An Economic Analysis of Georgia’s Black Belt Counties
Brigid Doherty1 and John McKissick2
1
Economist and 2 Coordinator, Professor, and Extension Specialist
CR-02-06
Center for Agribusiness and Economic Development
College of Agricultural and Environmental Sciences
The University of Georgia
April 11, 2002
Introduction
As with many states, Georgia has historically had areas of high poverty among its
counties. In Georgia, and in the Southeast, counties with high poverty levels have been
labeled “black belt” counties. This paper dissects the economies of black belt counties
and compares them with non-black belt counties. The main thesis of this paper is to
determine the ability of black belt counties to be economically competitive.
Data
For purposes of this paper, the black belt counties are as defined by The Black
Belt Data Book (Wimberley, Morris and Woolley). The Black Belt Data Book lists
counties by percentage of total population in poverty. Georgia counties with a percent of
total population in poverty of over twenty percent were included in this analysis. The
counties included are shown in red in the following map.
Map 1: Georgia Black Belt Counties
Catoosa
Dade Walke r
Fan nin
Wh itfiel d Murray
Gordon
Chattoo ga
Floyd
Union
Cherokee
Forsyth
Haralso n
Rabun
Jack so n
Barrow
Gwi nnett
Cobb
Pauldi ng
Town s
Wh ite
Habersham
Lum pk in
Steph ens
Dawso n
Banks Frankli n
Hall
Pickens
Barto w
Polk
Gilm er
Classification
Belt Counties
Non-Belt Counties
Hart
Elbe rt
Madi son
Clarke
Ogleth orpe
Ocone e
Wilkes
Walto n
De Kalb
Rockdale
Morgan
Ne wton
Clayton
Greene Taliaferro
Fulton
Do uglas
Carroll
Coweta
Heard
Meriwether
Tro up
Pike
Jaspe r
Butts
Spalding
Lamar
Monroe
Putnam
Hancock
Baldw in
Jone s
Colum bia
Warren
McDuffie
Hen ry
Faye tte
Lin col n
Richmo nd
Glascock
Burk e
Washin gton Jeffe rson
Upso n
Bibb
Harris
Talbot
Crawfo rd
Taylor
Muscogee
Do oly
Terrell
Lee
Clay
Cal houn
Do ugherty
Irw in
Wo rth
Jeff Davis
Coffee
Tift
Mille r
Mitchel l
Colquitt
Berrien
Bulloch
Evans
Effi ngham
Bryan
Chatham
Tattnall
Early
Baker
Can dler
Telfair
Ben Hill
Turn er
Tre utlen
Montgo mery
Wh eeler
Toom bs
Do dge
Wilcox
Crisp
Randolph
Laurens
Bleckley
Pulaski
Sum ter
Quitman
Scre ven
Em an uel
Hou ston
We bster
Jenk ins
John son
Peach
Maco n
Mario n
Chattahoo ch ee
Schle y
Stewart
Wilkinson
Twiggs
Atk inson
Lon g
Wayne
Bacon
McInto sh
Pierce
Ware
Cook
Liberty
Appling
Brantley
Glyn n
Lanie r
Se minol e De catur
Clinch
Grady
Thom as
Bro oks
Charlton
Cam den
Low ndes
Echols
1
Methods
Many methods exist for examining the structure and competitiveness of regions
and counties. Two such methods will be implemented in this paper. The first method
breaks the local economy into sectors. This can be done by sales and by employment.
This method also addresses the issue of productivity, manufacturing mix, and agriculture.
The second method examines receipts to households by their sources.
Input-output models allow for both analyses to occur. The input-output model
software called IMPLAN (IMpact Analysis for PLANning) was used for this report.
Input-output models are designed to illustrate the flow of money throughout the
economy. These models show the market value of goods and services exchanged by
industries to produce a final good. IMPLAN expands on input-output modeling by
including the Social Accounting Matrix (SAM). The SAM builds dimension into the
model on two levels. First, the SAM includes the market value between both industrial
and non-industrial groups (households, government, investment and trade). Second, the
SAM includes non-market valued exchanges, such as those between government and
households.
Several IMPLAN models were developed using 1999 data, the most current
available. Each black belt county was modeled to allow analysis of individual
economies. The black belt counties were also modeled as a group to allow for ease of
comparisons. Finally, the non-belt counties were modeled as a group.
Sector Analysis
As mentioned above, IMPLAN has a basic input-output structure as its main
component. This input-output structure traces flows of dollars among industries. The
industrial classification used in IMPLAN follows the Standard Industrial Classification
(SIC) scheme. Thus, industries can be grouped together into sectors. Each sector
contains a multitude of closely related industries.
