Benefits and burdens: a study of city and county long-term debt in Georgia [Feb. 1994]

enefits

AStudy of City &County Long-Term Debt In Georgia.

Benefits

Benefits' &Bu,rdens

.Burdens

GEORGIA DEPARTMENT OF
COMMUNITY AFFAIRS

Burdens

r
February 1994

fit
Benefits Benefits

AStudy of City &County Long-Term Debt In Georgia

Benefits &Burdens

GEORGIA DEPARTMENT OF
COMMUNITY AFFAIRS
1200 Equitable Building 100 Peachtree Street
Atlanta, Georgia 30303 (404) 656-3879
An Equal Opportunity Employer
If you are disabled and would like to receive this publication in an alternative fonnat, please contact the Georgia Department of Community Affairs at (404) 656-3879 or 1 (800) 651-7954 (TOD).

Burdens Burdens
r

Findings

1

I. Introduction

3

II. Methods and Limitations of the Data

4

Incompatible Fiscal Years

4

Inconsistent Reporting

4

Special Circumstances

4

Consolidated Governments

4

Authorities

5

Population Figures

5

Trend Analysis

5

III. Local Government Debt In Georgia

6

General Obligation Bonds

6

Revenue Bonds

7

Lease

7

Other Debt

8

IV. Trends And Uses Of Debt

9

Trends & 'lYpes of Debt

~

9

Uses of Debt

12

Conclusion

13

V. Debt Management

14

Long-Term Debt Benchmarks

14

(1) Overall net debt exceeding 10 percent of assessed valuation

14

(2) Overall net debt exceeding $1,200 per capita

15

(3) Overall net debt per capita exceeding 15 percent of per capita personal income

16

(4) Debt service exceeding 20% of operating revenues

16

Revenue Debt

17

Conclusion

18

VI. Conclusion

19

This report analyzes city and county longterm debt from 1986 to 1992. The purpose of the report is to look at debt trends among cities and counties and review how long-term debt is being used. The report also utilizes a series of credit industry benchmarks to determine if there are debt management problems among cities and counties. The following are some ofthe highlights from the report and the corresponding page numbers where a more thorough discussion of the topic may be found: As a group, Georgia's cities and counties have
relatively little debt. Very few communities exceed any of the ICMA/credit industry benchmarks evaluated in the report. (p. 18) Using the credit industry benchmarks, Georgia's cities and counties generally appear to have a significant amount of unused debt capacity. (p. 18) One study of state and local government debt burdens in the U.S. showed that in 1989 Georgia had less "full faith and credit" debt and more public non-guaranteed debt (which does not require voter approval) than the U.S. average. Since 1987, both cities and counties in Georgia have continued the national trend to proportionately less full faith and credit debt. (p.6) The most heavily used type of long-term debt for cities and counties is revenue debt. In 1992 almost 75% of all municipal debt and about 56% of all county debt were revenue bonds. (p.7)
Between 1986 and 1992, long-term debt for Georgia's cities and counties grew rapidly. County per capita general obligation (G.O.) and revenue debt increased 96%, from $130.11 to $254.98. Per capita G.O. and revenue debt grew at a slower but significant pace for cities, increasing about 31 %, from $654.54 to $857.85. (p. 9)
The five counties with the highest per capita long-term debt in 1992 were in the Atlanta metropolitan area. Statewide, 52 counties, or about one-third, reported no outstanding longterm debt in 1992. (p. 10)
Unlike counties, the cities with the highest per capita long-term debt were not in the Atlanta metro area but were smaller cities elsewhere in the state. Many of these cities are in resort

areas and/or in the rapidly growing northern

portion of the state. Statewide, 146 cities

reported no outstanding long-term debt in

1992. (p. 10)

The City ofAtlanta's airport debt is the single

largest public project for which city/county

long-term debt has been issued; over $849

million, paid for with proceeds from airport

operations.(p. 10)

For counties, the growth in debt is a function

of an increased emphasis on capital

expenditures over the past seven years. While

per capita county operating expenditures

increased by about one-halfbetween 1986 and

1992, per capita county capital expenditures

more than doubled. (p. 10)

The increase in municipal Using the credit industry

debt is a reflection of a benchmarks, Georgia's

general increase in total per capita expenditures, rather than an increased emphasis on capital investment.

cities and counties
generally appear to have a
significant amount of

Generally, per capita unused debt capacity

operating expenditures for

cities have increased at a faster rate than per

capita capital expenditures. (p. 11)

For both cities and counties, the most frequent

use of debt is for water and sewer facilities.

In 1992, cities and counties had combined

long-term debt of over $1.6 billion for water

and sewer facilities. (p. 12)

None ofGeorgia's 159counties have G.O. debt

loads that approach the Constitutional cap of

10% ofassessed valuation. The highest percent

of G.O. debt to assessed valuation among the

counties is 3.06%. Two cities are approaching

the Constitutional cap of 10% of assessed

valuation. Seven municipalities have net debt

(G.O., lease, and "other long-term debt")

which exceed the benchmark of 10% of their

net bond digest. (pp. 14 & 15)

Six municipalities have net debt exceeding

$1,200 per capita; however, four of the cities

do so because of"other" water and sewer debt

for which they also have system revenues.

(p. 15)

1

Two municipalities have net per capita debt exceeding 15% ofper capita personalincome, although one had sufficient water/sewersystem revenues to pay the principal and interest on that debt. (p. 16)
Eight counties and 34 municipalities have debt service expenditures exceeding 20% ofgeneral operating revenues. Of those jurisdictions, 4 counties and 17 cities have readily identifiable circumstances which made their debt service levels look abnormally high. (p. 16)

One hundred eleven jurisdictions were unable to cover the debt service on their water/sewer system long-term debt with net system revenues for three consecutive years (1990, 1991, and 1992). In 1992, 56 of those jurisdictions had shortfalls (net revenues minus debt service) equivalent to 10% or more of general fund revenues. The majority of the 56 jurisdictions, about 64%, had less than 1,000 population, which might indicate a need to address the financial health ofthe state's small utility systems. (pp. 17 & 18)

2

City and county government in Georgia is a $6.2 billion business, employing approximately 100,000 people with a total payroll of $1.72 billion and an additional $372 million in employee benefits. 1To those who worry about the cost and intrusiveness ofgovernment, these numbers may be cause for alarm. For others, these same numbers represent the foundation of the state's community life. And in fact, local governments in Georgia, as indeed all governmental entities, represent a trade-offbetween benefits and burdens.
The benefits we receive from Georgia's general purpose local governments contribute much to our quality of life. We look to our cities and counties in emergencies for police and fire protection and ambulance services. They furnish potable water and treat our sewage so that it may be safely returned to the land. They haul away our garbage and wrestle with the dilemma of its ultimate disposal. They provide us with recreation and leisure time opportunities, from the library's children's reading hour to festivals and sports. Counties and cities are an important component ofGeorgia's economic growth through their provision ofservices and infrastructure - our streets, airports, parks, water and sewer systems.
At the same time, these general purpose local governments (like all government) are supported by tax dollars and fees for services. This side of the service equation might be thought of as the burden of local government-the financial resources used by government which therefore are not available for private purposes such as personal spending or corporate investment.
The benefit and burden aspects of long-term public debt are a particularly important component of the public service equation. Public debt is important because it represents a major part of public capital investment. While all capital spending is not financed with debt, major investments such as construction ofairports,jails, libraries, and water and sewer treatment facilities generally are.
An example ofthe necessity for debt financing
I The total number offull-time and part-time employees ofthe 126 counties and 420 municipalities reporting in the 19?2 Wage and Salary Survey conducted by the GeorgJa Department of Community Affairs equalled 97,n7. Expenditure figures are from the 1992 Report ofl.ocal Government Finances, also published by the Georgia Department of Community Affairs.

is the well-documented crisis of U.S. public

infrastructure. The decay of our infrastructure is

underscored by dramatic incidents such as sewer

mains collapsing and day-to-day hassles such as

highway and airport congestion and unsafe

bridges. Any solution to this crisis depends on

public sector investment, much of which will have

to be financed by general purpose local

government debt.

