Low Density Counties with Different Types of Sociodemographic, Economic and Health/Mental Health Characteristics

Table of Contents
Introduction  | Analysis | Pure Types with Concentrations of Counties that are Low Density | Principal Characteristics of the Low Density Pure Types | Characteristics of the Low Density Pure Types | Location of the Low-Density Pure Types | 1980-1990 Sociodemographic Changes | References | Appendix A | Appendix B

Introduction

This letter presents the results of a statistical procedure, Grade of Membership (GOM), used to empirically develop a typology of counties that may be useful in describing key characteristics of isolated rural or "frontier" areas of the U.S. Using 1980 data, it clusters counties in the coterminous United States into types that have similar social, demographic, economic, and health characteristics (Goldsmith, Holzer, Woodbury, & Ciarlo, 1998). Emphasis was placed upon the identification of county types that h ave concentrations of "sparsely populated" (less than 7 persons per square mile) counties or "less densely settled" (between 7 and 15 persons per square mile) counties, collectively called "low density". Sparsely populated counties are also often referred to as "frontier" and less densely settled as "frontier-like". The stability of county clusters between 1980 and 1990 is also evaluated.

This effort was undertaken in order to help policy-makers make better-informed judgments about the health and mental health problems of rural populations, particularly those that live in low-density counties. Because of their remoteness, small populati ons, limited and fragile economic bases, as well as culturally diverse and often poor populations, these areas are likely to have difficulties supporting adequate health and mental health services. Such counties have special mental health service need, de mand and access problems. Moreover, all low density counties do not have the same levels of need for or access to services (Ciarlo, Wackwitz, Wagenfeld, Mohatt, & Zelarney, 1996).

Generally, the need for and access to services in such counties can be indexed by their social, demographic, economic and health characteristics. Based on this expectation, this work attempts to describe the social, demographic, economic, and health ch aracteristics (structural characteristics) that help to identify different types of frontier and frontier-like counties. Stated differently, the GOM analysis identified groups of counties with clearly different structural characteristics, without regard f or population density. Then, by focusing on those groups with high concentrations of low-density counties, it is possible to find those counties that are most "frontier" or "frontier-like" and that share similar structural patterns, including health and m ental health service needs.

Analysis

The Grade of Membership Procedure. Grade of Membership (GOM) is a statistical classification procedure with the capacity to simultaneously identify and clearly and precisely discriminate clusters of cases or units (such as counties) . The GOM analysis determines the probability with which a select set of variables are associated with a pure or latent type (PT), as well as the extent to which any given unit (such as a county) has the characteristics of one of the pure types (see Golds mith, et al., 1998). In this study, the analysis produced 27 pure types.

Variable Selection. Seventy-five variables were used in the GOM analysis of the 3064 coterminous counties of the United States (see Appendix A for a detailed list). Key variables included county level social rank (economic, occupational a nd educational status), household and family composition, housing characteristics, mobility, journey to work characteristics, ethnicity, and local economic activities (tax structure, expenditure for police and fir e service). Also, county data denoting the availability, use of, and need for health services were used in the analysis. The variables were extracted from the 1980 decennial census or other government statistics for counties for the same period. Their val ues were split into five categories that denote very low, low, moderate, high and very high levels of a variable. Generally, designation of very low to very high categories was based on the quartiles of the distribution of a variable in 1980. It should be noted that the variable values utilized were not for characteristics of subpopulations, but always for the total population. Variables such as the metropolitan-nonmetropolitan, rural-urban or density status of counties were not inclu ded, because the emphasis of this work was on the economic, social, and health variables rather than the usual geographic, central place or density characteristics. The overlap of the GOM types with these "external" geographic variables, however, is itsel f an interesting and important issue and will be discussed in the next section.