IMPLAN has ten major sectors; agriculture, mining, construction, manufacturing,
TCPU (transportation, communications and utilities), trade, FIRE (finance, insurance and
real estate), services, government and other. The agricultural sector contains all
production agriculture, forestry and services related to agriculture and forestry. It does
not include manufacturing related to agriculture, such as milk processing. Many
manufacturing activities rely on agricultural production in order to exist in the state. The
mining sector includes all activities related to mining. Construction includes the building
of new structures, but does not include all maintenance and repair activities. Some
maintenance and repair is included in services. Manufacturing contains all activities
related to manufacturing and has the most industries in the IMPLAN model. TCPU
stands for transportation, communication and utilities. This includes items like telephone
companies, utility companies and railroads. The trade sector is comprised of wholesale
and retail trade, which includes restaurants and grocery stores. FIRE represents finance,
insurance and real estate. Services are all items that are typically considered service
oriented, such as hotels, car repairs and health services. The government sector includes
2
both federal and state/local government. Finally, the other category exists to help balance
the model with inventory adjustments and other items.
IMPLAN can categorize both output and employment data by these sectors for
each economy. Thus, this paper will examine how output and employment differ in the
two groups of counties. This paper will then add employment compensation data to
examine productivity and wages.
Output
The first measure considered is output, which is the value of all goods and
services produced in the area of analysis. Total output for the black belt counties is $52.7
million dollars. This translates into a per capita value of approximately $41,000. Per
square mile, this is equivalent to $1.8 million. Total output for the non-belt counties is
$417.2 million. This translates into a per capita value of $62,000. Output per square
mile equals $14.2 million.
Chart 1: Output per Capita and per Square Mile, Georgia
1999
$70,000
$60,000
$50,000
$40,000
$30,000
$20,000
$10,000
$0
Belt
Non Belt
Output Per Capita
Output per Square Mile
(thousands)
Chart 2 shows the percentage value of output from each sector in the black belt
economies. Manufacturing provides the largest percentage of output in the counties,
followed by services and trade.
Chart 2: Black Belt Output
Government
9.6%
Services
12.7%
FIRE
8.4%
Trade
10.8%
TCPU
6.9%
Other
Agriculture
0.1%
6.0%
Mining
1.7%
Construction
7.1%
Manufacturing
36.8%
3
Chart 3 illustrates the same information as chart 2, but for the non-black belt
counties. Manufacturing is the largest contributor to the economy, followed by services
and trade.
Chart 3: Non Belt Output
Agriculture Mining
1.0%
0.1%Construction
7.8%
Government
7.4%
Services
Manufacturing
18.1%
24.4%
Other
0.1%
FIRE
13.8%
Trade
15.2%
TCPU
12.1%
Several interesting conclusions can be drawn from a comparison of the data and
the two pie charts. First, both groups of counties have manufacturing, services and trade
as their three biggest economic sectors in terms of output. However, the shares of the
economy held by these sectors are different. In black belt counties, manufacturing
accounts for a larger share of total output than in the non-belt counties. Second,
manufacturing and agriculture show dramatic differences in share of total output. Next,
on both a per capita and per square mile basis, the black belt counties are significantly
lower in terms of output. Fourth, the economy of the non-belt counties is more
diversified with a service orientation.
Employment
The second measure used is employment. Total employment in the black belt
counties is 664,066. This means that roughly fifty-two percent of the total population is
employed in the region. There are 23 employees per square mile.
Chart 4 shows employment by sector in the black belt counties. It is immediately
clear that although manufacturing contributes 37% of total output, it only employs 17.8%
of the workers in the counties. Trade and services are tied as the largest sectors of
employment in the belt counties.
4
Chart 4: Black Belt Employment
Agriculture
7.0%
Government
18.8%
Other
1.4%
Mining
0.7%
Construction
5.6%
Manufacturing
17.8%
Services
20.4%
FIRE
4.2%
TCPU
3.7%
Trade
20.3%
Employment in the non-belt counties is 4.2 million people. Thus, sixty-two
percent of the total population is employed in the non-belt counties. There are 142
employed persons per square mile.
Employment by sector in the non-belt counties is the focus of chart 5. The trend
of manufacturing having a lower percentage of employment than output is true in these
counties too. Service becomes the largest sector of employment, followed by trade.