At the same time, debt represents a long-term

burden on taxpayers. Local governments and

their citizens have to live for

decades with each capital

investment and debt decision.

And since there are political

and practical limits on tax

revenues available to local

governments, all spending

decisions represent trade-

offs-operating expenditures

and current services versus

capital investment and debt;

park facilities or libraries versus jails.
With over $4 billion in

"A man must properly pay the fiddler. In my

long-term city and county debt case it so happened that

in Georgia in 1992, the stakes are high for these decisions. That $4 billion in debt is good newslbad news for Georgia's

a whole symphony
orchestra had to be subsidized. n

citizens; it signals not only

JOHN BARRYMORE

important investment in public

ON DEBT

infrastructure, but it also repre-

sents a significant allocation of tax and service

charge dollars.

For these reasons, it is critical that we under-

stand local government long-term debt. Among

the questions that must be answered are:

What are the debt trends among cities and

counties?

What type of debt is being used and for what

purposes?

Is debt being used responsibly by local

governments?

From this review, policy implications for both

state and local governments will be discussed. The

ultimate objective is to provide a better under-

standing of the benefits and burdens of city and

county debt in Georgia.

3

,.

This study relies primarily on data from the

annual Report ofLocal Government Finances,

which is based on a survey administered by the

Georgia Department of Community Affairs

(DCA). The survey is sent to all cities and coun-

ties in Georgia and requests data on revenues,

expenditures, debt and cash investment. The re-

sponses are checked manually and through an au-

tomated edit process for inconsistencies and

errors. Nonetheless, the data have important limi-

tations of which the reader should be aware.

Incompatible Fiscal Years

Unlike many states, cities and counties in

Georgia are free to set the beginning and ending

dates of their fiscal year (the

The reader

12 month accounting period). The most common beginning

and analyst

and ending dates for the fis-

should be careful when comparing
an individual county or city
with the aggregate data.

cal year are July I to June 30 and January 1 to December 31, but any 12 month period may be used. The finance survey requests data for the most recent fiscal year which ended between July 1 and the

following June 30. As a result,

the data in the 1992 Report ofLocal Government

Finances would cover a fiscal year beginning July

1, 1991 and ending June 30, 1992 for some gov-

ernments and a fiscal year beginning January 1,

1991 and ending December 31, 1991 for others,

as well asothervariationsduringthatJuly 1, 1991

and June 30, 1992 period. Local governments are

not required to use audited data in responding to

the survey, although many do so.

Inconsistent Reporting

In addition to setting their own fiscal years,

cities and counties are free to develop their own

chart of accounts. Thus, there is considerable

variation in how expenditures are classified and

recorded. One area of variation that particularly

impacts this report is the classification of capital

expenditures. Because counties and cities do not

use a uniform definition of capital, some capital

outlay (i.e., equipment purchases) may becatego-

rized as operating expenditures by one jurisdic-

tion and as capital purchases by another. In

addition to these accounting variations, cities and

counties are organized and operate differently.

Functions handled by a general public works department in one community might be divided between more specialized solid waste, streets, and recreation departments in another community. While the survey form was developed to avoid the problem of reporting and service delivery differences, variations will remain in the way expenditures are reported by purpose and type.
Special Circumstances
Local governments operate in unique political environments and circumstances, making it difficult to generalize about them. Take, for example, the skewed per capita figures one often gets with resort communities. The population numbers used in per capita equations are permanent residents. Resort communities, however, provide facilities and services to a large number of visitors who would not be included in that permanent population. As a result, it is very common for resort communities to have percapita revenues, expenditures and debt that far exceed the norm. The same is true of central cities providing services for a daytime population considerably larger than their resident population.
Other communities have unique situations which skew their numbers, but which detailed analysis will reveal. For example, one Georgia county is under a court order to use general obligation bonds to pay a defaulted industrial development bond. These factors make it important for the reader and analyst to be careful when comparing an individual county or city with the aggregated data.
Consolidated Governments
Throughout this report, the consolidated government Columbus-Muscogee County is treated as a county government. The consolidated government ofAthens-Clarke County is treated as a county government beginning in 1992. Prior to
1992, the City of Athens and Clarke County are
treated as separate municipal and county entities. In the county government population groupings, Athens-Clarke County is included as part ofpopulation group B, counties with populations between 50,000 and 99,999. Columbus-Muscogee CountY is included in population group A, counties with populations of 100,000 and above. Because ofthe urbaillcity services component of these consoli dated governments, their per capita expenditures

4

r1

and revenues are higher than for unconsolidated counties as a group.
Authorities
The Georgia Department of Community Affairs began collecting financial information from certain independent authorities in 1992; since there is only one year's data and reporting requirements and response rate issues are still being worked out, the data was not included in this analysis. The finances for dependent agencies are included in this report. Dependent agencies are agencies which lack substantial administrative and fiscal independence from general purpose local government. They include special districts and dependent authorities such as airport authorities (controlled by a single general purpose local government), recreation boards, and waterl sewer authorities.

Population Figures
Much of this report's data is shown in per capita amounts. United States Bureau ofthe Census population figures were used to derive per capita amounts. For the years 1986 through 1989, the 1986 population estimates for cities and counties were used. For the years 1990 through 1992, the 1990 actual population counts were used.
Trend Analysis
Several sections of the report focus on trends and changes from FY 1986 to FY 1992. FY 1992 was the most recent year for which data was available at the time of this study. In spite of the existence of earlier data (beginning in FY 1982), FY 1986 was chosen as the base year because of the improved quality of financial reporting and survey editing from that year forward.

The 1992 Report of Local Government Finances is available from the Georgia Department ofCommunity Affairs. Ifyou are interested in obtaining a copy, please contact Linda Fordham, at (404) 656-5527. Surveying for the 1993 Report is underway and the report will be published Spring, 1994.

5

Long-tenn debt is most often, and most ap-

propriately, used for financing capital outlay (e.g.,

equipment and vehicles) and capital projects (e.g.,

land, construction, and renovation) with a useful

life that exceeds the financing tenn. Cities and

counties in Georgia have several financing av-

enues available to them.

General Obligation Bonds

General obligation (G.O.) bonds are the

traditional way in which local governments

finance capital improvements. G.O. bonds are

- - - - - - - - - - debt backed by "the full faith

"Every country has the and credit" of the issuing

government it

government. Put simply, a local

deserves."

government issuing G.O. bonds

JOSEPH OEMAISTRE

has pledged its taxing power-

- - - - - - - - - - - a police power of the state-to

repay the principal and interest. For the investor

that is a powerful pledge, making G.O. bonds a

relatively safe investment vehicle. That pledge,

in combination with their tax exempt status, also

nonnally makes G.O. bonds the least expensive

fonn oflong-tenn debt, in tenns ofinterest rates,

for local governments.

The State Constitution limits the amount of

G.O. debt that may be incurred by political sub-

divisions to 10 percent of the assessed value of

all taxable property. The local government must

also receive approval of a majority of qualified

Table 1: State & Local Government Debt Burdens Relative to the U.S. Average, 1989

Region! State

Full

Public

State

Total Faith & Non- Gov'tas%

Debt Credit Guaranteed of Total

U.S. 100

100

100

Northeast 99

111

70

65.1

Midwest 89

73

76

41.0

South 111

95

Alabama 97

118

Georgia 91

7S

Florida 122

64

North Carolina 83

66

South Carolina 108

90

Tennessee 87

103

102

36.2

65

41.5

134

17.9

182

18.8

110

20.1

160

40.7

77

21.5

West 164

165

134

40.9

Table adapted from: Roy Bahl and William Duncombe. "State and

Local Debt Burdens in the 1980s: A Study in Contrast," Public

Administration Review 53.1 (1992): 35.