Pure Types With Concentrations of Counties That Are Low Density

After executing the GOM procedure for 1980 county data, the next step was to identify the pure types that had concentrations of sparsely populated and less densely settled counties (as noted, collectively referred to as low-density counties). Pure type s were designated as having concentrations of sparsely populated or low density counties if the percentage of these counties in a pure type was equal to or greater than the average for these designations for all counties in the coterminous United States i n 1980 (13.9% for sparsely populated, and 23.2% for low density). Nine of the 27 pure types that were produced by the analysis (PTs 1, 2, 5, 10, 11, 13, 14, 23, and 27) met the designated criteria. Collectively, these nine pure types contain 79.5% of the 425 sparsely populated counties in the coterminous counties in the United States and 70.9% of the 712 low-density counties. Table 1 shows the distribution of the nine pure types by county population density categories. The pure types are listed in order o f decreasing proportions of low-density counties included in each type. Pure Types 2,1,5, and 10 contain the greatest percentages of sparsely populated and low-density counties; consequently, the structural characteristics of these pure types will be emph asized in this letter. The percentage of counties that are sparsely populated declines from 91.4% in PT2 to 25.20 in PT10; the percent of counties that are low density from 98.1 in PT2 to 57.72 in PT10. For the remaining pure types with above average perc entages of low density counties, the percent of low density counties in a pure type ranged from 38.26 in PT14 to 23.24 in PT 23.

Table 2 classifies the nine pure types by their degree of rurality using a common rural-urban continuum coding scheme (Beale, 1983). Examination of Table 2 reveals that all counties in Pure Types 2 and 1 are nonmetropolitan (not part of the daily labor market area of a big city) and the vast majority (at least 75%) have populations that are totally rural (less than 2,500 urban persons in a county) and without easy access to a metropolitan area (not adjacent to a metropolitan county). Pure Types 5 and 1 0 are slightly less rural than Pure Types 2 and 1. While the vast majority of counties in Pure Types 5 and 10 are nonmetropolitan, there are a few metropolitan counties assigned to these types (2.3% for PT5 and 0.8 for PT10). Also, consistent with the sli ghtly less rural character of Pure Types 5 and 10, there are fewer totally rural counties (less than 45% compared to over 78% for Types 2 and 1). Excluding PT13, the remaining pure types are considerably less rural than Pure Types 2, 1, 5, and 10 (less th at 21% to the counties in the remaining types are totally rural). Like Pure Types 2 and 1, almost all counties in PT13 are totally rural (83.0%)

Table 2. Percentage Distribution of Frontier-Like Pure Types by
1983 Beale Urbanization Classification*

* Source: Beale, C. (1983). Rural-urban continuum code 1980. Unpublished data. Washington DC: Economic Development Division, Economic Research Service, U.S. Department of Agriculture

** Adjacent refers to counties that are adjacent to metropolitan areas

Principal Characteristics of the Low Density Pure Types

The principal characteristics that describe the similarities and the differences among the nine pure types can be found in 12 key variables. These county variables include population size, working in county of residence (including working at home), eco nomic status, educational status, ethnicity, presence of husband-wife households, older housing, employment in resource dependent industries, and selected health conditions (including presence of physicians and hospital beds, and use of inpatient and outp atient health facilities). These characteristics of the pure types are summarized in Table 3, which is organized to show both the "defining characteristics" of the pure types (characteristics that distinguish a pure type from other pure types), and other characteristics that help describe their character and make-up.

