Any significant effort to address the causes of poverty requires an understanding of the features of the problem. This includes a disclosure of how many poor there are in Kentucky, where they are, who they are, and if possible, why they are poor. This chapter presents the relevant data, examines its significance, and offers conclusions regarding the development of state policy addressing poverty.

The federal government defines "poverty" as three times the annual cost of a minimally adequate diet, as determined by the U.S. Department of Agriculture. This standard varies by family size, number of children, and age of the head of household. In addition, an annual adjustment is made for inflation, as measured by the Consumer Price Index. Table 2.1 provides the latest federal poverty thresholds. For example, a family of three with an income of $11,821 or less would be classified as "poor" according to federal poverty thresholds.

A common misconception about poverty is that the federal poverty thresholds offer an accurate and complete definition of who is poor. The thresholds have been criticized for a number of conceptual and technical shortcomings, which tend in some cases to overestimate or to underestimate the extent of poverty. The conceptual basis of the poverty thresholds -- three times the cost of a subsistence diet -- is no longer valid, as food now constitutes a smaller share (about one-fifth) of the average household budget than when the measure was first constructed, while other expenses, namely housing and child care, comprise a larger share. Another conceptual problem is that dollar thresholds imply a stark, black-and-white world: one is either "poor" or "nonpoor". In reality, economic status is a continuous spectrum of grays, on which poverty thresholds are finally an arbitrary cutoff. Thus, the barely "poor" (earning just below the poverty threshold) are likely to have more in common with the near-poor (earning just above the threshold) than with those who are abjectly poor (earning less than half the threshold) -- though their official poverty status would suggest otherwise. The traditional poverty thresholds capture neither depth of poverty nor proximity to poverty.

Some critics contend that the poverty thresholds underestimate poverty by ignoring the overall rise in national living standards. Since their development in the 1960s, the thresholds have been adjusted only for inflation. The result is a measure of poverty based on a constant standard of living (in real terms), or an absolute concept of poverty. Relative poverty, on the other hand, would reflect rising living standards as well as rising prices. In this view, the current thresholds are too low, since they are based on a 30-year-old standard of living, and the extent of poverty is therefore understated. It should be noted that in periods of declining prosperity, the opposite would hold true: an absolute measure of poverty (such as the existing thresholds) would overstate poverty over time, in comparison with a relative measure. By some measures, living standards in this country have been in decline since the mid-’70s. Thus, the long-term bias of an absolute versus a relative measure of poverty in this country is not necessarily one of understatement.

A related but broader criticism is simply that the thresholds understate a minimally adequate livelihood -- whether by not accounting for rising living standards over time, or simply by being set too low at the outset. This view is supported by surveys of the public which inquire how much money is needed "to get by in this community"; respondents generally suggest levels up to 50 percent higher than the current poverty threshold.

Charges that the thresholds overstate poverty center not on the level of the thresholds, but rather on the way family income is measured in comparison with the thresholds. In general, only cash income is considered; in-kind assistance, such as food stamps and health care (Medicaid), is not taken into account, nor is wealth (assets), such as savings and property.

Despite these shortcomings, however, the federal poverty thresholds are valuable as the only widely available measures which are consistent across geographic areas and available at a relatively disaggregated level. In terms of understanding the demographic and geographic characteristics of poor Kentuckians, no superior alternative is available.

[Insert Table 2.1: 1994 Federal Poverty Thresholds]

According to the 1990 Census, 19 percent of Kentuckians (approximately 682,000 individuals) have incomes below the federal poverty line. The state's poverty rate was the sixth-highest in the nation for 1989, and has been consistently high over the last 30 years. In 1969 and 1979, the state poverty rates were 22.9 and 17.6 percent, respectively. Thus, while government programs may have mitigated some of the effects of poverty on families, it is clear that they have not been successful in significantly reducing the proportion of Kentuckians who are poor.

