II. Comparison of Costs
This section presents a comparison of costs across the three types of districts: Type 1, inhouse; Type 2, mixed districts; and Type 3, those that rely primarily on contractors. Two types of cost variables are examined over a fiveyear period from 1994 to 1998. The variables include cost per mile and cost per pupil. In addition, data is presented on the amount of reimbursement for operating expenses for the three district types.
Contracting Costs More Per Pupil
Table 2.1 and Figure 2.1 report the annual perpupil costs across district types (Types 13) over the fiveyear time period between 1994 and 1998. Three important findings are reported. First, the median perpupil costs in districts that rely exclusively on contracting (Type 3 districts) is significantly higher than the median per pupil costs for districts that rely on inhouse provision of services or purchase only a small amount of contracting. While the cost relationship remains relatively consistent throughout the time period, the gap between contracting and noncontracting districts fluctuates.^{12} For example, in 1994 the annual cost per pupil bused for the median noncontracting school district (the district at the center of the distribution) was $267. By contrast, the annual cost per pupil bused for the median school district that relied exclusively on contractors was $381, or 43 percent higher. In 1996, contracting districts’ median annual perpupil costs were 50 percent higher than noncontracting districts’ costs, while in 1998, the difference was 33 percent.
Table 2.1
Median annual cost per pupil for Ohio school districts 19941998^{13}
1994
1995
1996
1997
1998
Change
19941998Type 1:
Districts that use no contractors$267.45
$277.90
$289.78
$303.71
$312.74
16.9%
Type 2:
Districts that use some contractors$256.93
$268.33
$288.10
$316.08
$314.44
22.4%
Type 3:
Districts that use primarily contractors$381.30
$351.90
$435.15
$375.29
$416.10
9.1%
Percent difference between
Type 1 and Type 342.6%
26.6%
50.2%
23.6%
33.0%
Source: Ohio Department of Education, T1 Reports 19941998. The data excludes districts that did not report any inhouse or contractor busing activity.
Figure 2.1
Median Annual Per Pupil Cost, by Type, 19941998
Source: Ohio Department of Education, T1 Reports 19941998. The data excludes districts that did not report any inhouse or contractor busing activity.
A second finding is that the median annual perpupil cost for districts that purchased only part of the transportation through a contractor was similar to the noncontracting districts. In 1994, for example, the median cost perpupil in districts that used some contractor services was $257. Five years later, the figure had increased to $314 or $2 dollars less than noncontracting districts. The similarity in costs is largely explained by the fact that partial districts use only a small amount of contracting services, typically to transport a small number of privateschool children.
A final interesting finding is that the change in the annual per pupil cost for the median district between 1994 and 1998 was lower in contracting districts. After adjusting for inflation, the perpupil cost for the median contracting district was 1 percent less in 1998 than in 1994. By contrast, the perpupil cost of the median non and partialcontracting districts had increased 6 and 11 percent, respectively.
Although contracting districts were significantly more costly at both points in time, costs grew more slowly in the contracting districts.
Contracting Districts Cost More Per Mile
Table 2.2 and Figure 2.2 report the permile costs for the median districts across the three district types. The central finding is that the median district that relies primarily on contracting reports a much higher permile cost than the median noncontracting district. During each of the five years examined in the study, the cost per mile reported for the median contracting district was at least 50 percent higher than for the median noncontracting district. In 1994, for example, the median district that kept all transportation inhouse spent $1.85 per mile on pupil transportation, while the median contracting district spent $3.01 per mile, a difference of 63 percent. By 1998, the gap had shrunk but the median contracting district was still spending $1.11, or 53 percent more per mile than the noncontracting district.
Table 2.2 


1994 
1995 
1996 
1997 
1998 
Percent change, 
Type 1: 
$1.85 
$1.85 
$1.98 
$2.04 
$2.08 
12.4% 
Type 2: 
$1.91 
$2.04 
$2.03 
$2.11 
$2.15 
12.6% 
Type 3: 
$3.01 
$2.80 
$3.11 
$3.13 
$3.19 
6.0% 
Percent difference between 
62.7% 
51.4% 
57.1% 
53.4% 
53.4% 