Chart 5: Non Belt Employment
Agriculture
1.5%
Government
14.3%
Other
0.7%
Mining
0.1%
Construction
Manufacturing
7.2%
11.8%
TCPU
5.9%
Services
29.2%
FIRE
6.9%
Trade
22.4%
5
A comparison of the two employment pie charts reveals several items. Both
groups of counties have lower percentages of workers in manufacturing than they do
output in that sector. This suggests that manufacturing is more productive per worker
than other sectors. Services and trade become major employers in both groups of
counties. Government also becomes a larger portion of the pie in the employment
graphs.
These observations raise many issues surrounding output and employment in both
sets of counties. A major question is the role of manufacturing in the economy. Clearly,
employees in manufacturing sectors are more productive than other sectors if they can
produce more output with fewer employees. This observation raises speculation
regarding output per worker across sectors and across groups of counties. Another major
issue is the composition of manufacturing. A comparison of the types of industries in
each county group may reveal differences in the competitive nature of the groups.
Finally, the role of agriculture is in question. The next sections of this paper will attempt
to answer these areas of speculation.
Output per Worker
Chart 6 graphs output per worker in each of the sectors in both the belt and nonbelt counties. Total output per worker in the black belt counties is $79,000, while output
per worker in the non-belt counties is $100,000. Output per worker is essentially equal in
agriculture. The belt counties are more productive than the non-belt counties in mining.
However, in every other sector, the non-belt counties have higher output per worker.
Chart 6: Output per Worker by Sector, Georgia 1999
250,000
150,000
Belt
Non Belt
100,000
50,000
ric
ul
tu
re
C Min
on
in
M stru g
an
c
uf t i o n
ac
tu
rin
g
TC
PU
Tr
ad
e
FI
R
Se E
G rvi
ov ce
s
er
nm
en
t
O
th
er
-
Ag
Dollars
200,000
6
Output per worker can be considered as one measure of productivity.
Productivity in turn, can be viewed as a measure of competitiveness. Higher productivity
usually indicates a more efficient use of the factors of production; land, labor and capital.
Manufacturing Mix
Earlier analysis in this paper shows that manufacturing is a larger component of
the black belt’s total output, but that output per worker is lower in the belt counties than
the non-belt counties. Thus, it is logical to wonder about the composition of the
manufacturing industry in both the belt and non-belt counties.
Chart 7 shows the mix of industries in the manufacturing sector for the black belt
counties. In this chart, the industries are grouped at the 2-digit SIC level. This allows for
more detailed analysis of the types of manufacturing companies in the black belt
counties.
Chart 7: Manufacturing Mix, Black Belt Counties,
Georgia 1999
Scientific
Transportation instruments
Food Processing
equipment
1%
18%
Miscellaneous
7%
0%
Tobacco
Electrical equipment
Industrial machinery
Manufacturing
4%
9%
0%
Fabricated metals
Textiles
6%
9%
Primary metals
Apparel
2%
7%
Stone, glass, clay
2%
Furniture
Leather Products
Chemicals
1%
0%
4%
Wood Products
Printing/Publishing
15%
Rubber Products
2%
Pulp and paper
4% Petroleum Products
9%
0%
According to the chart, food processing and wood products are the two biggest
manufacturing industries in terms of output for the black belt counties.
7
Chart 8 illustrates the mix of manufacturing industries by output in the non-black
belt counties. In these counties, food processing and transportation equipment produce
the largest shares of total output.
Chart 8: Manufacturing Mix, Non-Black Belt
Counties, Georgia 1999
Transportation
equipment
14%
Scientific
instruments
Food Processing
1%
Miscellaneous 15%
1%
Electrical equipment
6%
Industrial machinery
6%
Fabricated metals
3%
Primary metals
3%
Stone, glass, clay
3%
Leather Products
0%
Rubber Products
4%
Tobacco
Manufacturing
5%
Textiles
14%
Apparel
2%
Chemicals
Furniture
7%
1%
Printing/Publishing
5%
Pulp and paper
Petroleum Products
7%
0%
Wood Products
4%
Agriculture
The percent contributed by agriculture to total output changed fairly dramatically
from the black belt counties to the non-belt counties. Since many of the counties
classified in the black belt are rural, South Georgia counties, it is not surprising that
agriculture plays a larger role in the economy of these counties. However, it is not clear
what role agriculture plays in contributing to competitiveness.
Output per worker was used earlier as a measure of productivity because it alludes
to the effectiveness of the use of land, labor and capital. Of these factors of production,
only land is not mobile. In other words, labor and capital can be adjusted by policy
decisions and other activities. Land cannot. Thus, it would be remiss not to explore the
value of agricultural production from the black belt counties versus non-belt counties.
