County Government Fleet Acquisition Practices: Service Demand or Budget Limitations

Abstract

Service provision through automobile use prompts the need for periodic replacement. Questions arise concerning service need versus popular trends. Any change has some type of cost impact and budget implications. The finance officer is charged with finding the balance between need and affordability. Very limited information exists concerning budget actor influence on fleet acquisition. This study examines fleet acquisition practices of professionally administered county governments. Findings indicate that the sheriff and departments can acquire specific makes and models, but only in conjunction with finance officer need assessment and the presence of appropriate facilities and financing mechanisms.

Keywords

local government finance county government finance personnel fleet acquisitions public vehicle replacement

Introduction

The use of all kinds of equipment among local governments is necessary depending on service demand. For governments with elevated population levels and especially in professional forms, the number and types of equipment are quite vast. It is entirely possible to have bulldozers, dump trucks, tractors, fire trucks, and specialized law enforcement vehicles all under one entity. Of course, automobiles are the primary vehicles of service provision for local governments with law enforcement usually the primary benefactors.

The budget strain associated with automobile acquisitions can create many problems for the finance officer. The financing mechanism or payment has to be determined and for how long if the purchases consist of traditional borrowing or even debt service. Further costs include maintenance and, in some cases, detailing. However, the initial determination has to be some type of need assessment. Although department heads as well as the sheriff have some basis for need, the finance officer usually has to implement the encumbrance in the original budget recommendation prior to any ratification. Therefore, the finance officer, especially more experienced officers, has some idea of specific needs as related to departments along with all cost impacts.

This exploratory study examines fleet acquisition practices of North and South Carolina county governments. Although there are numerous types of vehicles and equipment used to distribute and provide services, the focus of this article is primarily automobiles and pickup trucks that have gross vehicle weight ratings of less than 10,000 pounds. The primary question addressed is how the finance officer interprets service demand based on financial indicator policies that determine the availability of resources. Three different models will be introduced to evaluate these influences. Initial findings indicate that the finance officer indeed has a major role in vehicle acquisition as policy and funding drive the initial assessment of need, but once purchases are made, the finance officer usually satisfies specific requests due to available funding and facilities.

This study advances the literature in several ways. First, it demonstrates the challenge of balancing all types of requests with budgeting limitations and the alternate perceptions of service demand. Second, it provides the opportunity to explore county government budget orientation based on empirical indicators. Finally, it illustrates the importance of cost accounting and financial management when considering the variety of costs associated with fleet acquisitions.

The article is organized as follows: The “Previous Findings” section examines the literature surrounding the models that will be used to test for need interpretation and fleet acquisition. The “Data and Method” section provides a breakdown of the variables used for measurement followed by models testing for fleet acquisition methods including the amount of cash needed prior to initial purchase and the number of vehicles purchased at a given time. Finally, there will be a findings section with discussion.

Previous Findings

Theoretical Application

Finance officers are usually the first to approve some type of vehicle replacement that is requested by other senior government officials. As finance officers are providing most of the information for this study, budget orientation, more specifically control orientation, will be the primary theoretical construct. Budget formulators were first assigned some type of orientation by Allen Schick (1966). At the time, application of the initial framework (control) only applied to the federal government, and the focal point for budgetary decision making was the executive branch with spending limitations imposed by both the executive and legislature. The primary characteristic of control orientation is a primary concern for responsible financial management in which deficits are obsolete or kept minimized despite organization goals and objectives. A mixture of findings among state-level budget offices suggested that policy orientation was emerging as a result of gubernatorial policy initiatives (Thurmaier & Gosling, 1997). However, at the local level, control orientation appears to be rather prominent (Modlin, 2018; Thurmaier, 1995).