Table 2.2 provides a breakdown of the income level of poor Kentuckians expressed as a percent of the federal poverty threshold. This table provides information about the depth of poverty in the state. As the table shows, those whose incomes fall below the federal poverty line are not uniformly "poor." Approximately 8.5 percent of the population (300,000 individuals) subsist on incomes less than half the federal poverty threshold.

[Insert Table 2.2: Kentucky Population for Whom Poverty was Determined: By Poverty Status, 1989]

The most common geographic analysis of the poor in Kentucky is at the county level. This section begins with that traditional level of geographic analysis, but extends the analysis to incorporate Census information about poverty at the subcounty level as well.

County Analysis

Figure 2.1 classifies counties in Kentucky by ranges of poverty rates. Counties with the highest poverty rates are, with the exception of Fulton County, located exclusively in eastern Kentucky. Those with the lowest rates are found in urban areas, particularly Jefferson, Fayette, Boone, Kenton, and Campbell Counties.

Although the county poverty rate is useful for showing the concentration of poor persons within a given area, it provides no indication of the number of poor in that area. Figure 2.2 illustrates the number of poor in each county. (The counties with the highest peaks have the greatest numbers of poor. The poverty rate and number of poor in each county are provided in the Appendix.) For example, the highest peak is found in Jefferson County, which has the greatest number of poor. The second-highest peak rises above Fayette County, with the second-highest number.

[Insert Figure 2.1: Poverty Rates By County in 1989]

[Insert Figure 2.2: Number of Kentuckians Living in Poverty By County in 1989]

Table 2.3 lists the counties with the five highest poverty rates and those with the five largest numbers of poor in order to illustrate the importance of using both measures to gain a more complete picture of county-level poverty within the state. Owsley County's poverty rate of 52.1 percent is the highest in the state, and indeed the nation. However, because the county has such a small population, this very high poverty rate represents a relatively small number -- 2,570 people in poverty. Contrast this with Jefferson County, the state's most populous county. While its poverty rate of 13.7 percent is approximately one-fourth that of Owsley County, it represents nearly 90,000 people, or 30 times as many poor as in Owsley County.

While the counties with the highest poverty rates are concentrated in eastern Kentucky, the counties with the largest numbers of poor people are in urban areas. The high rates of poverty in eastern Kentucky are as much a reflection of the absence of the nonpoor as they are an indication of the presence of the poor. The presence of large numbers of nonpoor in urban areas tends to obscure the fact that urban counties are home to a large number of poor residents.

[Insert Table 2.3: Top Five Kentucky Counties by Rate of Poverty and Incidence: From Population for Whom Poverty was Determined, 1989]

Targeting assistance to counties with the highest poverty rates draws assistance away from the counties where most poor people live. The use of county poverty rate "hides" those who are poor in the midst of relative affluence. A policy which targets programs to the poor, no matter who lives around them, will have a very different distribution than a policy which targets programs to the poor in counties with the highest poverty rates. Consideration should be given to whether the goal of a particular antipoverty program is to reduce the poverty rate of areas, or to change the poverty status of individuals. While the two goals are not mutually exclusive, it is important to understand that achieving the former does not necessarily achieve the latter. Indeed, the two measures can even move in opposite directions: the poverty rate of an area can fall even as the number of poor increases -- if the number of nonpoor increases by a proportionately greater amount.

The geographic distribution of poverty is different for various subgroups of the poor. Poor children are more likely to be concentrated in both eastern Kentucky and in Jefferson and Fayette Counties. Approximately 138,000 of the 234,000 poor children in the state are found in urban areas. Jefferson, Fayette, Boone, Campbell, and Kenton Counties combined are home to nearly a third of the state's poor children.

Altogether there are approximately 91,000 elderly poor, and they are less likely than other subgroups of the poor to be concentrated in eastern Kentucky and urban areas, and more likely to be found in western and south-central Kentucky. Pulaski, Madison, Barren, McCracken, and Wayne Counties, in particular, have a relatively larger number of poor elderly than they have of poor children or of poor people in general.