• Population Size. Except for two pure types (PT14 and PT23), all the pure types have small populations (less than 45,000 persons) and nearly half of the pure types (PTs 1, 2, 10, and 13) have populations that are very small (under 15,000 persons ).
• Working in county of residence. Most of these low-density pure types have a high or very high percentage of employed persons that work in their county of residence (70% or more of labor force) or work at home (10% or more of labor force). Pure Type 13 differs from the other designated pure types in that a low or very low percentage of the residents of these counties work in their county of residence (less than 60% of the labor force).
• Economic status. A low or very low economic status is a characteristic for three (PTs 1, 5, 10) out of the four pure types with very high percentages of low-density counties. The remaining six pure types have a status that is "not low or very l ow".
• Educational status. Most pure types had at least a moderate educational status (50% or more of the population aged 18 and over have completed high school), and for some it was a defining characteristic.
• Ethnicity. Most of the low-density pure types, for which there is information on the white population from the GOM analysis, have high or very high percentages of white persons in their populations (90% or more). Only for PT5 is the percentage of the population white not at least high and a defining characteristic. The percentage of PT5 population that is white is generally between 80 and 90%. Consistent with having a population that is only 80 to 90% white, the presence of Hispanic persons in PT5 counties is both defining and at least high (5% or more). Moreover, for most of PT5 counties (67%), 20% or more of their populations are Hispanic. Of the 27 pure types produced by the GOM analysis, this is the only pure type for which this is true.
• Husband-Wife Households. With the exception of PT10 counties, counties in all of the other designated pure types generally have a high or very high percentage of husband-wife households (65% or more of the households). Moreover, a high or very high percentage of husband-wife households was a defining characteristic for PT1, PT5, and PT11 counties.
• Older Housing. Four of the seven low-density pure types for which there is information (PTs 1, 2, 10, and 11) are defined by high or very high concentrations of older housing (over 60% being built prior to 1951).
• Employment in Resource Dependent Industries. Only two of the pure types, PT1 and PT2, are characterized by high or very high concentrations of persons employed in resource dependent industries (30% or more). Employment in resource dependent ind ustries is often cited as a hallmark of frontier areas. This appears to be the case only in those pure types with the highest percentage of low-density counties.
• Health Services. Half of the low-density pure types (PTs 1, 2, 5, 13 and 14) generally have low or very low numbers of physicians per 1000 persons (less than 0.7). The remainder did not. For six out of the nine low-density pure types (PTs 5, 10 , 11, 13, 14 and 23), the number of hospital beds per 1000 persons is moderate or below (<1.1 per thousand). For seven out of the nine low-density pure types (PTs 2, 5, 10, 11, 13, 14 and 23), the number of inpatient days per 1000 persons is moderate o r below (<4). For four out of the nine low-density pure types (PTs 1, 5, 11 and 13), the number of outpatient visits per 1000 persons per year is generally low or very low (<6). Though a lack of health services is often described in extremely rural areas, this generalization does not appear to hold for all low density counties. The exceptions may be important to understand when considering the adequacy of frontier-area health and mental health services.

Table 3. Distribution of Designated Pure Types by Selected Variables

Note:

Characteristics of the Low Density Pure Types

The preceding characteristics of the low-density pure types help to paint a strong portrait of several types of frontier populations in different parts of the United States. The defining characteristics of the four pure types where the counties are pre dominately low density are presented below. These portraits echo the descriptions of frontier populations found in popular literature and relayed by mental health professionals in these areas (see Ciarlo et al. 1998).

The average person living in a PT2 county would be a white, high school graduate, most likely married, with a high to a very high per capita income who works in his/her county of residence, commutes less than 10 minutes to work and lives in an older ho me (built prior to 1951), which may be modular (10% or more) or rented (20 to 40%). Males are employed full-time, while women, if seeking work, are likely to be employed. Often, employment was in resource dependent industries (40% or more of the labor for ce) such as agriculture. Employment in service or manufacturing industries is likely to be very low (less than 10% of labor force).

The population in these counties was very small (<15,000) and stable. The ratio of adult males to females was unusually high; there were often more adult males than adult females. Most counties in this pure type had less than 1% low birth weight bab ies a year, though some had a very high (2%) percentage. The rate of death by cirrhosis of the liver was very low (<0.1/1000 persons) in most counties, but very high (18/1000 persons) in a few counties. The counties also had a very low percentage of di sabled adult males. Most counties had very high amounts of federal funds for agriculture and natural resources. Their taxes were very high and the local spending per person was very high, including for education and highways.

The average person living in a PT1 county would be white, married, and living with his/her spouse and their children. They would have lived in the same, small community for at least the last five years. During this time they would have owned their own, older (built before 1951) home. While household income would generally be low, most men and women would have a high school education. If male, he would be likely to be employed and working in a mid-level job. Often, this employment would be in a reso urce dependent industry. If female, she would be likely to work part-time. Both would work in their county of residence (or at home) with less than a ten-minute commute to work.

The population in these counties was very small (<15,000) and stable over a ten-year period. Most counties in this pure type had less than 1% low birth weight babies a year, a very low percentage of disabled adult males, and a very low crime rate. A lso, the counties are defined by moderate to high taxes and by moderate to very high expenditures for education. Importantly, medical service availability (number of doctors and hospital beds) is not a defining characteristic for this group of counties. H owever, the values for outpatient visits for this type did tend to be "low" or "very low".