Finance officers among the states represented in the study are expected to have control orientation characteristics for very prominent reasons. First, in North Carolina, there is a heavy state oversight process that can, among other things, restrict spending due to the fund balance requirement and auditing requirements of the state oversight body (Coe, 2007; Coe, 2008; Modlin, 2012). Although state oversight is not as extensive in South Carolina, many elected county councils are extremely conservative with some boards investigating journal entries and ledger postings (Modlin & Stewart, 2014). Second, an overwhelming majority of finance officers have accounting backgrounds suggesting more emphasis on financial management and positive balances in all funds (Modlin, 2012, 2016a). For larger counties, findings indicate that accounting mechanisms such as internal service funds (ISF) are used for the motor pool or for county garages (GARAGE) to further isolate costs for fleet operations (Modlin, 2011).¹ Third, local governments have different thresholds of what is defined as a capital project (CP). Vehicles with capital project assignments may have to be approved separately as part of a capital improvement plan (CIP) causing delay depending on funding needs.

Fleet Replacement Findings

The amount of research dedicated to official influence on fleet acquisition has been relatively nonexistent. However, there have been several studies that have focused on standardized measures for general machine replacement (VEHCHG). For the most part, all of the studies take in differing levels of use (category of service), fixed and variable costs, life cycle costing, salvage costs, and numbers and types of machines purchased (Chen, 1998; Hopp, Zydiak, & Jones, 1993; Jones, Zydiak, & Hopp, 1991; Michel, Bell, Bronson, Owens, & Roylance, 2000). More recent studies have attempted to control for private sector budget constraints, asset turnover, and have attempted to determine if leasing is a more cost-effective option (Hartman, 2000; Hartman & Lohmann, 1997; Karabakal, Bean, & Lohmann, 2000). Among private sector organizations, an infusion of revenue could provide immediate funding for vehicle replacement, whereas for local governments, this usually takes place only at the beginning of the fiscal year after budget ratification and debited departmental appropriations. For local governments that have vehicle replacement policies (VRP), some type of uniform benchmark is used and incorporates variables introduced in the models above. In almost all cases, new vehicle requisitions (VEHACQ) are the result of high mileage and/or poor performance (REASONACQ). In addition, low utilization use is the subject of some policies while others further control costs by purchasing similar if not the same makes consistently to sustain inventory and avoid additional employee training (Contra Costa County, California, 2015; Pinal County, Arizona, 2013).

Organizational Influences

Vehicle replacement policies provide much information concerning ideal replacement scenarios, but service demand is the origin of costs with multiple departments vying for the same replacement dollars or to have vehicle request become part of the CIP. A recent study found that the sheriff or county law enforcement is the primary fleet replacement recipient. In most cases, performance is a primary driver in law enforcement fleet requests, so vehicle make requests (SHERIFFMAKE) have proven to be quite important (Modlin, 2016b). The same could also be true for other service departments depending on service category (DEPMAKE). As mentioned earlier, many vehicle replacement policies do have different replacement criteria based on category use (Pinal County, Arizona, 2013). With service demand, emergency service personnel and law enforcement, need is usually established due to increased calls as a result of population changes (POP). For county governments and these two parties in particular, an increased number of calls in more rural, large counties create further reduction in cost effectiveness (COMILES).

In professionally administered county governments, the finance officer is the primary budget formulator. County finance officers with substantial experience (EXP) have substantial knowledge of a variety of departments including some idea of vehicle purchase frequency and normal use. Finance officer experience has also been found to be significant in various areas of the audit process; thus, more experienced finance officers are aware of audit outcomes with purchases (Modlin & Stewart, 2014; Modlin, 2016a). No matter the request, proper funding has to be available (FUNDREQ). Committed fund balance monies are revenue sources in some cases (Gloucester County, Virginia, 2015). Previous research has determined that most southeastern county governments do indeed have a cash management plan (CMP; Modlin, 2012). Other financing needs of county personnel (COEMP) also have to be taken into consideration such as information system state mandates, various inspections, and the escalating costs of health care premiums along with claims payables. Therefore, finance officers representing these governments are capable of properly determining departmental fleet needs (FINRATE) and affirmation of requests.