8
Map 2: Farm Gate Value Per Acre, Georgia 2000
Farm Gate Value Per Acre
0 - 500
500 - 900
900 - 1700
1700 - 3000
3000 - 7000
Map 2 shows farm gate value per acre in each county in Georgia. Counties
classified as black belt tend to be the same counties with low farm gate value per acre.
Conversely, counties with high farm gate value per acre are mainly those not in the belt.
Counties with high values per acre tend to those with high value per acre commodities,
such as poultry and vegetable production. Charts 9 and 10 display the break down of
farm gate income by major category. The belt counties have higher concentrations of
crops and forestry, which are typically low value per acre enterprises.
Chart 9: Farm Gate Value by Category, Belt
Counties 2000
Vegetables
9%
Other
11%
Animals
28%
Forestry
11%
Crops
41%
9
Chart 10: Farm Gate Value by Category, Non-Belt
Counties 2000
Vegetables
7%
Forestry
8%
Crops
21%
Other
4%
Animals
60%
Compensation per Worker
This paper has explored sources of output and employment. It has considered
productivity issues and looked at the mix of manufacturing companies and the role of
agriculture. However, the sector analysis is not complete without a discussion of
compensation per worker.
Compensation per worker by sector is shown in chart 11. Total compensation per
worker in the belt counties is $24,000. Compensation per worker in non-belt counties is
$33,000. Non-belt counties receive higher compensation per worker in every sector
except mining. One conclusion is that employees are being compensated for higher
output per worker with higher pay.
It is important to note that employee compensation here is defined as wages and
benefits. Self-employment income is not included in this definition. Thus, lower
averages of compensation may be noted in areas such as agriculture. There are more
part-time workers, thus lower values per person, in agriculture. This combined with the
fact that many agricultural producers claim their income as self-employment income,
explains the relatively low value of agricultural compensation per worker.
10
Chart 11: Compensation per Worker
Belt
Non Belt
Ag
ric
ul
tu
re
M
C
on inin
g
st
ru
M
c
an
uf tion
ac
tu
rin
g
TC
PU
Tr
ad
e
FI
R
Se E
G rvic
ov
es
er
nm
en
t
O
th
er
60,000
50,000
40,000
30,000
20,000
10,000
-
Household Receipts
Having completed sector analysis using basic input-output theory, one can now
advance another step by using the Social Accounting Matrix (SAM). To review, the
SAM adds dimension to an input-output model by adding flows between industrial and
non-industrial sources and by adding non-market values for good and services
exchanged.
The SAM in IMPLAN essentially has thirteen components to trace dollar flows.
These are; industry, commodity, employee compensation, proprietary income, other
property income, indirect business taxes, households, federal government, state/local
government, enterprises (corporations), capital, inventory, and trade.
Since the SAM is designed to illustrate the exchange of monetary value
throughout the economy, the matrix itself has both rows and columns. The rows show
receipts to a sector (say households) from the other sectors. The columns show the
expenditures by one sector (say employee compensation) to another (households).
While the SAM itself is interesting, for purposes of this paper, the focus will be
on the household row. This will reveal the amount of dollars transferred to households
from the other sectors. Some interpretation of the sectors will be needed. Wages and
benefits (employee compensation) represent dollars provided to households as payment
for work. However, compensation for self-employment is part of the proprietary income
category. Property income includes monies received at households from the ownership
of property, like rents and dividends. Households do not receive indirect business taxes
as this sector captures taxes. Receipts to households from households include household
11
payments of interest to other households, such as personal notes, contracts for deed and
other inter-household loans. The federal government essentially makes two types of
payments to households; interest payments and transfer payments. Interest payments
cover dollars returned to holders of government bonds and other securities. Social
security, veterans benefits, food stamps, direct relief, earned income credit, and aid to
students are among the transfer payments. State and local governments also make
interest and transfer payments to households. Interest payments are mainly for bond
holders. Transfer payments are mostly comprised of welfare payments and
unemployment compensation. Receipts to households from enterprises are entirely from
dividend income, or the distribution of corporate profits. Capital represents dis-savings
or withdrawals of capital by households to support their consumption. Households have
no inventory adjustments. Trade receipts to households include the value of goods and
services exchanged across state and federal boundaries.
Total
Total receipts to black belt households were $32.9 million in 1999. The break
down of these receipts by source is shown in chart 12. Wages and benefits provide the
largest source of income to households. Enterprises (corporate dividends) provide the
second biggest portion of receipts, followed by the federal government.