6

voters voting in a bond election in order to issue G.O. debt.2
Constitutional amendments may be used to exempt localities from these requirements. For example, the City of Atlanta "may issue a total of $4 million in bonds in any fiscal year for municipal corporate purposes without a referendum" (GA.L. 1986 p. 4824); may "incur bonded indebtedness without the necessity of a referendum for the purpose of constructing and repairing water and solid waste disposal systems" (GA.L. 1986 p. 4818); and has a maximum indebtedness of 12% ofthe total tax digest (GA.L. 1986 p. 4816). The City of Savannah has two Constitutional amendments each authorizing it to "issue general obligation bonds for street and drainage improvements and all costs necessary therefor including acquiring rights-of-way without submitting the issuance to the voters" (GA.L. 1972 p. 1521; GA.L. 1977 p. 1583). Savannah is limited to $2 million in bonds per fiscal year which are exempt in this manner.
The requirement for voter consent has kept many Georgia cities and counties out ofthe G.O. bond market for all but the most popular capital projects. One study ofstate and local government debt burdens in the U.S. showed that in 1989 Georgia had less "full faith and credit" debt and more public non-guaranteed debt (which does not require voter approval) than the U.S. average. Where the U.S. average is set equal to 100, Georgia "full faith and credit" debt was 75 and its public non-guaranteed debt was 134.
One difficulty with the analysis above is that it includes State debt, as well as the debt of special purpose local governments such as school districts. The 1987 Census of Governments does provide a clue to the composition ofmunicipal and county debt.
One interesting trend among U.S. counties and cities is their movement away from "full faith and credit" debt. In 1987 Georgia counties held a larger percentage of their debt as "full faith and credit" debt (39.4%) than did U.S. counties as a whole (27.2%), while Georgia cities held a smaller percentage of debt as "full faith and credit" debt (16.5%) than U.S. cities (31.1%). Since 1987, both cities and counties in Georgia have continued the nationwide trend to proportionately less "full faith and credit" debt.3

Table 2: County &. Municipal Long-Term Debt

Georgia's counties hold more G.O. debt, as a percentage of total debt, than Georgia's cities. About 30% of total county debt in 1992 was in the form ofG.O. bonds, while only about 15%of municipal debt was G.O.; however, cities have more per capita G.O. debt than do counties, $147.60 compared to $89.77 in 1992.4
Revenue Bonds
The most heavily used type of long-term debt for cities and counties is revenue debt. Local governments issuing revenue bonds pledge the revenue derived from the project to repay the bond.5 Because the issuing government is not backing the debt with its taxing power, revenue bonds do not require approval of the voters. State law does restrict the types of activities which may be financed by cities and counties with revenue bonds. Some activities, such as hospitals, are not included on the list, pushing cities and counties to seek enabling legislation to

U.S. Counties/Georgia Counties

Full Faith & Credit

U.S.

Georgia

86-87 27.2 39.4

81-82 76-77 43.1 67.5

Non-Guaranteed

U.S. 70.5 51.1 27.1

Georgia 60.6

U.S. Municipalities/Georgia Municipalities

86-87 81-82

Full Faith & Credit

U.S. 31.1 39.3

Georgia 16.5

76-77 55.3

Non-Guaranteed

U.S. 66.3 55.8 38.3

Georgia 83.5

U.S. Department of Commerce. Bureau of the Census, 1987 Census ofGovernments,4.3 and 4.4 (Washington. D.C.: Government Printing Office): Table I, p.3.

Lease financing is of two types. The traditionallease involves the local government and a single lessor to whom the city or county (the lessee) makes equal payments for the lease period, usually less than 10 years. The lessor might be a

create an authority which may incur revenue debt for that purpose. Revenue debt is commonly used for enterprise activities such as water and sewer systems, electric utilities, and airports. In 1992 almost 75% of all municipal debt and about 56% of all county debt were revenue bonds.
Lease
In 1988, the Georgia General Assembly enacted legislation allowing local governments

bank, an association ofgovernments, or the equipment manufacturer. To the citizenry this type of lease financing would seem very similar to the installment payments which they make on cars, furniture, and appliances. For cities and counties, the traditional lease option provides a way to meet vehicle and equipment needs. These purchases are often of such a small size and the leases are ofsuch a short duration as to make bonds imprac-

to enter into multi-year lease and installment purchases.6 In 1989,
the Georgia Supreme Courtjoined the majority of other states in
upholding this type of financing
and ruling that it does not constitute debt.
2 GA. CONST. art IX, Section V, Paragraph 1.
3 Since the 1992 Census of Governments data is not yet available, it is not clear if the U.S. figures will also reflect a continuing shift toward non-guaranteed debt.
School debt is not included in any of the county debt figures discussed in this report and, thus, would not account for the heavieruse ofG.O. debt by counties.
S GA CaNST. art IX, Section VI, Paragraph I.
6 0.C.G.A.36-60-13.

Table 3: Percent Long-Term Debt by Type

Georgia General Purpose Local Government

All Counties

86

87

88 *89 90

Full Faith & Credit 32.76 39.39 40.68 32.09 31.88

Non-Guaranteed 67.24 Revenue 67.24 Other Lease

60.61 60.61

59.32 598.32

67.91 68.12 64.15 62.34
3.76 5.78

All Cities

86

Full Faith & Credit 17.45

Non-Guaranteed Revenue Other Lease

82.55 82.55

87 16.48
83.52 83.52

88 11.75
88.25 88.25

*89 90 14.56 14.55
85.44 85.46 82.51 78.09
2.93 7.37

91 **92 29.4 30.2
70.6 69.8 63.15 55.6
7.45 6.2 7.9
91 **92 15.47 15.49
84.53 84.5 77.19 74.51
7.34 7.42 2.57

First year that data on "other" debt were collected.
** First year that data on lease financing were collected.
GA Department of Community Affairs, Repon ofLocal Government Finances, 1986-1992

7

tical. Both the Association County Commission-

ers of Georgia and the Georgia Municipal Asso-

ciation operate lease programs for their members.

The second type of lease financing is a more

recent phenomenon and is used for larger pur-

chases and capital projects. This type of lease is

referred to as Certificates of Participation (COPs)

because investors purchase the
A national debt, if it is lease obligations in the form of

not excessive, will be to certificates. COPs are similar to

us a national blessing. " municipal bonds in that they are

ALEXANDER HAMILTON

underwritten and sold to investors.7 The COPs option provides

a way for local governments to meet the need for

large equipment purchases and capital facilities

without worrying about the Constitutional G.O.

debt limit and without seeking voter approval.

These advantages ofCOPs financing are particu-

larly important when financing mandated, but

sometimes unpopular, facilities such as jails or

solid waste facilities.