Data and Method

Several types of data were used to analyze fleet acquisition practices. First, survey research was used to obtain finance officer rating of official needs (FINRATE), the amount of funding required prior for purchases (FUNDREQ), the usual number of vehicles acquired at a given time (VEHACQ), the presence of a vehicle replacement policy (VRP), the reason for the vehicle change (REASONACQ), specific make requested by department heads (DEPMAKE), specific make requests for the sheriff (SHERIFFMAKE), finance officer experience (EXP), the presence of a cash management plan (CMP), if fleet are designated as capital projects (CP), and if there is a county garage (COGARAGE) and internal service fund for that garage (GARAGEISF). After multiple rounds of survey dissemination, responses were received from approximately 34% of finance officers in both North and South Carolina. Responses from each state were approximately the same with every budget classification represented from units with total expenditures of less than US$25million to those of more than US$100million.

Alternate sources of data were used to obtain information for the remaining predictors. The U.S. Census Bureau (2016b) provided information for the number of citizens within each county for each state (POP) as well as the number of square miles for each county (COMILES) (U.S. Census Bureau, 2016a). Comprehensive Annual Financial Reports (CAFR) were the sources for the number of employees (COEMP) in North Carolina county governments and the South Carolina Association of Counties (2016) for South Carolina county employees.

Three different dependent variables are used to determine finance officer interpretation of need versus service demand. First, finance officer rating is based on a 5-point scale with “5” coded as really needed. The second dependent variable examines the level of funding required prior to purchases. At the most, counties determined that more than US$1million is needed prior to purchases with less than US$100K representing the lowest amount. The final dependent variable represents the actual number of vehicles purchased at any given time with the maximum amount representing more than 20. All of the variables and definitions are listed in Table 1.

Table 1.

Definitions of Variables for Measurement.

Variable	Definition	Measurement
FINRATE (Dependent)	Finance officer rating of official need	5 = really needed
FUNDREQ (Dependent)	The amount of cash required prior to purchase	5 = more than US$1million
VEHACQ (Dependent)	Average number of vehicles purchased at given time	5 = more than 20
VRP	Dummy variable for the presence of a vehicle replacement policy	1 = policy
REASONACQ	Dummy variable for how replacement is established	1 = miles only
DEPMAKE	Dummy variable for department head make preference	1 = make preference
SHERIFFMAKE	Dummy variable for sheriff make preference	1 = make preference
COMILES	Discrete variable for county number of square miles	4 = more than 750
POP	Discrete variable for population of county	5 = more than 100,000
CMP	Dummy variable for presence of a cash management plan	1 = yes
EXP	County finance officer experience	5 = more than 10 years
COEMP	The number of county government employees	5 = more than 1,000
CP	Dummy variable for autos classified as capital projects	1 = yes
COGARAGE	Dummy variable for county having own garage	1 = yes
GARAGEISF	Dummy variable for county garage internal service fund	1 = yes

There are virtually no models for fleet determination needs in the literature that encompasses actor influence. All of the collected data for this study have provided the opportunity to fill this void. Table 1 suggests that the dependent variables are based on an ordinal scale; therefore, three separate ordered logistic regression models are analyzed based on the predictors as a whole with the primary linear regression model listed below.

\begin{matrix} FINRATE = β_{0} + β_{1} V R P + β_{2} REASONACQ + β_{3} DEPMAKE \\ + β_{4} SHERIFFMAKE + β_{5} COMILES + β_{6} POP + β_{7} CMP + β_{8} EXP \\ + β_{9} COEMP + β_{10} CP + β_{11} COGARAGE + β_{12} GARAGEISF . \end{matrix}

The use of ordered dependent variables prompts the illustration of a model that can compensate for the multitude of responses resulting from predictor changes as well as changes within the actual dependent variable. For instance, the random disturbance term ( $\in_{i}$ ), which has a logistic distribution, suggests that FINRATE, for instance, could easily fall somewhere between four and five (really needed) versus having a precise measurement.

Y_{i}^{*} = \sum_{K = 1}^{K} β_{K} X_{k i} + \in_{i} = Z_{i} + \in_{i} .

Many of the discrete variables used in the study are on a 5-point scale with the exception of county square miles which is on a 4-point scale. The remaining predictors are dummy variables. In most cases, coding is based on the presence of the activity, whereas REASONACQ is coded based on miles and performance as the primary determinants of new vehicle acquisitions.

