ISPRS, Vol.34, Part 2W2, “Dynamic and Multi-Dimensional GIS", Bangkok, May 23-25,2001 
293 
Such additions are too significant to be neglected. 
(3) Accessibility of Low-Wage Workers 
Job accessibility for low-wage workers deserves special 
attention for two factors. The first is the addition of 
unemployment (a total of 39,079 in the study area) to the 
demand side of labor market. One may argue that workers of 
other wage groups may be unemployed too. It is perhaps 
reasonable to say that most unemployment competes for the 
low-wage jobs. The second is the impact of vehicle 
availability. Workers without vehicles have to rely on public 
transportation, carpool, walking or bicycles. For simplicity, al 
workers without vehicles are assumed to be low-wage 
workers 2 , and use public transits. “[Ajccessibility is a 
measure of supply, namely potential mobility, is not a 
descriptor of behavior.” (Salomon and Mokhtarian, 1998, 
p. 131) There are people who choose to use public 
transportation or walk or bicycle, not because of their 
financial limitation. For workers without vehicles, 
dependence on public transits means spending 110.4% 
more time than drive-alones for the same commute trips. 
Using superscript C or X to differentiate job accessibility for 
workers with or without personal vehicles respectively and 
superscript 1 to indicate the low-wage group (g=1), equation 
(5) can be written as: 
•w 
7=1 
(6) 
A} 
\x 
(7) 
where v) =YjS\-r k )w\d k p + r k w[(td kj r p ] for both (6) 
¿=i 
and (7) since both workers with and without vehicles 
compete for the same jobs. Note that W k 1 is the summation 
of low-wage workers and unemployment at TAZ k, t is the 
time ratio between public transit riders and drive-alones 
(t=2.1104 in this study), and r k is the percentage of workers 
without personal vehicles at TAZ k. In the Cleveland region 
in 1990, 3.75% workers have no vehicles. This is translated 
into 12.04% low-wage workers without vehicles. 
A combined job accessibility for low-wage workers 
is the weighted average of A 1C and A . 
A} =Q- n )Aj C + n A} X . (8) 
(4) Overall Job Accessibility 
Similar to equation (4), the overall (total) job accessibility is 
the average of all five accessibility measures weighted by 
the number of workers in each wage group: 
5. Findings 
examine Figure 3 and tell that one place has better job 
access than the other. Job proximity and accessibility 
indexes discussed in the previous two sections can be used 
to evaluate the spatial patterns. See Figures 4 and 5. 
Findings are summarized in the following. 
(1) Better Job Proximity for Lower-Wage Workers 
Job proximity simply reveals how far resident workers are 
from their suitable jobs in terms of physical distance. The 
index sets the baseline for job accessibility, and represents 
perhaps the commonly-perceived locational advantage. 
In first question is how job proximity varies among workers of 
various wages. In order to address the issue, the 808 TAZs 
(excluding 74 TAZs without resident workers) in the study 
area are grouped into eight categories based on the mean 
wage rates. Table 2 shows the number of TAZs, average job 
proximity and other variables (to be discussed) in each 
category. A smaller interval ($2,500) is used to divide TAZs 
with wage rates of $15,000-$25,000 because many TAZs fall 
in the range. From Table 2, the mean wage rate of $30,000 
is a turning point. Among the 706 TAZs with mean wage 
rates below $30,000, the higher the mean wage rate of a 
TAZ, the farther the TAZ is from the jobs (i.e., a larger D 
value). Specifically, for TAZs below $20,000, an average 
increase of $2,500 in mean wage rate leads to an increase 
of about 2-km distance from jobs; for TAZs between 
$20,000-30,000, an increase of $5,000 is associated with an 
increment of 1,5-2.5 km from job. Among the 102 TAZs with 
mean wage rates above the $30,000 level, the trend is 
reversed. In all TAZs with mean wage rates lower than 
$30,000, TAZs with higher wage rates are farther away from 
jobs. This trend is reversed in areas with mean wages above 
$30,000. 
Table 2. Job Proximity, Accessibility and Others in TAZs with 
Various Mean Wage Rates 
Mean 
Wage 
Rate 
No. 
TAZs 
Mean 
Job 
Proximity 
(D, in 
km) 
Mean Job 
Accessibility 
(A) 
Mean 
Commute 
Range 
(R, km) 
Mean 
Commute 
Time 
(minu.) 
< 
$15k 
84 
14.98 
1.0866 
7.51 
21.82 
$15- 
17.5k 
144 
16.97 
1.1194 
9.12 
22.19 
$17.5- 
20k 
137 
18.99 
1.0778 
10.07 
20.84 
$20- 
22.5k 
146 
19.96 
1.0666 
11.41 
20.91 
$22.5- 
25k 
108 
21.55 
0.9876 
12.01 
21.59 
$25- 
30k 
86 
23.03 
0.9897 
12.83 
21.85 
$30- 
40k 
64 
21.24 
1.0225 
12.63 
21.85 
2: 
$40k 
38 
18.23 
1.0404 
11.91 
20.82 
Total 
808' 
19.30 
1.0565 
10.72 
21.48 
'Among the total 882 TAZs, 74 have no resident workers. 
Figure 3 shows the job density pattern in 1990. In addition to 
the major job concentration in downtown Cleveland, jobs are 
scattered across the whole region with several noticeable 
suburban concentrations such as Lorain downtown, 
Cleveland International Airport, Solon Industrial Parks and a 
service office center in southeast. One cannot simply 
2 The CTPP does not have the information of vehicle 
availability by wage groups. The data of vehicle availability 
by household incomes indicate that only low-income 
households face the problem of no vehicles. 
In the classic urban economic model, locations of workers 
are determined by the tradeoff between transportation and 
housing costs. On the one side, higher-income workers 
could live farther away from jobs than lower-income earners 
if their saving in housing cost exceeds the increased cost in 
commuting. On the other side, the relationship could be 
reversed if the increase in commuting cost is more than the 
saving from housing. The above analysis supports that 
higher-wage workers live farther from jobs up to a certain 
wage level. Beyond that, high-wage workers value the time