In spite of court rulings to the contrary, finan-

cial analysts and local governments generally

treat lease financing as debt for planning and fi-

nancial management purposes. The essential na-

ture of the equipment and facilities purchased

with lease financing and the lease agreements

themselves make the annual lease payments a

7 Roy F. Smith, "Certificates of Participation for Lease Financing," Government Finance Review. December 1988: 25.

long-term obligation in reality if not in law. Thus this report will treat leases as long-term debt.
Because of questions about the cost of using COPs rather than G.O. bonds-particularly for public facilities-and because COPs are sometimes seen as a ploy to avoid the voters, controversy has arisen surrounding their use both in Georgia and nationally. While COPs appear to be gaining in popularity among local governments, they constitute a relatively small proportion of local government long-term debt. In 1992, the first year in which lease data was collected as part of the Georgia Report of Local Government Finances, 2.5% ofall municipal debt ($24.54 per capita) and 8% of all county debt ($23.59 per capita) was lease financing. These figures include both traditional lease financing and COPs.
Other Debt
Georgia cities and counties also have loan programs available to them through the State and federal government. Forexample, the Georgia Environmental Facilities Authority (GEFA) makes loans to local governments for capital needs such as water and sewerimprovements. Occasionally, as part of local government finance reporting, a city or county will list "bank loans" in this section as "other debt." These other types of longterm debt constitute a small proportion of total long-term debt-6.2% of 1992 total county debt and 7.4% of total 1992 municipal debt.

8

CLaw t.

r

Trends &Types of Debt
Between 1986 and 1992, long-tenn debt for Georgia's cities and counties grew rapidly. For counties, per capita G.O. and revenue debt grew from $130.11 to $254.98, a 96% increase.8 Per capita G.O. and revenue debt grew at a slower but significant pace for cities, increasing about 31 %, from $654.54 to $857.85.
Not surprisingly, in both cities and counties, population size is related to per capita debt, with larger counties and cities generally carrying the greatest amount of debt. Population size is not, however, related to debt growth. While Georgia's smallest counties have the lowest per capita G.O. and revenue debt, their growth has been quite dramatic. In 1986, the per capita G.O. and revenue debt in the 26 counties with populations between 10,000 and 14,999 was $8.21. By 1992 those counties had G.O. and revenue debt of$38.31 per capita-a 366% increase. The 45 counties with populations under 10,000 started the period with $10.35 of G.O. and revenue debt per capita. By 1992, that figure had increased 325% to $43.99.
Interestingly, only the 34 counties with populations between 15,000 and 24,999 did not significantly increase their per capita G.O. and revenue debt load. For those counties, G.O. and revenue debt increased by only 1.5% during the seven year period (from $30.33 in 1986 to $30.80 in 1992); however, during that period, these counties' per capita G.O. and revenue debt fluctuated from a high of $45.45 in 1988 to a low of$29.90 in 1991.
For cities, the three population groups that showed the most dramatic growth of long-tenn debt were the municipalities with populations between 25,000 and 49,999; 2,500 and 4,999, and less than 500. Between 1986 and 1992, the per capita G.O. and revenue debt in these population groups grew 117%, 82%, and 74%, respectively. Someofthat dramatic growth may have been the result of a large debt load undertaken by one or two cities in a population group. Forexample, the city with the highest per capita revenue debt statewide had a population under 500 and 1992
8 Because the Department ofCommunity Affairs did not collect data on "other long-term debt" until 1989 and data on lease financing until 1992, only G.O. and rev~ue debt can be tracked for the entire seven year pe_ nod, 1986-1992.

Chart 1: Per Capita G.O. & Revenue Debt

19Sf-1992

Counties t------+III- 800

tIl Cities

';%
i ' ! I -mAiFliif-:-

I 61 '.)';

;

.

I

-

-

m,. .'i---j flJ:h'a-:'-

-f,:ilM.~'-Wr{iiilh''

'

"':!'.'

;~'

~'.,,:'.,,;:.1 '

-

-

~I~
.:.*..:..::r.;:'.,1

-

600

~ &

G

I : ! I ~~t-I~ 1

;1,..,".,.,...';*':1:.,.,.....;.,.:...

400

1

-JIm.:~, 1][. 1"- 1"- ~.,:~";t.'".i'~ r i~ ' ..:. .ii; '..:,'.[i,..:,:.,.

.
:.[.:1.

",,;NIT'

200 $$$

1986 1987 1988 1989 1990 1991 1992

Chart 2: Percent Change Per Capita G.O & Revenue Debt, 19Sf-1992

rC_lT_I_E_S_--.-

.,---

r-

Population Less Than 500

E!llF1'Im 500-999

1,000-2,499

2,500-4,999

5,000-9,999

10,000-24,999
'." ., .x.., .,. . v 25,000 - 49,999

.1IlB1IlBm1 50,000 &Above

"', .'".' All Cities

120

90

60

30

%

Chart 3: Percent Change Per Capita G.O & Revenue Debt, 19Sf-1992
COUNTIES

Population Less than 10,000

10,000-14,999

15,000-24,999

25,000-49,999
50,000-99,999
100,000 &above

All counties

400

300

200

100

%

9

revenue debt of $10,000 per capita (compared to $710.19 for all municipalities) and a resort community with a year round population ofunder 500 had the largest amount ofper capita G.O. debt among all the state's cities; $8,235.29 per capita compared to an average of $147.66. The debt which both of these small communities are carrying is for water/sewer facilities.
Total long-term debt for counties and cities is even more significant when lease purchase obligations and other long-term debt are included. In 1992, per capita long- term debt for all counties was $297.04. That numberis distorted by the eight largest counties (with populations of 100,000 and above) which had 1992 per capita long-term debt of $531.36. The other 150 counties in the remaining five population groups had significantly less per capita long-term debt.
The five counties with the highest per capita long-term debt in 1992 were in the Atlanta metropolitan area. Two counties with populations under 10,000 are also on the list of the ten counties with the highest per capita long-term debt in 1992. Fifty-two counties, or almost 1 in 3, reported no outstanding long-term debt in 1992.
Total per capita long-term debt for

Table 4: 1992 Per CapIta Long-Term Debt

COUNTY

Population Group ##

100,000 & above

9

50,000-99,999 21 25,000-49,999 24 15,000-24,999 34

10,000-14,999 26

Less than 10,000 45 AU Counties 159

($$$s) 531.36 167.16 79.68 48.42 47.06 62.21 297.04

CITY

Population Group ##

($$$s)

50,000 & above

4 1,758.13

25,000-49,999

9

705.24

10,000-24,999 29

712.49

5,000-9,999 46

571.67

2,5~,999

71

551.89

1,000-2,499 102

490.72

500-999 94

278.41

Less than 500 166

354.44

All Cities 524* 953.11

* of 536 munIcIpalitIes surveyed

municipalities in 1992 (see Table 4) was about $953.As with counties, the four largest cities carry significantly more total per capita long- term debt than the remaining 520 cities. However, the debt load of the four largest cities, as well as for the total ofall municipalities, is distorted by the over $849 million in revenue debt carried by the City of Atlanta for the airport. In order to appreciate the magnitude of Atlanta's commitment to this facility, consider that without Atlanta's airport debt total percapita municipal debt in 1992 would have been approximately $600, rather than $953. The per capita long-term debt for the four largest Georgia cities would drop to $570 from $1,758. Atlanta airport debt is revenue debt, paid from the proceeds of airport operation; however, the importance of this facility may be more easily grasped if stated in per capita terms: the Atlanta airport debt amounts to $131.19 for every citizen in the state of Georgia.
Interestingly, unlike counties, the cities with the highest per capita total long- term debt were not in the Atlanta metropolitan area. Only two cities in the Atlanta metro area are among the top ten municipal per capita long-term debt holders. This difference may reflect the high growth rate and suburbanization of the unincorporated (county) areas surrounding the City ofAtlanta. It is important to note that of the 524 municipalities reporting, almost 28%, 146 cities, reported no outstanding long-term debt in 1992.
For counties, the growth in debt is a function ofan increased emphasis on capital expenditures overthe past seven years. While per capita county operating expenditures increased almost 50% between 1986 and 1992, county per capita capital expenditures (e.g., expenditures for construction and purchase ofequipment, land, and structures) increased by 127.7%. The higher growth rate of capital expenditures versus operating expenditures held true for all county population groups except for counties with populations between 15,000 and 24,999. For these 34 counties, per capitaoperating expenditure growth outpaced per capita capital expenditure growth, increasing more than 55% from 1986 to 1992 versus about 43% for capital expenditures.
For counties as a whole, the growth rate for per capita capital expenditures associated with general fund activities was particularly dramatic