Result and Discussion

Budget formulation can become quite complicated when fleet purchases are concerned. Finance officers have to ascertain need based on departmental requests and information concerning various automobiles or, in the case of larger governments, fleet managers. In any case, the finance officer has some initial input on fleet needs as the purchase acquisition process begins with that office once budget ratification takes place. Table 2 examines finance officer assessment of need based on unit budget size. For the most part, there is agreement with other official assertion of need, but this is not absolute. More finance officers stated additional vehicles were somewhat needed more than really needed. There also appeared to be more urgency among finance officers in the lowest budget category. This could very easily be related to budget constraints. No finance officer reported that additional or replacement vehicles were not needed at all.

Table 2.

Finance Officer Need Rating by Budget Size.

Budget size	Not really needed	Need somewhat	Needed	Really needed
100M+	0	2	6	2
75M-100M	2	0	3	2
50M-75M	0	3	3	2
25M-50M	2	2	2	3
>25M	1	6	7	2
Total	5	13	21	11

Need interpretation can be highly subjective depending on the official. Finance officer experience level and knowledge of various departments had some expectation of influencing actual purchases. In Table 3, it appears that as experience level increases, the probability of purchasing multiple automobiles on a more frequent basis actually decreases suggesting that officials have to make a stronger case for replacement vehicles. In addition, most vehicle purchases consist of five vehicles or less compared with the other categories. Both tables also provide some support for control orientation theory. The tables demonstrate that finance officers as a whole do suggest vehicles are needed, but purchases are made in a selective manner.

Table 3.

Finance Officer Experience Level by Number of Vehicles Purchased.

Years experience	1 vehicle	2-5 vehicles	6-10 vehicles	11-20 vehicles	20+ vehicles
10+ years	1	5	1	0	1
7-10 years	0	3	0	3	0
5-7 years	1	2	4	2	4
3-5 years	0	4	1	2	4
<3 years	3	3	1	1	2
Total	5	17	7	8	11

Most of the counties in the study did not have an official vehicle replacement policy (Table 4). In addition, approximately half of the sample stated that vehicle miles were either the primary reason or only reason for a new vehicle. The remainder stated that poor performance and specification issues were also a problem. Departmental staff did not have a make preference whereas law enforcement personnel did have a preference. A recent study discovered that law enforcement personnel prefer either the Dodge Charger or Ford Interceptor (Modlin, 2016b). For county governments, the number of miles covered for service does have an impact on fleet replacement. For this sample, county size was just more than 500 square miles. Among the two states in the study, only five exceed 1,000 square miles.

Table 4.

Descriptive Statistics.

Variable	n	M	SD	Minimum	Maximum
VRP	50	0.32	0.4712	0	1
REASONACQ	50	0.50	0.5051	0	1
DEPMAKE	50	0.28	0.4536	0	1
SHERIFFMAKE	50	0.60	0.4949	0	1
COMILES	50	2.56	0.8369	1	4
POP	50	3.04	1.6031	1	5
CMP	50	0.69	0.4684	0	1
EXP	48	2.87	1.4013	1	5
COEMP	50	2.73	1.3659	1	5
CP	50	0.62	0.4903	0	1
COGARAGE	50	0.58	0.4986	0	1
GARAGEISF	50	0.10	0.3058	0	1

The organizational factors illustrated that governments within the sample represented various sizes. There was an average of about 700 county government employees per unit within the sample. Most units did have a cash management plan and a county garage, but not a service-related internal service fund. As mentioned earlier, internal service funds are indicative of larger governments and provide a more isolated examination of vehicles within actual departments and, in some cases, the actual vehicle itself.

