparison of demographics of the
U.S. population with the population 1-mile proximate to facilities
reporting form R in RY 2003 and subject to either the Phase II proposed
rule or the Phase II alternative 5000/2000 lb rule. 

Data

RY 2003 frozen TRI data, released April, 2005. U.S. Census 2000 Summary
File 3 (SF 3) Block Group data. Census 1999 estimates of poverty.

Data Summary

In RY 2003 there were 21,489 facilities that reported 80,169 form R's.
The proposed Phase II rule would allow 7,644 of these facilities to have
used form A for 14,496 of these forms R's. The Phase II alternative
5000/2000 rule would allow 6,631 of these facilities to have used form A
for 11,971 these form R's. The form R’s considered are those that
could be converted under the proposed rule (or alternative rule) and not
under the status quo.

facilities reporting form R in 2003	form R's filed in 2003	Phase II
proposed 5000 lb	Phase II alternative 5000/2000 lb



facilities affected	form R's converted	facilities affected	form R's
converted

21,489	80,169	7,644	14,496	6,631	11,971



The group of Phase II 5000/2000 facilities is a subset of Phase II
facilities, and the same is true for the form R’s converted.

To study socio-economic characteristics of the population near these
facilities, proximate populations within 1 mile were derived using
Census block groups.  Block groups are clusters of blocks and may
comprise between 300 and 3,000 people, though they have an optimal
population size of 1,500. To derive the proximate population, all Census
block groups were examined. A Census block group is proximate if it lies
wholly or partially within a predefined distance from a TRI facility.
 A proximity distance of 1 mile was used. All block groups were run
through a proximity calculation, and if the block-group is proximate to
a TRI facility (meaning its centroid is within 1 mile from one or more
TRI facilities) it was designated as proximate. This formed a dichotomy
of block groups for population study.

The three proximate population groups that are discussed and compared
with the general U.S. population with respect to selected demographic
statistics are the following –

TRI proximate population – population within 1-mile proximity to
facilities that filed a form R report for RY 2003.

Phase II proximate population - population within 1-mile proximity to
facilities that filed a form R report for RY 2003, but could have used
form A instead under the proposed rule 

Alternate proximate population - population within 1-mile proximity to
facilities that filed a form R report for RY 2003, but could have used
form A under the alternative 5000/2000 lb rule.       

     

Associated with each facility is the population 1-mile proximate to the
facility. Each of the three population groups named is formed by
aggregating the facility proximate populations for facilities that
belong to the group with care taken to omit double-counting when
proximate areas overlap.

Analysis of Minority

In 2000, the percent minority in the U.S. population was 31.8%.

The percent minority for the TRI proximate population is 41.8% which is
higher than the U.S. rate of 31.8%.

The percent minority in the proximate population associated with the
Phase II proposed rule is 43.5%, which is slightly higher than that for
the TRI proximate population.

The percent minority in the proximate population associated with the
Phase II alternative rule is 43.8%, which is slightly higher than that
for the TRI proximate population but nearly the same as that for the
phase II proximate population.

The percent minority for each of the three proximate population groups
is clearly larger than the U.S. rate. The percent minority for both the
Phase II proximate population and the Phase II alternative proximate
population is slightly larger than that of the proximate population for
all TRI facilities. This difference from the general population appears
to be characteristic of populations proximate to all TRI facilities
rather than related to Phase II facilities.

Even though the percent minority in each of the proximate population
groups is higher than the U.S. rate, it should be noted that 72% of all
TRI facility 1-mile proximate populations have percent minority below
the U.S. rate; for Phase II facilities it is 70% and for Phase II
alternative facilities it is also 70%. Further analysis shows that this
statistic appears different from previous aggregated results and
deserves repeating: For any facility reporting a form R in RY 2003 the
percent minority in its 1-mile proximate population is twice as likely
to be below the U.S. rate. This last result appears to differ from the
results for the percent minority in the population 1-mile proximate to
all facilities. This happens because about 1000 facilities have large
minority proximate populations.

Analysis of Poverty

In 1999, 12.9 percent of the population was below the Census poverty
level. The percent below poverty within 1-mile proximity of facilities
that filed a form R report for RY 2003 was 16.5%. 

The percent below poverty within 1-mile proximity of facilities that
filed a form R report for RY 2003 but could have used form A instead
under the proposed rule was 17%. 

The percent below poverty within 1-mile proximity of facilities that
filed a form R report for RY 2003 but could have used form A instead
under the alternative rule was also 17%. 

Minority and Poverty Demographics of populations within 1-mile proximity
to a facility that filed at least one form R report for reporting year
(RY) 2003, versus the U.S. Population

 	Within U.S. Population	Within 1-mile proximity of all facilities that
filed a form R report for RY 2003	Within 1-mile proximity of those
facilities that filed a form R report for RY 2003, but could have used
form A instead under the Phase II proposed 5000 lb rule.	Within 1-mile
proximity of those facilities that filed a form R report for RY 2003,
but could have used form A instead under the Phase II alternative
5000/2000 lb rule.

% Minority	31.8%	41.8%	43.5%	43.8%

% Below U.S. Census Bureau Poverty Level	12.9%	16.5%	17.0%	17.0%



Potential Reporting of Less Detailed Data Under the Proposed Rule

Of the 7,644 facilities where the potential reporting of less detailed
data is at issue under this rule, 28% (2,129) of them could have 100%
reporting of less detailed data, that is, all form R’s could be
converted to form A’s. This represents 21% of the total potential data
at issue.  The aggregate proximate population for these facilities is
46.8% minority and 17.6% of this population is below the poverty level.
Some 409 of these facilities have no 1-mile proximate population.