Chart 12: Household Receipts by Source, Black Belt
Counties, Georgia 1999
Foreign Trade
0.22%
Domestic Trade
0.02%
Capital
10.95%
Commodity
0.02%
Wages /Benefits
39.91%
Enterprises
18.03%
State/Local Govt
1.92%
Fed Govt
13.53%
Self Emp Comp
7.25%
Rents, Dividends
Households 6.36%
1.78%
Non-belt households received $205.5 million in receipts. Chart 13 graphs the
amount of receipts by source for non-belt counties. Wages and benefits comprise over
half of receipts by households. Federal government and capital make up the next two
highest categories.
12
Chart 13: Household Receipts by Source, Non Black Belt
Counties, Georgia 1999
Enterprises
5.91%
State/Local Govt
1.06%
Domestic Trade
0.03%
Capital
8.70%
Foreign Trade
0.19%
Commodity
0.01%
Fed Govt
9.19%
Households
2.00%
Wages /Benefits
58.85%
Rents, Dividends Self Emp Comp
7.03%
7.04%
In comparing the two pie charts on household receipts, two things immediately
stand out. First, households in non-belt counties receive a large amount of their income
from wages and benefits. Second, households in belt counties receive more of their
income from enterprises (corporate profits) than non-belt counties. A comparison of the
two reveals more about the use of the factors of production as well. Capital in these
graphs represents the amount of capital that must be exchanged for dollars in order for
households to maintain consumption. Households in belt counties are receiving a higher
percentage of their consumption from capital than non-belt counties.
Per Capita
While the composition of total household receipts is revealing, it is difficult to do
direct comparisons. Total per capita in the black belt counties is $26,000 while total per
capita household receipts in non-belt counties is $30,000. Chart 14 shows receipts to
households by source on a per capita basis. The same two items that were noticeable in
the pie charts also are of note in this chart.
13
Chart 14: Household Receipts by Source, Per Capita, Georgia 1999
$20,000
Belt Counties
Non Belt Counties
$18,000
$16,000
$14,000
$12,000
$10,000
$8,000
$6,000
$4,000
$2,000
n
Tr
ad
om
e
es
t ic
Tr
ad
e
ita
l
re
ig
C
ap
D
Fo
s
ag
e
W
C
om
m
od
ity
/B
en
Se
ef
its
lf
Em
p
R
C
en
om
ts
p
,D
iv
id
en
ds
H
ou
se
ho
ld
s
Fe
d
St
G
at
ov
e/
t
Lo
ca
lG
ov
t
En
te
rp
ris
es
$-
Per Square Mile
Comparisons of household receipts per square mile can also illustrate
competitiveness. Total household receipts per square mile in the black belt counties are
$1,152,880. Total receipts per square mile in the non-belt counties are $7,207,862.
Chart 15 shows these receipts by their source. On a per square mile basis, the non-belt
counties consistently have higher value per mile.
Chart 15: Household Receipts by Source, Per Square Mile, Georgia 1999
$4,500,000
Belt Counties
Non Belt Counties
$4,000,000
$3,500,000
$3,000,000
$2,500,000
$2,000,000
$1,500,000
$1,000,000
$500,000
tic
Tr
ad
e
Tr
ad
e
om
es
D
C
ap
ita
l
Fo
re
ig
n
iv
id
en
ds
H
ou
se
ho
ld
s
Fe
d
G
St
ov
at
t
e/
Lo
ca
lG
ov
t
En
te
rp
ris
es
D
om
p
C
m
p
R
en
ts
,
/B
en
ef
its
Se
lf
E
W
ag
es
C
om
m
od
ity
$-
14
Conclusions
This paper has explored the competitiveness of Georgia’s black belt counties
compared to the non-belt counties. Two main approaches were taken to accomplish this
project. First, a sector analysis was performed. The analysis identified two main areas in
which the groups of counties differ: manufacturing and agriculture. These two areas
were further explored. Productivity of the sectors was also considered. It was shown that
black belt counties are less efficient in the use of some factors of production.
Second, an analysis of household receipts was done. This analysis revealed that
households in the black belt counties receive less of their income from wages and
benefits and more from enterprises than do non-belt counties.
Our analysis leads to several conclusions about characterizing the black belt.
First, the belt counties have a lower output of goods and services. Second, the counties
are more dependent on low wage manufacturing than non-belt counties. Third, there is a
low value per acre of agriculture. Finally, the belt counties are more dependent on
government and dividends for household income, while their household income is about
$8,000 per person lower than non-belt counties.
15