Three different logistic regression models were used to evaluate finance officer need determination against available funding and acquisition models (Table 5). In the first model, REASONACQ and DEPMAKE had the strongest relationships with finance officer need rating. If number of miles was the primary reason for vehicle replacement, finance officers were 16 times more likely to state replacement was necessary. Indeed, finance officers representing governments with a VRP, which was also significant, stated high levels of need at a 75% rate compared with 62% without. Most of the mileage benchmarks stated ranged between 100K and 150K miles with some South Carolina counties going as high as 200K. The likelihood also increases substantially if department heads state a specific make preference. Interestingly enough, however, finance officer need assessment decreases by 1.4930 if there is no county garage.

Table 5.

Ordered Logistic Regression Models Determining Need.

Variable	Finance officer rating model	Funding required model	Acquisition model
Variable	Coefficient (OR)	Coefficient (OR)	Coefficient (OR)
VRP	1.5279 (4.6084)*	1.2417 (3.4613)	2.2446 (9.4367)**
REASONACQ	2.7963 (16.3849)**	2.0247 (7.5741)**	1.0312 (2.8046)
DEPMAKE	2.4969 (12.1445)**	0.4475 (1.5644)	−3.5895 (0.0276)**
SHERIFFMAKE	−1.3638 (0.2557)	−2.5300 (0.0797)**	3.5256 (33.9738)**
COMILES	0.9115 (2.4881)*	0.2093 (1.2328)	0.8484 (2.3358)
POP	−0.3046 (0.7374)	0.8879 (2.4301)*	0.8204 (2.2714)*
CMP	0.2009 (1.2225)	−0.8241 (0.4386)	−0.5501 (0.5769)
EXP	−0.2425 (0.7846)	0.0704 (1.0729)	0.2648 (1.3032)
COEMP	0.5296 (0.7846)	−1.2278 (0.2929)**	0.5035 (1.6545)
CP	−0.0320 (0.9685)	−0.1121 (0.8940)	−0.9131 (0.4013)
COGARAGE	−1.4930 (0.2247)**	0.5367 (1.7103)	2.0252 (7.5776)**
GARAGEISF	1.7076 (5.5160)	3.0967 (22.1257)**	0.9781 (2.6594)
Threshold 1	−0.4209	1.7172	3.8905
Threshold 2	2.1132	2.0991	9.0411
Threshold 3	5.4543	2.5322	11.0012
Threshold 4		2.7638	13.5496
N	48	48	48
Log likelihood	−37.1280	−45.6409	−36.1636
LR Chi-square (12)	22.97**	21.06**	54.44***
McFadden’s Pseudo-R²	0.2363	0.1875	0.4295

Note. Cell entries are unstandardized parameter estimates. OR = odds ratio; LR = likelihood ratio.

p ≤ .10. **p ≤ .05. ***p ≤ .001 (two-tailed test).

The next model which examines the predictors against available funding also had many significant findings. Again, REASONACQ was the most influential predictor as finance officers would need an extra US$250K if miles were the primary determinant for replacement. County size was also factor with this model as POP and a GARAGEISF were significant and positive indicating more idle cash was needed prior to purchases with higher levels of service populations and if there was an internal service fund for the garage. COEMP was significant and negative indicating more cash was needed for counties with slightly lower levels of employees. This finding could suggest that in many cases, other types of funding were made available to benefit county employees as a whole such as specific supplies and equipment or even something more necessary such as unexpected increases in health insurance claims. In this study, finance officers in counties with lower numbers of county employees stated that in most cases, less than US$100K or less was needed prior to a purchase. These counties are also more than likely to only purchase two or less vehicles at a time. The SHERIFFMAKE somewhat verifies this finding as additional funding is necessary when law enforcement does not specify a make preference. Both of the first two models were significant at the .05 level suggesting that the predictors provide reliable distinction between the differing levels of the dependent variables.

Actual vehicle purchases were tested against the predictors in the final model. The most compelling finding was SHERIFFMAKE. If the sheriff specified a make preference, the likelihood of additional five purchases increases by 34 times. In most cases, when there was a preference, there was an increase of only one to five cars. Interestingly enough, however, more purchases take place when department heads do not make a preference. VRP and COGARAGE were also significant and positive indicating that the presence of both increased the probabilities of additional purchases. The service population was also significant for this model indicating that with every additional 250K people, an extra one to five vehicles would be purchased.