Black Kentuckians, particularly those who are poor, reside in Jefferson, Fayette, Christian, and Hardin Counties. Of the 160,000 black Kentuckians in poverty, most (63 percent) live in these four counties. While also relatively populous, Boone, Kenton, and Campbell Counties account for only 2 percent of black Kentuckians living in poverty.

Nearly 52 percent (51,519) of female-headed families with related children under 18 in Kentucky are living in poverty. This is in contrast to 13.7 percent (57,214) and 29.9 percent (5,683) for married-couple and male-headed families, respectively. Although their greatest concentration is in urban areas, female-headed families with children under age 18 are found throughout the state. Wherever they live, they are likely to be poor.

To complete the Census counts each decade, the Bureau of the Census divides the country into geographic units that are smaller than counties or cities. Specifically, "block groups" are defined to be comparable in size to a large neighborhood. In the 1990 Census, the Bureau defined approximately 3,500 block groups in Kentucky, with an average of 1,044 residents per block group. The Bureau of the Census makes some of the data it collects available for these subcounty geographic areas. When this data is analyzed, patterns emerge which are different from those exhibited in the county-level analysis.

As shown earlier, analysis of county poverty rates indicates that areas with the highest concentrations of poverty are found almost exclusively in eastern Kentucky. On the other hand, if the poverty rate is shown at the block-group level, it becomes apparent that there are pockets of poverty throughout the state. The overwhelming majority of block groups (3,443) contained at least some poor persons. Thus, while poverty rates are highest in eastern Kentucky, and while most poor people inhabit urban areas, there is really no area of the state that is unaffected by problems associated with poverty. Antipoverty programs which target recipients on the basis of either poverty rate or the number of poor address the problems of being one of mostly poor or one of many poor, but may not address the plight of the poor who fall outside those statistically defined areas.

A combined poverty index addresses some of the limitations regarding the use of either the poverty rate or the number in poverty as the sole measure of poverty in an area. This index presents the over- or underrepresentation of poverty in a local area -- either for all people or for some group of people -- relative to the average rate of poverty for Kentucky. It is stated in terms of people: specifically, the number of people by which poverty in the area, and for the relevant group, overrepresents or underrepresents the overall state average. The combined index accounts for both the concentration (or percent) of poor and the incidence (or number) of poor.

For example, assume that a local area with a population of 1,000 has 290 people living in poverty, or a poverty rate of 29 percent. If this area had the same rate of poverty as the state as a whole (19 percent), it would have only 190 people in poverty (rather than 290) and would perfectly represent the state in terms of poverty rate. However, with 290 people in poverty, the area could be described as disproportionately poor, or overrepresented in poverty relative to the state. The combined index measures the magnitude of that overrepresentation by stating how many people would have to be raised above poverty for the local area's poverty rate to equal the state's. In this case, the local area is overrepresented by 100 poor people (290 minus 190); that is, if 100 fewer people were in poverty, the area's poverty rate would equal the state average of 19 percent. The same applies to the opposite case of underrepresentation: the index measures how many more people would have to be impoverished for the area's poverty rate to equal the state's.

A set of maps was developed to display the under- and overrepresentation of the poor in each of the block groups in Kentucky. The first map (Figure 2.3) presents the index for all of Kentucky. However, because the populous areas contain so many block groups that are obscured by the scale of the full state map, separate maps are presented for Jefferson, Fayette, and the northern Kentucky counties of Kenton and Campbell as well (Figures 2.4 - 2.6).

In terms of the geographic distribution of poverty at the subcounty level, the maps reveal an important difference between urban and Appalachian poverty. Eastern Kentucky contains approximately 30 block groups in which poverty is overrepresented by at least 400 people. Jefferson, Fayette, and northern Kentucky have 20, 9, and 2 such block groups, respectively, with the same amount of overrepresentation. In particular, Jefferson and Fayette Counties have numerous small areas where poverty is extremely overrepresented -- areas in close proximity to more prosperous areas. On the other hand, eastern Kentucky contains relatively few prosperous areas, as measured by the combined index. Thus, although they are similar in magnitude (according to this index), the poverty of eastern Kentucky is spread over a large geographic area and is relatively uniform, while urban poverty is concentrated in small pockets near large areas of relative affluence.