There are a total of 1,786 facilities (23%) with no 1-mile proximate
population so the potential reporting of less detailed data near these
is somewhat mitigated. This represents 23% of the total data at issue.

Some 8% of facilities have high percent minority (80% or more) in their
respective proximate populations. This represents 9% of the total
potential data at issue.

The following table partitions the facilities into quadrants by
splitting percent minority on the 31.8% minority in the U.S. population
and percent below poverty on the 12.9% below poverty level in the U.S.
This allows one to focus on reporting of less detailed data by high and
low values of minority and poverty.

	Reporting of Less Detailed Data, Phase II proposed 5000 lb



Facility 1-mile proximate populations



Number form R's converted









% Below Poverty at or below U.S. level	% Below Poverty above U.S. level
No proximate population

	% Minority at or below U.S. level	          4,598 	          1,843 	 	
      6,441 

% Minority above U.S. level	          1,206 	          3,509 	 	      
4,715 

No proximate population	 	 	          3,340 	       3,340 

	          5,804 	          5,352 	          3,340 	      14,496 



For minority we find that 33% (4715 of 14496) of potential reporting of
less detailed data occurs in proximate populations where the percent
minority is greater than that in the U.S. population.  For poverty we
find that 37% (5352 of 14496) of potential reporting of less detailed
data occurs in proximate populations where the percent below poverty is
greater than that in the U.S. population.

We find that 24% (3509 of 14496) of potential reporting of less detailed
data occurs in proximate populations where both the percent below
poverty and percent minority are greater the U.S. rates, respectively. 
About 32% of the potential reporting of less detailed data occurs when
minority and poverty are both less than the U.S. rates, respectively.

Total potential reporting of less detailed data under the proposed rule
is 18.1% (=14496/80169).

Potential Reporting of Less Detailed Data Under the Alternative Rule

Of the 6,631 facilities where the potential reporting of less detailed
data is at issue under this rule, 26% (1,746) of them could have 100%
reporting of less detailed data, that is, all form R’s could be
converted to form A’s. This represents 20% of the total potential data
at issue. The aggregate proximate population for these facilities is
47.2% minority and 17.5% of this population is below the poverty level.
Some 329 of these facilities have no 1-mile proximate population.

There are a total of 1,493 facilities (23%) with no 1-mile proximate
population so the potential reporting of less detailed data near them is
somewhat mitigated. This represents 22% of the total data at issue.

Some 9% of facilities have high percent minority (80% or more) in their
respective proximate populations. This represents 9% of the total data
at issue.

The following table partitions the facilities into quadrants by
splitting percent minority on the 31.8% minority in the U.S. population
and percent below poverty on the 12.9% below poverty level in the U.S.
This allows one to focus on reporting of less detailed data by high and
low values of minority and poverty.

	Reporting of Less Detailed Data, 

Phase II alternative 5000/2000 lb

	Facility 1-mile proximate populations

	Number form R's converted









% Below Poverty at or below U.S. level	% Below Poverty above U.S. level
No proximate population

	% Minority at or below U.S. level	          3,876 	          1,486 	 	
      5,362 

% Minority above U.S. level	          1,038 	          2,948 	 	      
3,986 

No proximate population	 	 	       2,623 	       2,623 

	          4,914 	          4,434 	       2,623 	      11,971 



For minority we find that 33% (3,986 of 11,971) of the potential
reporting of less detailed data occurs in proximate populations where
the percent minority is greater than that in the U.S. population. For
poverty we find that 37% (4,434 of 11,971) of the potential reporting of
less detailed data occurs in proximate populations where the percent
below poverty is greater than that in the U.S. population.

We find that 25% (2,948 of 11,971) of the potential reporting of less
detailed data occurs in proximate populations where both the percent
below poverty and percent minority are greater than the U.S. rates,
respectively.  About 32% of the potential reporting of less detailed
data occurs when minority and poverty are both less than the U.S. rates,
respectively.

Total potential reporting of less detailed data for the alternative rule
is 14.9% (=11971/80169).

Conclusions

The EJ results for the proposed Phase II rule are very nearly the same
as those for the Phase II alternative rule. There appears to be some
disparity with respect to minority and poverty in population 1-mile
proximate to facilities when compared to U.S. rates.  This disparity is
probably associated with all TRI facilities not just those affected by
the Phase II proposed rule or its alternative.  This apparent disparity
is driven by about 1000 facilities in urban areas that have large
minority populations within proximity -- the percent below poverty
within proximity of these 1000 is generally much higher than the U.S.
rate.  For example, the percent minority in the population 1-mile
proximate to all facilities is larger than the U.S rate, but on a
facility-by-facility basis the percent minority in the facility 1-mile
proximate population is twice as likely to be below the U.S. rate.  If
these 1000 facilities were removed then the percent minority in the
population 1-mile proximate to the remaining facilities would be the
same as the U.S rate. 

In terms of potential reporting of less detailed data, the proposed rule
would result in more potential reporting of less detailed data (18.1%)
than the alternative rule (14.9%).

William P. Smith, PhD 

U.S. Environmental Protection Agency

Washington, DC 20460

202 566-0636

Smith.Will@epa.gov