The models present some interesting findings, but some conclusions could be the result of timing and proximity. For instance, the models suggest significant findings in make preferences for various officials differing according to the models. For department heads, it would be relatively simple to express fleet needs at any time due to finance office proximity and more intimate knowledge of the budget process. Furthermore, as the third model indicates, if there is no make preference, attainment becomes more probable. The sheriff is more likely to appeal first to the manager and later to commissioners in budget hearings if necessary. In the case of the third model, the make preference could represent a change in a more high-performing model just introduced, and the elevated number of purchases could reflect that change. A couple of respondents did state that law enforcement had an interest in particular vehicles due to suggested performance indicators.

Conclusion

This exploratory study has examined fleet acquisition practices of professionally administered county governments to contrast need determination with financial limitations. The models taken concurrently provide an insight into acquisition practices. In the first model, the finance officer determines that there is a relatively high level of need if there is a VRP in place with miles as the primary indicator of replacement need. Subsequently, in the third model, that same policy determines the number of vehicles acquired. Within this same model and information from the second model, it can be determined that sheriff make preference does determine the number of vehicles purchased, but because this variable was negative in the second model, these vehicles were already budgeted as either next fiscal year departmental spending priorities or capital budget priorities. Department head vehicle requests with no make preference (third model) are just added to the purchases. The county garage variable, also significant with two of the models, points to the urgency of continual repair for service use. The first model indicates a higher urgency need related to a lack of a garage suggesting contracting could be more time consuming and less service availability while additional purchases in the third model points to adequate resources in place for service and repair. The service population was also significant in two of the models illustrating public needs are a major influence as well.

The study also verifies the existence of control orientation at the county government level (Thurmaier, 1995). First, finance officer need determination was based on either some type of policy or other indicator that replacement was needed versus just a policy preference. Second, financial management sustainability was prevalent with responses. Some type of funding indicator was needed with replacement such as normal departmental spending or a capital budget acquisition. Table 3 provides additional evidence as more experienced finance officers are less likely to have multiple purchases at any one time. In addition, the results state maintenance and repair costs are also considered with the internal service fund finding.

There are also limitations with the study. First, North Carolina has a very extensive state financial oversight process that has a fund balance requirement as mentioned earlier while for the most part, South Carolina counties have conservative councils that encourage spending limitations with an emphasis on transparency. As this is the case, fleet spending would not be a priority in the event funding problems were to emerge. Second, these are professionally administered county governments with county administrators or managers that have significant accountability with overall budget and financial management responsibilities. Although the finance officer performs the initial budget formulation, the manager has final approval prior to distribution to the commissioners or council. With other forms of governments, such as the county executive or even commission, certain spending decisions made by elected officials may not consider all costs associated with various purchases and thus create a financial strain on the unit as a whole if unchecked. Third, fleet needs are highly subjective. The various VRPs discussed in this study along with many others had alternate mileage levels as benchmarks prior to replacement. Each party has a different assessment of need with respect to actual miles placed on a vehicle or even performance.

Some items of inquiry can also emerge as a result of this type of study. The first would be how taxpaying citizens of jurisdictions interpret fleet replacement need. In counties and jurisdictions that have lower per capita income levels, demand and need become questionable especially if the age of the fleet on average is significantly less than that of the county population as a whole. Second, commissioner or council opinion on need, if any, would be interesting. For the most part, elected officials usually do not express opinions on these types of purchases unless faced with budget cut decisions. Finding the balance between service need and demand can be difficult, but this study has pointed out that finance officers have met the challenge.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biography

Steve Modlin is an independent researcher with the National Coalition of Independent Scholars. His research focuses on U.S. local government budgeting and financial management practices.

References

Chen

(1998). Solution algorithms for the parallel replacement problem under economy of scale. Naval Research Logistics, 45, 279-295. doi:10.1002/(SICI)1520-6750(199804)45:3<279::AID-NAV3>3.0.CO;2-4

Coe

C. K.

(2007). Preventing local government fiscal crises: The North Carolina approach. Public Budgeting and Finance, 27(1), 39-49.