10

from 1986 to 1992, increasing about 181% compared to a 58% increase in per capita ~pe~ating expenditures. Again, only among counues 10 the 15,000 to 24,999 population group did growth in general fund per capita operating expenditures outpace the growth in per capita capital expendi-
tures. Per capita enterprise fund expenditures vary
among the county population groups, but also reflect the trend toward a greateremphasis on capital expenditures. Among counties with populations between 15,000 and 49,999, that growth has been astonishing. From 1986 to 1992, counties with populations between 25,000 and 49,999 increased per capita enterprise capital expenditures 2700% and counties with populations between 15,000 and 24,999 increased their per capita enterprise capital investment 1352%. This is largely water and sewer investment.
The increase in municipal debt is a reflection of a general increase in per capita expenditures, rather than an increased commitment to capital investment. With only a few exceptions, percapita operating expenditures for cities has increased at a faster rate than per capita capital expenditures (see Charts 4 and 5). Among the four cities with populations over50,000, per capita capital expenditures actually decreased between 1986 and 1992, while total per capita expenditures increased. Only in their enterprise activities did these four cities show growth in per capita capital expenditures; 8% compared to about a43% increase in total per capita enterprise expenditures. This shift away from capital investment may be cause for some concern.
As discussed earlier, non-guaranteed debt is the most prevalent type of long- term debt used by Georgia's general purpose local governments. There is, however, a departure from that pattern for counties, depending on population group. The 30 largest counties (including the two consolidated governments), with populations over 50,000, had a majority of their long-term debt in revenue bonds. For the 71 counties with populations under 15,000, the majority oflong-term debt was in the form ofgeneral obligation bonds (62% for counties with populations between 10,000 and 14,999; about 59% for counties with populations under 10,OOO-see Table 5).
The use of lease .purchase financing varies

County Operating City Operating

m County Capital

~ City Capital

--;il--,800

6 0 0 ~ - - = - - - r m - - _ _ _ _ f H - - - H i t _ _ - _ _ f J - - f H - - _ _ _ _J;- . , ~-lm---H} ..--fA---__1ln--___+q.-__..-tf--__rlit__l 400

f--.--Mt-. .fj1.....--I.-fH---IHmt__......__fiJ-. .fH---I.-fi4----i 200

1986 1987 1988 1989 1990 1991 1992
Chart 4: City & County Per Capita Expenditure by Type .. 1986-1992

Chart 5: Capital Spending as a Percentage of

T

Total Expenditures, 1986-1992

County

25

City

20

15

10

5
%
1986 1987 1988 1989 1990 1991 1992

Table 5: 1992 Long-term Debt by Type

COUNTY

Percentage

Population Group Rev GO Other

100,000 & above 56.9 31.8 4.2

50,~99,999

57.9

21.5

13.1

25,~9,999

44.7

36.0 12.9

15,~24,999

38.0

25.6 27.3

10,~14,999

19.2

62.2

9.8

Less than 10,000 11.8 58.9 19.7

Lease 8.0 7.5 6.4 9.0 8.8 9.6

CITY Population Group
50,000 & above
25,~9,999 10,~24,999
5,~9,999
2,506--4,999 1,000-2,499
500-999 Less than 500
All Cities

Percentage

Rev GO Other Lease

86.0 10.0

1.1 2.9

57.8 37.5

3.5

1.1

53.3 33.2 12.2 1.3

61.6

6.4 25.2 6.8

70.6

1.1 26.7 1.6

76.5

4.0 18.5 1.0

12.8

0.6 26.4 0.2

55.1 22.3 22.5 0.1

74.5 15.5

7.4 2.6

11

considerably among Georgia's counties and cities. A substantial majority of cities and counties have no lease debt at all; about 81 % of all counties and 90% of all cities reported no outstanding lease debt in 1992. As a group, cities and counties held a similar amount of per capita lease financing debt in 1992-$24.54 for cities compared to $23.59 for counties. However, because cities have significantly more total long-term debt than counties, the proportion of long-term debt held in the form of lease financing was less for cities than for counties-about 3% of total long-term debt for cities compared to 8% for counties.
Per capita. the eight largest counties held $42.61 in lease financing debt in 1992, compared to $50.51 per capita in lease debt held by the four largest cities. As a percentage of total long-term debt, leases constituted a larger proportion of long-term debt for the large counties than for the large cities (8% oftotal long-term debt for the five largest counties compared to 2.9% for the largest cities). Gwinnett County's $98 million in Certificates of Participation (COPs) for water and sewer facilities was a major contributor to the rela-

Table 6: 1992 Per capita Long-Term Debt By Purpose

COUNTY Purpose

$ Per Capita

Water & Sewer

131.37

*Public Safety

35.70

Public Buildings

13.81

Highways

14.90

Parks

13.82

Multipurpose

24.97

All Other

58.52

* Includes Police, Fire and Jail

CITY

Purpose $ Per Capita

Water & Sewer

345.07

Other Enterprise

*368.4/**7.96

Education

35.44

***Public Safety

4.23

Public Buildings

14.72

Highways

19.14

Parks

11.49

Multipurpose

69.48

All Other

103.45

* Includes Atlanta airport debt
** Excludes Atlanta airport debt

*** Includes Police, Fire and Jail

12

tively high level of lease debt held by the five largest counties. The debt service on Gwinnett's water/sewer COPs are being paid with system revenues, serving in effect as revenue bonds.
Uses of Debt
The heavy use ofrevenue debt by counties and cities, vis-a-vis other types of debt, is also reflected in the purposes for which local governments issue debt. For both cities and counties, the most frequent use of debt is for water and sewer facilities. In 1992, cities and counties had combined long-term debt ofover $1.6 billion for water and sewer facilities: counties had just over $851 million ($131.37 per capita) in long-term debt for water and sewer facilities and cities had almost $814 million ($345.07 per capita) in water and sewer long-term debt.
Among cities under 10,000 population, water and sewer facility debt accounts for 80% or more of total long-term debt. The heavy investment in water and sewer facilities reflect a number of factors: continuing state and federal mandates for water and wastewater treatment, population growth, and investment for future economic development. In terms of the mandates for clean water, the Environmental Protection Agency (EPA) has estimated that municipal spending on wastewater treatment will increase 85% from 1987 to 2000 and municipal spending on drinking water will increase 50% during the same period.9 Additionally, utilities of all types are very capital intensive, which is why they are generally classified as natural monopolies, while most other city and county services are labor intensive.
There are two major exceptions to this emphasis on water and sewer facilities. The first exception is among the smallest counties, with populations under 15,000, which primarily hold debt for public buildings, multipurpose facilities, and other purposes such as libraries, hospitals, and equipment. The second major exception is the City of Atlanta's airport revenue debt (almost $850 million outstanding at the end offiscal year 1992) which makes spending among all cities for airports seemingly the most prevalent type ofdebt.
Despite the prevalence of county-held water and sewer debt, from 1986-1992 it actually de-
9 "Cities Awash in Clean Water Mandates," Fiscal Notes (Texas Comptroller of Public Accounts): 7.
,~.