A similar analysis was conducted of the over- and underrepresentation of subgroups of the poor, by block group. The conclusion from this analysis is that poor children are more overrepresented than the general population of poor in both eastern Kentucky and in Jefferson and Fayette Counties. Conversely, the elderly poor are less overrepresented in the urban areas, and more overrepresented in south-central and western Kentucky. Black Kentuckians, and particularly those who are poor, reside primarily in Jefferson, Fayette, Christian, and Hardin Counties. Female-headed families exist across the state, but wherever they live, they are likely to be poor. Their greatest overrepresentation is in urban areas.

Finally, use of all three tools to measure poverty in Kentucky (the county poverty rate, the number of poor per county, and the combined poverty index) helps to challenge another myth about poverty in Kentucky: the misconception that poverty occurs mostly in the eastern portion of the state. While counties in this geographic area have the highest poverty rates, the greatest number of poor Kentuckians are found in urban counties, particularly Jefferson, Fayette, Boone, Kenton, and Campbell Counties. The combined index reveals that Jefferson and Fayette Counties, taken together, account for nearly the same number of block groups overrepresented by at least 400 people in poverty as do all counties of eastern Kentucky combined.

While knowing how many Kentuckians are poor and where they live aids in the design of antipoverty efforts, it is of little help in understanding why certain Kentuckians fall into, and often remain in, poverty. This section presents a brief description of the demographic characteristics of poor Kentuckians in an attempt to further such an understanding.

Two assumptions guided the selection of the major groups for analysis in this section. First, the inexorable demographic changes associated with the aging of a large "baby boom" generation and the rise of a much smaller "baby bust" generation make Kentucky's economic future profoundly a matter of its children’s future. A high-quality labor force of sufficient size to allow a small state like Kentucky to compete in a global labor market cannot be assured if a large number of its children must enter that market burdened with the long-term disadvantage of growing up in poverty.

Second, in considering why children are poor, it is immediately apparent that children are poor not through any action of their own, but because the adults responsible for their care do not, or cannot, obtain sufficient incomes to raise themselves and their children out of poverty. Thus, the question of why children are poor is that of why the adults responsible for them are poor. Examining the characteristics of the latter group is a major focus of the analysis. Within this group, there is particular policy interest in exploring the characteristics of the working poor.

Table 2.4 summarizes some of the major demographic characteristics of adults with children in poverty. First, they are more likely to be young (19-25), female, black, and rural (non-farm) than adults living with children not in poverty. Second, according to Census data, they are more likely to live in a two-person household (one adult and one child) or in a household with 6 or more persons than is the comparison group. Third, adults with children in poverty are less likely to be in married-couple families and more likely to be female heads of households. Similarly, they are less likely to be married and more than twice as likely to be divorced, separated, or never married.

The single largest set of differences between adults with children in poverty and those with children above poverty involves reported work status. Of the adults with children in poverty, 31 percent reported that they lived in a family in which no one had worked in the previous year, compared to 1 percent in the other group of adults. Conversely, 54 percent of the adults with children not in poverty reported that 2 or more family members had worked in the previous year, compared to 21 percent of the adults with children in poverty. Even this statistic is somewhat misleading in that it indicates whether any person in the household worked at all, even if only for a few days, in the previous year.

[Insert Table 2.4: Characteristics of Adults in Households with Children]

The interruption in work status for adults with children in poverty is highlighted by the fact that two-thirds of them reported that they had not worked the previous week, compared to 27 percent of the other group of adults. Although 47 percent of adults living with poor children reported having worked in the previous year, only 12 percent of them reported that they worked in all 52 weeks. The corresponding percentages for the adults with nonpoor children were 84 percent and 51 percent, respectively. Overall, the differences in the amount of time worked are profound.