Coe

C. K.

(2008). Preventing local government fiscal crises: Emerging practices. Public Administration Review, 68, 759-767.

Contra Costa County, California. (2015, October). County vehicle equipment acquisition and replacement policy, and clean air vehicle policy and goals. Administrative Bulletin, p. 508.4.

Gloucester County, Virginia. (2015, September). County facilities, maintenance, repair, and replacement fund policy. Administrative Policy Section, p. 505.

Hartman

(2000). The parallel replacement problem with demand and capital budgeting constraints. Naval Research Logistics, 47(1), 40-56. doi:10.1002/(SICI)1520-6750(200002)47:1<40::AID-NAV3>3.0.CO;2-T

Hartman

Lohmann

(1997). Multiple options in parallel replacement analysis: Buy, lease or rebuild. The Engineering Economist, 42(3), 223-248. doi:10.1080/00137919708903180

Hopp

Zydiak

Jones

(1993). A further note on parallel machine replacement. Naval Research Logistics, 40(4), 575-579. doi:10.1002/1520-6750(199306)40:4<569::AID-NAV3220400410>3.0.CO;2-F

Jones

Zydiak

Hopp

(1991). Parallel machine replacement. Naval Research Logistics, 38(3), 351-365. doi:10.1002/1520-6750(199106)38:3<351::AID-NAV3220380306>3.0.CO;2-U

10.

Karabakal

Bean

Lohmann

(2000). Solving large replacement problems with budget constraints. The Engineering Economist, 45, 290-308. doi:10.1080/00137910008967554

11.

Michel

Bell

Bronson

Owens

Roylance

(2000). County vehicle services: Preventing wear, repairing tear. Popular Government, 65(2), 32-35.

12.

Modlin

(2012). “County Government Finance Practices: What Independent Auditors are Finding and What Makes Local Government Susceptible,” Journal of Public Budgeting, Accounting & Financial Management, 24(2), 558-578.

13.

Modlin

(2016B). “County Government Fleet Purchasing Practices: Financial Efficiency v. Need Assertion.” Journal of Public and Nonprofit affairs, 2(1), 15-30. doi:https://dx-doi-org.web.bisu.edu.cn/10.20899/jpna.2.1.15-30.

14.

Modlin

(2011). “The Declining Use of Internal Service Funds: How Local Governments Are Changing the Allocation of Indirect Costs,” Journal of Public Budgeting, Accounting & Financial Management, 23(2), 151-165.

15.

Modlin

(2016A). “Increasing Transparency and Efficiency: An Examination of County Government Note Disclosures.” Public Administration Research, 5(2), 59-69.

16.

Modlin

(2018). “Local Government Finance Officer Orientation Revisited: Actually, It’s All About Control.” International Journal of Public Administration, 41(3), 1-10.

17.

Modlin

Stewart

(2014). “Staff Involvement in the External Audit Process: Reassessing Independent Auditor Satisfaction Levels Among Professionally Administered County Governments.” Public Administration Quarterly, 38(2), 243-269.

18.

Pinal County, Arizona. (2013, October). Vehicle replacement. Pinal County Policy and Procedure, pp. 1-4.

19.

Schick

(1966). The road to PPB: The stages of budget reform. Public Administration Review, 26, 244-245.

20.

South Carolina Association of Counties. (2016). 2016 county profiles publications. Available from www.sccounties.org

21.

Thurmaier

(1995). Execution phase budgeting in local governments: It’s not just for control anymore! State & Local Government Review, 27, 102-117.

22.

Thurmaier

Gosling

J. J.

(1997). The shifting roles of state budget offices: Gosling revisited. Public Budgeting and Finance, 17, 48-70.

23.

U.S. Census Bureau. (2016a). Geography division based on TIGER/Geographic Identification Scheme. Retrieved from http://www.census.gov/geo/www/tiger/index.html

24.

U.S. Census Bureau. (2016b). Population estimates program. Retrieved from https://www.census.gov/programs-surveys/popest.html