An average resident of a PT5 county would be white (generally, 80 to 90% of county populations), married and living with his/her spouse and their children, often, in a town. Mothers with children are not likely to be employed (generally, less than 50%). He/she would work within the county and have a commuting time of less than ten minutes. Household income would be low and there is a good chance the family would be living at or below poverty level (15 to 30% of PT5 county populations, have incomes at or below the poverty level). He/she might not have completed high school (30 to 70% of persons 18 and over in PT5 counties, complete high school). If male, he would be working in a low occupational status job (generally, 40 to 45% of the male labor for ce); if female, chances that she would be working in a high status job exceed those of males. The likelihood that the average resident would be employed in resource dependent industries is less than that of PT2 or PT1 counties but still sizable (generally , between 20 and 40% of the labor force). The likelihood of employment in manufacturing industries is very low (less than 10% of labor force).

Unlike PT2 or PT1 counties, the populations of PT5 counties generally include some Hispanic persons (at least 5% of county populations and often 20% or more). Counties in this pure type have a high to very high percentage of low birth weight babies. Th e percentage of adult, disabled males is moderate to low. Most counties have very high levels of federal funds for community resources and moderate local taxes. Expenditures for education and highways in these counties are moderate.

An average person in a PT10 county would live in a county with a small population (15,000 or less) that is characterized by concentrations of elderly persons (median age of county residents is 54 or greater), few children, low economic status and r esidence in older (built prior to 1951) stand-alone home (90% or more of the dwelling units). While not as high as PT2 or PT1 counties, persons in PT10 counties generally work in their county of residence. Many of the residence of these counties, particul arly the elderly, either live alone or with non-related adults and receive social security payments.

Employment in resource dependent industries, like agriculture, is often high (30 to 40% of the labor force) but below that of PT2 or PT 1 counties. Further, male labor force participation is moderate and often part-time. Most of these counties have a m oderate rate of doctors (MD and OD) per 1,000 population (between 0.7 and 1.1 per 1000). Reflecting an aging population, the death rate in these counties is very high.

All of the defining characteristics (characteristics that distinguish one pure type from another) of the nine low-density pure types are further detailed in Appendix B (for even more detailed descriptions see Goldsmith, et al., 1998). This appendix fir st presents a detailed analysis of PT1. Thereafter, the characteristics of each of the remaining low-density pure types are discussed in comparison to Pure Type 1. When a defining characteristic of a pure type is clearly different from that of PT1 (i.e., no overlap in the distribution of the variable), the information about the variable is underlined.

Location of the Low-Density Pure Types

The location of the counties in Pure Types 2,1,5, and 10 - the types in which low-density counties predominate - can be found in Map 1, Map 2, Map 3 and Map 4. This information is summarized below. The location of the remaining pure types with above average percen tages of low density counties can be found at the following web site: http://psy.utmb.edu/research/frontier/gom/gom.htm.

• Pure Type 2. Overwhelmingly, the PT2 counties are sparsely populated (91.4%) and are concentrated in the west north central states of Kansas and Nebraska (27.1% of the 96 sparsely populated counties of this type), the west south central state o f Texas (21.9%), and in the mountain states of Colorado and Montana (14.6%).
• Pure Type 1. Approximately 91% of the PT1 counties can be found in the west north central states of Iowa, Kansas, Minnesota, Nebraska, North Dakota and South Dakota. Of these, the sparsely populated counties are concentrated in the states of Ka nsas, North Dakota and South Dakota (74.0% of the 73 sparsely populated counties in the type), whereas the densely settled counties are concentrated in the states of Iowa and Minnesota (71.8% of the 39 densely settled counties in the type). The less dense ly settled counties can be found in Nebraska (40.5% of the 37 counties).
• Pure Type 5. Concentrations of the 86 PT5 counties are found primarily in the west south central state of Texas (66.3%) and the mountain state of New Mexico (11.6%). Sparsely populated PT5 counties are also found primarily in Texas (54.8 of the 31 counties) and New Mexico (25.8% of the 31 such counties). Densely settled counties (74.2% of the 31 such counties) and less densely settled counties (70.8% of the 24 such counties) are concentrated in Texas.
• Pure Type 10. Of the 123 Pure Type 10 counties, 46.3% are found in the west north central states of Iowa, Kansas, Missouri, and Nebraska and 41.5% are found in the west south central states of Oklahoma and Texas. The state with the highest numb er of sparsely populated and less densely settled counties is Texas (58.1% of the sparsely populated PT10 counties and 25.0% of the less densely settled counties). After Texas, Kansas had the most sparsely populated or less densely settled counties (12 or 16.9%). Missouri, followed by Texas, had the most densely settled PT10 counties (13 and 10 out of 52, respectively).