Since women are overrepresented in the group of adults with children in poverty, it might be assumed that a substantial amount of the disparity lies in the difference in work status between the sexes. Controlling for the gender of the respondent, however, does not greatly lessen the disparity. To simplify the comparison, a total number of work hours was calculated for each respondent by multiplying the number of weeks worked in 1989 by the usual number of hours worked per week. A person working 40 hours per week for 52 weeks works a total of 2,080 hours. This amount was divided into four equal amounts with those working fewer than 520 hours classified as working 25 percent of full-time hours or less, and so on for 50 percent, 75 percent, and 100 percent of full-time hours. Those who worked more than 2,080 hours were classified as working more than 100 percent of full-time hours.

As Table 2.5 clearly indicates, the differences in the amount of time worked are profound, even controlling for gender. For example, 35.7 percent of women with nonpoor children worked 0 to 25 percent of full-time hours, while 77.1 percent of women with poor children worked the same amount. For men, only 9.4 percent of men with nonpoor children worked 0 to 25 percent of full-time hours, while 52.8 percent of men with poor children worked similar hours.

[Insert Table 2.5: Percent of Full-Time Hours Worked in 1989 - Adults Living with Children]

Table 2.6 provides the reported sources of income for adults with children in poverty. Although it is commonly believed that most poor people are on welfare, this notion is incorrect. Only one-fourth of the adults living with children in poverty reported receiving any income from public assistance. And the largest proportion of these adults (42%) reported receiving wages and salary as their main source of income.

[Insert Table 2.6: Percent of Respondents Reporting Any Income from Various Sources - Adults with Children in Poverty]

While instructive for understanding who is poor, a summary of demographic characteristics does not answer the key question of why some adults do not (perhaps cannot) provide sufficient incomes to raise themselves and their children out of poverty. Demographic characteristics explained only 34 percent of the variation in family income in this sample. Education accounted for the greatest share of this variation (22 percent). Apparently, other variables relating to development, psychology, community, and welfare program restrictions must be studied in order to better answer the question of why. Analysis which moves beyond assessment of the individual characteristics of the poor is necessary to begin to understand the community dynamics that also affect how many people will be poor and who those poor will be.

For the purpose of this demographic analysis, "children" are defined as persons 18 years or younger. Table 2.7 summarizes the characteristics of this group. As the table shows, most poor children in Kentucky are white, although black children are significantly overrepresented in the group. Put another way, the poverty rate for black children is much higher than it is for white children. In terms of family characteristics, poor children are much more likely to live in households with 1 or 2 persons, or with 6 or more persons, than are nonpoor children. Thirty-eight percent of poor children live in female-headed households. Fifty-one percent live in married-couple families, compared with 87 percent of nonpoor children. Seventeen percent of poor children reported not attending school as of 1989, while 12 percent of nonpoor children were not in school.

Interestingly, while the majority of adults with children in poverty live in rural areas, the majority of poor children themselves live in urban areas. The difference is accounted for by the fact that rural children are more likely to live in married-couple families that have more adults present, while urban children are more likely to live in families with a single adult, usually a woman. Also, over half of all poor children live in families with incomes 50 percent or less of the federal poverty threshold. Thirty percent of "nonpoor" children live in families with incomes between 100 and 200 percent of the federal poverty line. While these children may not be classified as poor, they are best described as "near-poor," since they live in families with income levels that might not withstand any disruption -- such as a temporary layoff or the addition of another family member.

Forty-eight percent -- nearly half -- of Kentucky's children live in families with incomes 200 percent of the federal poverty level or below. Among black children, 70 percent are below 200 percent of the poverty level, and 28 percent are below 50 percent. In terms of family type, 39 percent of children in married-couple families are below 200 percent of the poverty level, and 7 percent are below 50 percent. The corresponding figures for female-headed families are 80 percent and 37 percent, respectively.