creased as a proportion of all county debt, from
about 66% to 45%. This change reflects a growth in county debt for other purposes, rather than an actual decline in water and sewer investment. For example, county debt for parks increased from $0.30 per capita in 1986 to $13.82 in 1992; for highways from $2.32 to $14.90; and for public buildings from $3.73 to $13.81. At the same time, since 1986 total per capita county debt for water and sewer facilities increased from $84.90 to $131.37. Interestingly, the "other" category grew particularly quickly between 1986 and 1992, from $8.08 per capita to $58.52 per capita. This category reflects such diverse county activities as hospital and library capital spending, as well as reporting of lease financing for vehicles and equipment not included in the 1986 survey.
For cities, water and sewer facility debt comprised about the same proportion of total debt from 1986 to 1992, although, like counties, actual percapita debt for water and sewer increased (from $197.34 to $345.07). During the same period, however, municipal debt for public buildings increased from $3.44 per capita to $14.72; for streets and highways from $8.93 to $19.14; and for parks from $3.90 to $11.49. One area reflecting a large increase in municipal debt is edu-

cation, which increased over 128% from 1986 to 1992, from $15.51 per capita to $35.44. Twentyfour cities have school districts. In many cases, the debt for facilities and equipment for these city school districts is issued.by the municipality. As with the counties, municipal debt for "other" purposes has risen dramatically, from $10.61 per capita to $103.45 per capita. The other category for city debt reflects a broad range of projects including hospitals, storm waterprojects, resource recovery, and equipment and vehicle purchase. As with counties, this increase is at least partially the result of changes in the financial survey form.
Conclusion
The seven years between 1986 and 1992 have witnessed considerable growth in city and county long-term debt. For counties, this growth is a reflection of an increased emphasis on capital spending compared to operating expenditures. Cities and counties now use debt for more diverse purposes, although debt for water and sewer facilities predominates for both types of governments. Georgia's cities and counties use more non-guaranteed debt than "full faith and credit" debt, a pattern which became more pronounced between 1986 and 1992, and reflects national trends for both types of governments.

13

The International City/County Management Association (ICMA) publishes a handbook for local government entitled Evaluating Financial Condition, A Handbookfor Local Government. to This handbook was devised to help local governments: Better understand local government financial
condition-the forces affecting it and obstacles associated with measuring it; Identify existing and emerging financial problems; and Develop remedial actions to deal with these problems.I I The handbook's indicators identify trends or conditions, but as the handbook's authors point out, "No single indicator is conclusive."12 Each must be evaluated in light of the jurisdiction's unique political, administrative, and financial history and environment. They are useful, however, to alert governmental officials, financial analysts, and citizens to situations which may need to be addressed. Based on credit industry benchmarks, the ICMA handbook provides a list of warning signals for overall net debt. Net debt is "net direct bonded debt of the local government, as well as the bonded debt ofoverlappingjurisdictions that

is geographically applicable to the local government."13 Overlapping jurisdictions are jurisdictions such as school districts which issue debt against all or part of the same tax base as the county or city. By evaluating net debt, the ability of a community's tax base to repay all of the debt issued against it is measured.
Because ofthe difficulties with obtaining data on overlapping debt for all 159 counties and 538 municipalities, an attempt will not be made to use the ICMA net debt indicator here.14 This analysis will use a more liberal interpretation ofnet debt as the jurisdiction's direct debt minus self-supporting debt (revenue bonds). This means that communities which exceed these benchmarks, not considering overlapping debt, should be doubly attentive to the possibility of a debt management problem. Since overlapping debt will not be considered here, all debt for educational purposes was excluded from the analysis in order not to penalize those municipalities which issue debt on behalf of their city school districts. County school district debt is not reported by the counties in their Report ofLocal Government Finances. Three of the six ICMA indicators for long-term debt are examined here.1S

Long-Term Debt Benchmarks (1) Overall net debt exceeding 10 percent of assessed valuation

This credit industry benchmark matches Georgia's Constitutional requirement that G.O. debt not exceed 10% ofassessed value ofall taxable property. This indicator can be analyzed in two ways. First, are any of Georgia's cities or counties at or approaching the Constitutional cap of 10% ofassessed valuation for G.O. bonds? Any community in this position has limited ability to meet future capital needs.
Of the 159 counties, the highest percent of G.O. debt to assessed valuation was 3.06%. The low ratio of debt to assessed valuation reflects Georgia local government's relatively light use of "full faith and credit" debt. Apparently, the re-
10 Sanford M. Groves and Maureen Godsey Valente, Evaluating Financial Condition, A Handbook for
LocalGovernment (Washington, D.C.: ICMA, 1986).
11 Ibid., p. I. 12 Ibid., p. 5.

quirement for bond elections is a greater obstacle to the counties' use ofG.O. debt than the 10% cap.
No municipality exceeds the 10% of assessed valuation cap, although two are approaching it. Both are using G.O. debt for water/sewer facilities. Four municipalities which did not report collecting a property tax reported having G.O. debt in 1992. Communities do not have to levy a property tax in order to have G.O. debt. In fact, many cities and counties use the special purpose sales tax as the revenue source for their G.O. debt even
13 Ibid., p. 81.
14 Individual counties or cities can do the analysis including overlapping debt more easily and should consider using these benchmarks as part of a coordinated community financial planning process.
IS The three benchmarks not used in this report were excluded because they either did not reflect current Georgia law or because the data was not readily available.

if they levy a property tax. If a community does not levy a property tax, however, there is not always a separate municipal property tax digest. Thus, there is no direct way to evaluate compliance with the state's Constitutional cap.
The second way to evaluate this indicator is to compare net long-term debt to total assessed valuation. This method excludes revenue debt for which proceeds from the project are pledged, but includes other debt which might require payment from general funds. This method flags communities which are carrying long-term debt burdens which, while not "full faith and credit" debt, may not be self-supporting debt. In Georgia, the use of other debt is such a minor part of the total county debt picture that, not surprisingly, the addition of other debt to G.O. debt does not appreciably change the ratio of debt to assessed valuation. On the other hand, seven municipalities exceed the 10% benchmark when comparing net long-term debt to the net bond digest (the

taxable value of property).16 All but one of these cities' debt is primarily for water/sewer facilities.
The one city whose net debt is not for water/ sewer facilities has "other long- term debt" and lease financing for various purposes (Le., fire protection, law enforcement/jail, streets, and garbage collection). The other six cities have "other longterm debt" (e.g. GEFAandFarmer's Home loans) for water and sewer . In 1992, one ofthese six cities' water/sewer systems generated sufficient revenues to pay part of the water/sewer debt service, the others did not. Of the six cities which do not meet this benchmark due to water/sewer debt, three also do not meet the benchmarks for enterprise activity debt discussed later in this report. Another of these cities' water system was still under construction at the time of the 1992 report and not yet generating revenues. The municipality building the new water system exists primarily for that purpose. In 1992, it spent only $2,754 on other activities.

(2) Overall net debt exceeding $1,200 per capita

This benchmark uses a per capita figure, $1,200, to alert officials to the possibility that the net debt burden in that community is excessive. This benchmark should be evaluated cautiously since it does not reflect a given community's ability to pay. Communities which have a lot of wealth, either in the form ofhigh personal income or high property values, may be able to comfortably carry more than $1,200 in per capita net debt. Additionally, communities which can export their tax and debt burden may also have a greater debt capacity. Exporting the tax burden can be accomplished through sales taxes paid largely by tourists or other nonresidents or property taxes paid by nonresident property owners or industry. At the same time, communities need to recognize the danger associated with relying too heavily on tax exportability. Economic forces, such as a downturn in tourism or the construction of a competing mall, could put the community in financial jeopardy.