1980-1990 Sociodemographic Changes

To determine if dramatic changes took place in the social and demographic characteristics of the designated pure types counties between 1980 and 1990 (which might alter the composition of the pure types or the classification of individual counties), th e distributions for eight key variables (total population, percentage of the labor force working in county of residence, percentage of population in poverty, percentage of persons 25 and over with less than 9 years of education, percentage of population w hite, percentage of population Hispanic, percentage of households that were married couple households (husband-wife families), and number of physicians (both MD's and OD's) per 1000 persons ) were compared for both 1980 and 1990. The results of the compar isons indicate that, in general, the sociodemographic characteristics of the designated pure types either changed very little or changed in the same direction as all counties in the United States. The variables that were relatively stable include the size of the population of counties, the percentage of the labor force that work in their county of residence, percentage of population that was white, percentage of population that was Hispanic, and the number of physicians per 1000 population. Important chan ges did take place in the distributions of the remaining variables.

The 1980-1990 changes for each of the variables that were compared are presented in Chart 1. In general, populations tended to increase slightly, white percentages declined slightly, and working outside one's county of resident increased slightly. For other key variables, changes were also in the same direction, but greater. These included a reduction in persons with less than nine years of schooling, and a reduction in the percentage of husband-wife households. But there were fewer consistent trends f or health-related variables. For example, some pure types increased in physicians per 1,000 population, while other decreased. Similarly, economic status improved for some pure types, but not for others.

Chart 1: Sociodemographic and Economic Change in the Pure Types Between 1980 and 1990

REFERENCES

Ciarlo, J. A., Wackwitz, J.H., Wagenfeld,M.O., Mohatt,D.F., & Zelarney, P.T. (1996). Focusing on "Frontier": Isolated Rural America (Letter to the Field No. 2). Denver, Colorado: Frontier Mental Health Service Resource Network, University of Denver.

Goldsmith, H.F., Holzer III, C.E., Woodbury, M.A. & Ciarlo, J.A., with Stiles, D. (1998) Frontier-Like Counties with different Types of Sociodemographic, Economic and Health/Mental Health Characteristics: A Grade of Membership Analysis. Frontier Mental Health Services Resource Network. Manuscript in preparation

.

Appendix A

Quartile Distributions of the 1980 Variables Used to Conduct the Grade of Membership Analysis of All Counties in the Coterminous United States.