Source: Ohio Department of Education, T1 Reports 19941998. The data excludes districts that did not report any inhouse or contractor busing activity.
Figure 2.2
Median Cost Per Mile, by District Type, 19941998
Source: Ohio Department of Education, T1 Reports 19941998. The data excludes districts that did not report any inhouse or contractor busing activity.
As noted in the discussion on costs per pupil, differences between noncontracting and partially contracting districts are minimal. The median partialcontracting district always spent less than ten cents more per mile than did noncontracting districts. Moreover, the differences over time across the three district types appear less dramatic than in the cost per pupil analysis. After adjusting for inflation, the cost per mile for the median contracting district was 4 percent lower in 1998 than in 1994, while it increased by 2 percent in both the median non and partialcontracting districts.
Local Taxpayers Bear Greater Cost in Contracting Districts
School districts that rely exclusively on contractors bear a higher percentage of their transportation operating costs than districts that provide transportation services inhouse or rely only partially on contracting. Table 2.3 compares the percentage of district costs covered by the state across each school district type from 1994 to 1998. It should be noted that due to time and data constraints, this study focuses exclusively on funds allocated to districts to cover a portion of their operating costs.^{15}
Table 2.3
Percentage of Costs Reimbursed by the State,
By District Type, 19941998
1994 
1995 
1996 
1997 
1998 

Type 1: No Contractors 
34.6% 
34.6% 
34.8% 
35.6% 
38.8% 
Type 2: Some Contractors 
31.7% 
31.8% 
32.5% 
33.3% 
35.7% 
Type 3: Primarily Contractors 
27.8% 
26.9% 
27.2% 
27.8% 
29.7% 
Percentage difference Between Type 1 and Type 3 
6.8% 
7.7% 
7.6% 
7.8% 
9.1% 
Source: Ohio Department of Education, T1 Reports 19941998. The data excludes districts that did not report any inhouse or contractor busing activity
Figure 2.3
Percent of Costs Reimbursed by State, 19941998
Source: Ohio Department of Education, T1 Reports 19941998. The data excludes districts that did not report any inhouse or contractor busing activity.
During each of the five years, median contracting districts saw significantly less of their costs covered than the districts that did no contracting. Because their actual costs were also higher, this meant that local taxpayers in contracting districts bore an even greater share of the burden. For example, in 1994 the median percentage reimbursement allocated to noncontracting districts was 35 percent. By contrast, the median amount for contracting districts was just 28 percent. Five years later, both types saw their reimbursement amounts increase. The median dollar amount reimbursed in contracting districts, however, remained 9 percentage points (and 23 percent) less than the median for noncontracting districts, as shown in Figure 2.3.
^{12}Type 3 always has higher costs, but mixed districts (Type 2) have lowest costs from 199496, and then second lowest in 1997 and 1998.
^{13}An analysis of variance (ANOVA) was conducted to see if the three group means for cost per pupil were statistically different. For three of five years the ANOVA test indicates that the group means were statistically different at the .05 confidence level. Since the primary comparison is between Type 1 and 3 districts an independent sample ttest was computed. In each case the test was statistically significant, indicating a strong difference in the group means for Type 1 and 3. The following are the F value and probability value for each year: 1998, F Value 4.27, Pr> 0.01; 1997, F Value 2.34, Pr> .09; 1996, F Value 4.84,Pr> .0082; 1995, F Value 1.76, Pr> .17; 1994, F Value 6.72, Pr> .0013.
^{14}An analysis of variance (ANOVA) was conducted to see if the three group means for cost per mile were statistically different. For four of five years the ANOVA test indicates that the group means were statistically different at the .05 confidence level. Since the primary comparison is between Type 1 and 3 districts an independent sample ttest was computed. In each case the test was statistically significant, indicating a strong difference in the group means for Type 1 and 3. The following are the F and probability values comparing three group means: 1998, F Value 11.75 Pr>0; 1997, F Value 2.26, Pr>.0736; 1996 F Value 8.87, Pr>.0002; 1995, F Value 3.28, Pr>.038; 1994, F Value 3.36, Pr>.035.
^{15}The state also reimburses districts for a portion of the cost incurred to replace and add new buses.