No county's net per capita debt exceeds $1,200. Six municipalities exceed the $1,200 per capita benchmark, four ofthem because of"other" water/sewer debt (i.e., loans for water/sewer facilities from GEFA). In order to obtain these loans, a community must pledge general fund revenues; however, the loans are made based on an analysis ofsystem revenues and the community's overall financial health. Thus while the water and sewer debt held by the four cities is not revenue bond debt, 1992 system revenues were sufficient to pay at least part of the principal and interest.
Local Government Net Per Capita Debt

1200

900

600

300

16 The municipalities which have no separate property tax digest could not be evaluated against this benchmark.

Cities Counties Benchmark $$$

a;

15

(3) Overall net debt per capita exceeding 15 percent of per capita personal income

This indicator is also known as the S & P Index, named for Standard and Poor's, the bond rating firm that developed it. The S & P Index, unlike the preceding benchmark does reflect a community's ability to pay. Like the $1,200 per capita standard, however, it does not consider the exportability ofthe tax burden. Communities with significant revenues (e.g., sales taxes) from tourists or other non-residents may be able to handle a higher debt load than the indicator recognizes. Additionally, the reality is that for some communities long-term per capita debt of 20% or more of per capita personal income may be politically acceptable, while in other communities debt of5 or 10% of personal income would be considered too high. This benchmark was analyzed using 1990 net per capita debt in order to use 1990

Net Per Capita Debt as a Percent of Per Capita Personal Income

Cities Counties
Benchmark

4

8

12

16

Census figures for per capita personal income. Two jurisdictions exceed the 15% benchmark.
One city makes this list because of "other" long-term debt for water and sewer. Again, as discussed above, this is not revenue debt but communities generally utilize system revenues for some or all of the debt repayment. In 1990, the city's system revenues were sufficient to pay the principal and interest on this debt. Additionally, by 1992 per capita net long-term debt in this city had been reduced to $26.24.
The other city which exceeds this benchmark used general obligation debt to construct a water system. This city's situation is a good example ofhow financial analysis using per capita figures may fail to adequately reflect the situation in a resort community. While the city's year round population is under 200, its peak season population is about 1500. The water system, for which the G.O. bonds were issued, has 600 customers and the city's property tax valuation is almost $35 million. The G.O. debt was issued in 1987 for 20 years. While $1.5 million is a substantial debt load for the community, by all appearances it has the financial capacity to carry the debt.

(4) Debt service exceeding 20% of operating revenues

Another way to evaluate local government debt is to analyze the amount spent on debt service. Debt service is defined here as principal and interest paid on net direct debt (long-term debt excuding revenue bonds and debt for educational purposes) plus interest paid on short-term debt. A credit industry benchmark holds that debt service exceeding 20% of operating revenues is a potential problem.I? Operating revenue is defined here as total tax revenues plus service charges and other revenues (e.g., interest earnings, fines and forfeitures). This benchmark is an indicator of debt management problems because a high debt service burden can limit a jurisdiction's expenditure flexibility-witness the problems faced by the federal government for that reason.
One caution here is that the data from the Report of Local Government Finances does not distinguish among types of short-term debt
17 Ibid., p. 83.

[e.g., Tax Anticipation Notes (TANs), Bond Anticipation Notes (BANs), or Revenue Anticipation Notes (RANs)] nor for the purpose of the short-term debt (e.g., general fund short-term borrowing versus enterprise fund short-term borrowing). Throughout this analysis, selfsupporting debt (revenue bonds) and debt for educational purposes has been excluded. In this instance, however, interest paid on BAN's for revenue or educational bonds or for RANs for enterprise activities may be included in debt service. Corresponding revenue from enterprise activities is not included as part of the operating revenue figures. To the extent that interest on short-term debt for enterprise activities, either BANs or RANs, is included without including enterprise revenues, a community's debt service load will be overestimated. Additionally, the revenue figure used here does not reflect funds held as unappropriated fund balances which may be available to help pay debt service.

Eight counties and 34 municipalities exceed this benchmark. Three of those counties and four municipalities retired over 60% oftheirnet longterm debt (long-term debt excluding revenue bonds and debt for educational purposes) outstanding at the beginning of the fiscal year, making their debt service burden look abnormally high.
As with the other benchmarks, a number of communities have "other" long- term debt which, while not revenue debt, is for water and sewer utility projects. Thirteen cities and one county exceed this benchmark because of the debt service on their "other" debt for water and sewer projects. In all of these cases, the communities reported

Debt service as a Percent of Operating Revenues

20

15

10

5

enough revenue from their utility systems in 1992, after covering operating expenditures, to pay all orpart ofthe debt service on their water and sewer debt and thereby reduce their debt service to revenue ratio below 20%. That leaves 4 counties and 18 cities which do not have readily identifiable circumstances to account for debt service payments over 20% of revenue.

Cities Counties
Benchmar~
%

Revenue Debt
The four benchmarks described above address the management of net debt; however, good management of self-supporting debt is also critical to the financial health of a community. This is particularly true in Georgia where the majority of debt is for enterprise activities, especially water and sewer systems. While communities do not generally pledge their "full faith and credit" to cover debt for enterprise activities, the reality is that they will use their taxing power to make good on enterprise debt rather than have the city's or county's future bond rating adversely affected or have the enterprise activity (e.g., the water and sewer system) go into receivership.
To identify enterprise activities posing potential debt problems, water/sewer system, gas and electric utility, and airport debt were evaluated. Two standards were used to identify potential debt problems.
First, the analysis simply looked at whether enterprise activities were being operated as selfsupporting activities. If an enterprise activity is not self-supporting, the burden forthat service has been shifted from the service's users to taxpayers generally. If property taxes subsidize an activity, then owners of undeveloped land are paying for a system they do not use and to which they may not even have access. While the existence of the enterprise activity may make their property more valuable, there may be an issue of fairness associated with large general fund subsidies.
In order for an enterprise activity to be com-

pletely self-supporting, revenues need to be adequate to pay both operating expenses and debt service. Recognizing that a single year ofrevenue shortfalls does not necessarily mean that an enterprise activity is not self-supporting, the benchmark was established to flag activities in which debt service payments exceeded net system revenues (system revenues minus operating expenses) for three consecutive years-1990, 1991, and 1992!8 The reader should be aware that the requirement that the system not be self-supporting for three consecutive years may miss problems ifsystems are losing customers and revenues or have new loans.
One hundred and eleven water/sewer systems did not meet this benchmark. Five are operated by counties; the other 106 are municipal. The shortfalls ranged from less than 1I10th of a percent-meaning that only a small amount ofthe 1992 debt service could not be paid with 1992 net revenues--to over 1,OOO%-meaning that system revenues paid very little ofthe operating and none of the debt service in 1992.
Many of these III systems are in small communities and the actual deficit is relatively little money. In the case ofthe water/sewer system with the shortfall over 1,000%, the deficit was only about $30,000.
II For this analysis, both revenue debt and "other longtenn debt" are included in the figures for water and sewer debt service. This was done in order to include GEFA and Fanner's Home loans. For the other enterprise activities, only revenue debt was included in the debt service figures.