	County Health/Quality of Life Characteristics	Very Low/Low**	Moderate	High/Very High
	Risk Factors
1.	% 81 births low birth weight	<1.0/1.0-1.3	1.3-1.6	1.6-2.0/2.0+
2.	Cirrhosis deaths per 1000 pop	<.01/.01-.07	.07-.13	.13-.18/.18+
3.	Death rate	<7/7-8	8-11	11-13/13+
4.	1981 known crimes/1000 tot pop (1980)	<15/15-30	30-45	45-60/60+
5.	1981 violent crime/1980 tot pop	<15/15-30	30-45	45-60/60+
21.	Inpatient days/1000pop/year	<2/2-4	4-7	7-10/10+
22.	Outpatient visits/1000 pop	<2/2-6	6-10	10-15/15+
52.	% males 16-64 not inmates with disability/handicapped & unable to work*	<2/2-4	4-6	6-10/10+
53.	% females 16-64 not inmates with disability/handicapped & unable to work*	<2/2-4	4-6	6-10/10+
	Service Availability
19.	# MD+OD/1000 pop	<0.4/0.4-<0.7	0.7-1.1	1.1-1.5/1.5+
20.	# hospital beds/1000 pop	<2/2-4	4-7	7-10/10+
	County Population Growth and Commuting Patterns
6.	% works within county of residence*	<50/50-60	60-70	70-80/80+
7.	% works at home*	<2/2-5	5-10	10-15/15+
8.	Median minutes reported commute	<10/10-15	15-20	20-30/30+
	Change and Stability*
29.	% change 70-80 pop*	<0/0-10	10-30	30-40/40+
30.	Total # persons (thousands)*	<15/15-45	45-150	150-300/300+
	Mobility/migration
54.	% total persons diff house than 1975 (mobile persons)*	<30/30-40	40-50	50-60/60+
55.	% total persons 5+ diff county than 1975 (migrants)*	<15/15-20	20-30	30-40/40+
	County Socioeconomic Status
	Economic Status
40.	Total per capita income (thousands)*	<6/6-7	7-9	9-10/10+
70.	Median Household income (thousands)*	<10/10-13	13-16	16-20/20+
71.	% persons below poverty level*	<10/10-15	15-20	20-30/30+
72.	Median value owner occupied dwelling units (thousands)*	<20K/20-30	30-40	40-50K/50+
73.	Median value renter occupied dwelling units (dollars)*	<100/100-150	150-200	200-250/250+
38.	% income div, interest, rent	<10/10-15	15-20	20-25/25+
39.	% income transfers	<10/10-15	15-20	20-25/25+
24.	70-80 dif % pop in poverty*	<-15/-15 to -10	-10 to -1	-1 to 1/1+
	Educational Status
50.	% total persons 18+ comp 4 years high school*	<40/40-50	50-70	70-80/80+
51.	% total persons 25+ with < 8 years education*	<10/10-20	20-30	30-40/40+
	Occupation Status
46.	% total males low occupational status (operators, fabricators, laborers including farm and agricultural, service workers and workers in fishing/forestry occupations)*	<30/30-35	35-40	40-45/45+
47.	% total males high occupational status (executive, administrative and managerial occupations (except farm) or professional specialty occupations)*	<10/10-15	15-20	20-25/25+
48.	% total females low occupational status (see 46)*	<30/30-35	35-45	45-50/50+
49.	% total females high occupational status (see 47)*	<13/13-17	17-20	20-25/25+
	County Ethnicity
25.	% pop minority*	<5/5-15	15-30	30-45/45+
26.	% pop Hispanic*	<1/1-2	2-5	5-20/20+
27.	% pop white*	<60/60-80	80-90	90-98/98+
28.	% pop black*	<1/1-5	5-20	20-40/40+
	County Demographic/Household Composition
	Marital Status
31.	% females 15+divorced/separated*	<4/4-6	6-8	8-10/10+
32.	% males 15+ divorced/separated*	<4/4-6	6-8	8-10/10+
	Household Composition
56.	Median household size*	<2.3/2.3-2.4	2.4-2.5	2.5-2.8/2.8+
57.	Total males/100 females18-64 in HH (sex ratio)*	<90/90-95	95-99	99-106/106+
58.	Median age Householders*	<45/45-48	48-51	51-54/54+
59.	Total persons <18/100 persons 18-64 in HH (youth dependency)*	<45/45-50	50-55	55-60/60+
60.	Total persons 65+/100 persons 18-64 in HH (aged dependency)*	<15/15-20	20-25	25-30/30+
61.	% Households with/householder+spouse present (married couple HH)*	<50/50-60	60-65	65-70/70+
62.	% Families with 1 parent and own children (one parent families with children)*	<10/10-15	15-20	20-25/25+
63.	% fem 16+ with own children <18 working*	<45/45-50	50-60	60-65/65+
64.	% units with householder age 65+ who live alone (aged living alone)*	<6/6-9	9-11	11-14/14+
65.	% units with total householders that are nonfamily (nonfamily HH)*	<18/18-21	21-25	25-28/28+
41.	# total persons in group quarters*	<100/100-500	500-1500	1500-100K/100K+
	County Housing Characteristics
23.	% occ housing built prior to 1951*	<40/40-50	50-60	60-70/70+
66.	% occ units non-modular detached*	<70/70-80	80-85	85-90/90+
67.	% occ units modular*	<5/5-10	10-15	15-20/20+
68.	% occ units rented*	<20/20-25	25-30	30-40/40+
69.	% occ units 1 person/room	<2/2-4	4-6	6-10/10+
75.	% lots city/suburban or less 1 acre*	<25/25-50	50-80	80-95/95+
	County Industrial Base/Employment Structure
	Federal Funds
9.	Fed funds($) for agriculture/natural resources per capita	<70/70-200	200-500	500-1K/1K+
10.	Fed funds($) for community resources per capita	<150/150-300	300-600	600-1K/1K+
11.	Fed funds($) for space per capita	<30/30-80	30-150	150-500/500+
12.	Fed funds($) for human resources per capita	<10/10-20	20-60	60-100/100+
13.	Fed funds($) inc security per capita	<7/7-9	9-11	11-13/13+
	Local Taxes
14.	77 local taxes per capita	<100/100-200	200-400	400-500/500+
15.	77 local cur expenditures per capita	<300/300-450	450-600	600-750/750+
16.	77 local cur educational expenditures per capita	<200/200-250	250-350	350-400/400+
17.	Local fire and police expenditures per capita	<10/10-220	20-40	40-70/70+
18.	77 local cur highways per capita	<20/20-40	40-60	60-90/90+
	County Employment Levels and Participation
33.	% total females 16+ in labor force including military*	<45/45-50	50-55	55-60/60+
34.	% employed in manufacturing*	<10/10-20	20-30	30-40/40+
35.	% employed in services*	<10/10-15	15-20	20-25/25+
36.	% employed in state/local government*	<10/10-15	15-20	20-25/25+
37.	% employed in resource dependent industry*	<10/10-20	20-30	30-40/40+
74.	Total males 16+ in labor force including military*	<70/70-75	75-80	80-85/85+
42.	Males part-time employment*	<9/9-11	11-13	13-15/15+
43.	Females part-time employment*	<25/25-30	30-35	35-40/40+
44.	% total males 16+ in labor force unemployed*	<3/3-5	5-7	7-10/10+
45.	% total females 16+ in labor force unemployed*	<3/3-5	5-7	7-10/10+