Recognizing that communities may wish to subsidize their enterprise activities but that a large subsidy limits the community's flexibility in other general fund activities, a second, less strict benchmark was also used. This benchmark only flagged enterprises that were not self-supporting for the three years and had shortfalls equivalent to 10% or more of total 1992 general fund own-------....-.----- source revenues (total tax
Because Georgia's cities revenues plus service and counties are generally charges, fees, fines, and other financially conservative, that revenues). The 10% was only conservatism translates into used to eliminate systems careful debt management. from the list which had rela-
tively small subsidies, but there is nothing magical about the 10% subsidy figure used here. A community may feel subsidies higher or lower than 10% are acceptable. Additionally, the 10% general fund subsidy figure does not mean the community actually used general funds to pay the shortfall between net water/ sewer revenues and debt service. Capital reserve funds, enterprise fund retained earnings, or transfers from other enterprise activities (e.g., electric utilities) may have been available for this purpose.
One county and 55 municipal water/sewer systems exceeded this benchmark. Sixteen systems had shortfall percentages less than 100%, indicating that only a proportion of the debt could not be covered by net revenues (revenues minus operating expenditures). Three systems had shortfall percentages equal to 100%, indicating that system revenues were equal to operating expenditures with no net revenues available to cover debt service. The remaining 37 systems had shortfall percentages over 100%, indicating that system revenues were insufficient to cover system operating expenditures as well as debt service. In these instances, the water/sewer system may not necessarily be carrying a heavy debt load, but the utility system as a whole is not operating as a selfsupporting enterprise fund. For example, at the end offiscal year 1992 one municipal system only had $9,000 in water/sewer revenue debt outstanding, having retired $8,000 of the $17,000 which was outstanding at the start ofthe fiscal year. What caused that system to exceed the benchmark was that system revenues only covered about 67% of operating expenditures.
The-majority of jurisdictions exceeding this

benchmark are small; over 64% have populations under 1,000. Some of these systems may not be metered, charging a flat rate for water/sewer. Some may not have a large enough customer base to operate as a true, self-supporting enterprise activity. Additionally, for some ofthe smallest communities, the utility is the primary function ofcity government. The shortfalls experienced by these small systems involve relatively small amounts of money, averaging $30,628 and representing about 28% ofown-source general fund revenues.
The number of jurisdictions exceeding this benchmark is cause for concem. While this benchmark does not indicate a widespread crisis in local government utility systems, it does suggest that a number of communities may require technical or other assistance to improve the financial status of their water/sewer utility systems.
Conclusion
Georgia's counties and municipalities appear to have relatively few debt management problems. Thirteen counties exceeded one of the five benchmarks, none exceeded more than one.19 A number of municipalities, particularly the small jurisdictions, exceeded one of the five benchmarks; fourteen municipalities exceeded two benchmarks and one exceeded three. No city or county exceeds the limits on all five. While exceeding one or more of these benchmarks does not necessarily mean that a community is in debt trouble, it does mean that public officials may wish to evaluate the debt load and their debt management practices and consider how the present situation might affect future development plans.
The benchmarks described indicate cities and counties have a lot ofunused debt capacity. They are generally a long way from using their Constitutionally authorized 10% of assessed valuation for G.O. bonds. The availability of unused debt capacity does not suggest that our counties and cities ought to take on higher debt loads; rather, it does indicate an ability to meet local capital needs in the future. Because Georgia's cities and counties are generally financially conservative, that conservatism translates into careful debt management.
19 Only the least restrictive of the two enterprise activity benchmarks was counted, since the 56 jurisdictions which exceeded that benchmark automatically exceeded the more restrictive benchmark.

18

City and county long-tenn debt in Georgia has grown rapidly during the past seven years, but remains relatively low compared to ICMNcredit industry benchmarks. Since 1986, county G.O. and revenue debt has increased 96% while municipal G.O. and revenue debt has increased 31 %. Forcounties, the increase in long-tenn debt is part ofa greater emphasis on capital spending. County per capita capital expenditures increased about 128% between 1986 and 1992, compared to a 48% increase in per capita operating expenditures.
The growth in municipal long-term debt is part of a general increase in per capita expenditures, rather than a change in spending patterns. While county per capita spending has grown more quickly than city capital spending, cities still invest more per capita in capital projects than do counties - $133.20 percapita for cities compared to $97.64 for counties. Nonetheless, the gap in capital spending is narrowing. In 1986, cities spent $71.95 more per capita on capital than counties. In 1992, cities only spent $35.56 more per capita on capital than counties. This at least partially reflects the impact ofsuburbanization on the counties, as many counties are called on to deliver urban services to their citizens.
The majority of city and county debt is nonguaranteed debt. Because 1992 was the first year that lease/COPs data was collected on the Report of Local Government Finances, it is unclear to what extent this financing option will add to or accelerate the movement away from "full faith and credit" debt. The use of lease financing bears watching in the future. But even without a long track record of lease financing, there has been a shift away from G.O. bonds, a trend that is also evident nationwide. While slight, that shift points to an interesting policy question. With the availability of non-guaranteed debt options, particularly lease financing/COPs, the Constitutional requirement for voter approval of G.O. debt is essentially irrelevant.
Another important aspect oflocal government finance in Georgia is the use of the special purpose sales tax to pay G.O. debt. The special purpose sales tax requires voter approval, but makes the Constitutional cap of 10% of the assessed value ofall taxable property an inappropriate limitation for those G.O. bond issues. A good casein point are the four municipalities which reported

G.O. debt but which did not report any property

tax collections.

The Constitutional cap of 10% of assessed

valuation may have been a reasonable safeguard

against excessive debt when local governments

primarily relied on property taxes for their revenues and before localities had widespread access to the lease financing option. It no longer reflects current local

Long-Term
Per Capita Debt $$$ 400+
200>399.99 100>199.99 50>99.99 25>49.99
10>24.99 .50>9.99
0

Number of
Counties 10 15 17 15 17 17 16 52

government financial practices. If state policy

makers wish to control local government debt, a

more relevant safeguard would be to somehow

link non self-supporting debt to all own- source

revenues. At the same time, the citizens of

Georgia might be better off if the cheaper G.O.

bonds were a more attractive financing vehicle for

local governments.

The primary obstacle to the use ofG.O. financ-

ing appears to be the need for voter approval. The

hesitancy to hold bond elections most likely grows

from a number of factors, including:

the risk of a "no" vote on a vital public facil-

ity such as a court mandated jail facility;

the expense and time required to hold a bond

election and mount an "education&l"

campaign.

The time factor involved in the G.O. bond

option can be critical for local governments when

making a debt financing decision. For example,

a metro-Atlanta county's decision to use COPs

to acquire and renovate their county admin-

istration complex was at least partially prompted

by the need to act quickly on the opportunity to

acquire that facility.

If G.O. financing is the preferred financing

route, some possibilities to make it more attractive

might include a Constitutional change to lift the

requirement for voter approval on all or part of a

city's or county's maximum general obligation

indebtedness. The Constitutional amendments

which give Atlanta and Savannah partial

exemptions from the requirement provide a

precedent for such a change.

South Carolina may provide some insights

19

% of All Counties
6.29 9.43 10.69 9.43 10.69 10.69 10.06 32.70

into the impact of a change in the Constitutional requirement for G.O. debt. South Carolina cities and counties may incur G.O. debt up to 8% of property tax valuation without voter approval. Voter approval is required for any G.O. debt over the 8% cap, with no maximum cap. In 1992, South Carolina counties reported having $ I02.79 per capita in outstanding G.O. debt, compared to Georgia counties' $90.77 per capita.20
At the same time, South Carolina counties reported having only $0.33 in per capita capital lease purchase debt in 1992 and a total of $12.54 per capita in non-enterprise fund bank loans, lease, and "other" debt compared to Georgia counties' $24.03 in reported 1992 per capita lease debt. While differences in the two states' financial reports make absolute comparisons difficult, it does indicate that the absence of an election requirement does not necessarily result in high levels of G.O. bonded indebtedness.
20 S.C. State Budget and Control Board, Division of Research and Statistical Services, "Annual County Financial Report," 1992.

The financial health of the state's water and sewer systems does need to be closely monitored. This is particularly true for the smallest water and sewer systems, the financial health of which is essential to state economic development efforts. Technical assistance in the form ofrate studies and capital planning assistance, as well as training for improved utility system financial management may adequately address the problem. Another option would be to closely tie system permitting to a demonstration of financial health and economic viability.
While county and city debt has grown significantly since 1986, there is no evidence to suggest that debt is being poorly handed. As a group, Georgia's cities and counties have relatively little debt. Almost 30% of the counties and cities have no long-term debt at all. Few communities exceed any of the ICMA/credit industry benchmarks. While this does not mean that all of Georgia's general purpose local governments are financially healthy, it does not indicate any pressing need to limit city/county long-term debt.

20