* Denotes variables from the 1980 decennial Census (more detailed definitions of the variables can be found in NIMH Series BN-No.4 [1984]). Other variables are from the 1986 Area Resource File or unpublished USDA files.

** Designations vary from very low to very high based on quartiles for a particular variable such that for 0 to the first quartile is designated very low and from the first quartile to the second is low and so on.
 

Appendix B.

Pure Type Characteristics

	Pure Type 1	Pure Type 2	Pure Type 5	Pure Type 10	Pure Type 11
Social and Economic Characteristics	Small and declining populations with low mobility that work in their county of residence Economic status low but educational/occupational status moderate Population almost totally white Husband-wife households, many with children Housing older, owned detached single dwelling units Male labor force participation high and Female part-time employment high Employment in resource dependent industries generally high Local taxes not low/very low Local educational expenditures at least moderate and Local expenditures fire/police low/moderate Federal fund human resources generally moderate	Small nonmobile populations that work in county of residence. Commuting to work time very low Economic and educational status generally greater than PT1 counties  More persons in low or high status occupations than PT1 counties More adult males than females Very few persons in group quarters Very few one parent families (implication concentrations of Husband-wife households) Very few Black persons but a few Hispanic persons Housing older, often modular and rented Male labor force participation very high/high Male part-time employment very low/low Unemployment very low/low Employment in manufacturing or service industries low Employment in resource dependent industries high Local taxes very high Local expenditures for education and highways very high/high Federal funds for agriculture/ natural resources very high/high and Federal funds human resources low/very low	Many work in county of residence and have short commute to work (generally greater than PT1 counties) Economic status low but educational status moderate/low Male occupational status low, female occupational status moderate Concentrations of white (less than PT1 counties) husband-wife households with children Few one parent families with children Some Hispanic and perhaps Native American residents Male labor force participation at least moderate, female moderate or below Female unemployment low/moderate  Employed mothers below average  Employment in manufacturing very low and employment in resource dependent industries moderate (below PT1 counties) Local taxes moderate Local expenditures education moderate/high (usually exceed PT1 counties) and local expenditures highways moderate/low Federal funds communication generally very high and Federal funds human resources at least moderate	Generally work in county of residence (level generally below PT1 counties) Small aging populations Economic status low Many nonfamily households, older persons living alone and few households with children Federal funds for income security very high Female labor force participation and unemployment low/very low Male labor force participation moderate/low Male part-time employment very high/high Employment in resource dependent industries high/moderate (but below PT1 counties) and employment in State/local government predominantly low		Commuting to work time low (exceeds PT1 counties) Working at home high/moderate (PT1 counties are very high) Predominantly white Husband-wife households Economic state moderate/high  Educational/occupational status exceed PT1 counties More women, including mothers, in the labor force High male labor force participation Female unemployment low and part-time employment very high Moderate employment in resource dependent industries Federal funds agriculture/natural resources moderate/high and Federal funds human resources at least moderate Local educational expenditures moderate/high Local expenditures fire/police high/very high
Risk Factors and Service Availability	Very low percent of births that are low birth weight  Very low number of crimes per 1000 persons	Very low percent of births that are low birth weight  Very low number of crimes per 1000 persons Very low percent of adult males with a work disability	High or very high percent of births that are low birth weight Moderate or low percent of adult males with a work disability	High or very high death rate Low or moderate number of physicians per 1000 persons	Very low percent of births that are low birth weight  Low percent of adult males with a work disability

 
 

	Pure Type 13	Pure Type 14	Pure Type 23	Pure Type 27
Social and Economic Characteristics	Small populations that work outside county of residence and have commuting to work times that are high/very high Few males in high status occupations (inference socioeconomic status of counties low) White husband-wife households Most live in single dwelling units that are not in small towns/cities Employment in resource dependent industries is moderate (below PT1 counties) Federal funds for space very low Local expenditures for fire/police low/very low	Migration/mobility, growth high/ very high (exceeds PT1 counties) Commuting to work time low/moderate Moderate economic status and high occupational status  High/moderate educational status Predominantly white (but below PT1 counties) Some disrupted families, and husband-wife households moderate Residence in older housing low Some housing modular, overcrowded and in small towns/cities Federal funds human resources at least low Employment in resource dependent industries very low/low Female unemployment high/very high Local taxes moderate/high Local expenditures education at least moderate and local expenditures fire/police high	Population generally work in county of residence (below PT1 counties) Population growth moderate Economic status moderate Young, large families (most likely husband-wife) Housing generally not old Male labor force participation high Employment State/local government low Federal funds income security low(below PT1 counties)	Small populations (generally, greater than PT1 counties) that generally work in county of residence Commuting to work time low (greater than PT1 counties) Males in high status occupations moderate (PT1 counties are low) Females in high/low status occupations moderate Persons in group quarter moderate Single dwelling unit moderate (below PT1 counties) Rented units moderate (above PT1 counties) Units in small towns/cities moderate Employment in resource dependent industries low  Federal funds space generally low Federal funds human resources moderate Local taxes moderate Local expenditures fire/police moderate
Risk Factors and Service Availability	Very low death rate from cirrhosis Very low number of inpatients days and outpatient visits per 100 persons Very low number of hospital beds per 1000 persons	Low number of hospital beds per 1000 persons	Moderate percent of births that are low birth weight Moderate or low death rate Low inpatient days Moderate outpatient visits Moderate number of hospital beds per 1000 persons	Moderate crime and violent crime rates Very high inpatient days Moderate number of physicians per 1000 persons

---

Write us with comments on our site This project is supported by the Center for Mental Health Services, Substance Abuse and Mental Health Services Administration Contract No. 280-94-0014

Frontier Mental Health Resource Network
Please send comments and suggestions on this home page to Dennis F. Mohatt at dmohatt@wiche.edu
http://www.wiche.edu/MentalHealth/Frontier/frontier.asp