The single largest difference between poor and nonpoor children is the work status of the adults with whom they live. Thirty-one percent of all poor children live in a household where no one worked in the previous year. Only 1 percent of nonpoor children lived in a household where no one worked. Twenty-two percent of poor children lived with two nonworking parents, and 30 percent lived only with nonworking mothers. However, 27 percent of poor children lived in families with one working parent, and 3 percent lived with two working parents. The comparable figures for nonpoor children were 54 percent and 29 percent, respectively.

Note that the comparison is for number of workers only and does not control for number of hours worked, industry, or occupation. Therefore, it seems reasonable to conclude that the key predictor of whether a child is living in poverty is the presence of at least one working adult in the family. In terms of raising children out of poverty, it appears that it is the presence of a worker that matters more than the quality of the work, in terms of wages or salary.

[Insert Table 2.7: Characteristics of Children in Poverty (Age 18 and Under)]

Policy interest has focused recently on the plight of the working poor. It is widely held that those who are expending significant effort to move their children out of poverty may have a compelling claim for assistance. However, any analysis of the "working poor" must recognize that there is no universally accepted definition of this group. There are two steps involved in defining "working poor." First, of course, one must define "poor." When some discuss the plight of the working poor, they have in mind those who work full-time, but are still unable to access the main features of "the American dream," such as home ownership, reliable transportation, and some measure of economic security. The problem with this conception is that there is no firm measure of "poor." Therefore, for the purposes of this analysis, individuals are classified as "poor" if they live in families with incomes below the federal poverty line. As noted earlier, this income is substantially below the amount most survey respondents suggest as an adequate income for a family to "get by."

The second step involves determining when an individual is working. The Bureau of the Census defines "worker" as anyone who answered "yes" to the question, "Last year [1989], did this person work, even for a few days (italics added), at a paid job or in a business or farm?" This is the data used to quantify the number of "workers" in a family, as discussed above. However, in defining the "working poor" for this analysis, the decision was made to refine the definition of "worker" to include only those who reported having worked at least 76 percent of full-time hours in the previous year. Note that this definition of "working" does not include those who find themselves part of the contingent labor force -- either through working part-time (fewer than 30 hours a week) or intermittently (fewer than 39 weeks during the year). Contingent workers comprise an increasing share of the labor force, and are more likely than those defined as "working" in this report to have incomes below the poverty line.

When the above definitions are used, there are few Kentuckians who can be classified as "working poor." According to 1990 Census data, a relatively small percentage (15.5 percent) of adults with poor children were working more than 75 percent of full-time hours (Table 2.8). Examined from another direction, only 6 percent of adults classified as working more than 75 percent of full-time hours have family incomes at or below 100 percent of poverty. These percentages yielded a sample size too small for further analysis.

To increase the sample size, those who worked more than 75 percent of full time hours and had incomes between 100 and 200 percent of the poverty threshold were defined as the "working near poor." Therefore, for the purpose analyzing their characteristics, adults working more than 75 percent of full-time hours and falling below 200 percent of poverty are classified as the working poor and near-poor. Those working over 75 percent of full-time hours with incomes over 200 percent of poverty are designated the "working nonpoor." According to these definitions, 26 percent of working adults are classified as poor and near-poor, and 74 percent are classified as "nonpoor."

An analysis of their demographic characteristics reveals that the working poor and near-poor tend to be younger and are more often black and rural than the working "nonpoor." The working poor and near-poor are more likely to live in families of 2 persons or 5 or more persons. In addition, they are more likely to be responsible for the care of a greater number of children than the comparison group. Also, the working poor and near-poor are less likely to be in married-couple families and more than three times as likely to be in female-headed families.

In terms of educational background, the largest category of respondents in working poor and near-poor groups includes those with a high school education but no more. Persons in either group, however, are much more likely to have less than a high school education than the working "nonpoor." With respect to work status, the working poor and near-poor are much more likely to live in families with only one worker and much less likely to live in families with two or more workers.

Although this section discusses only a few characteristics, an examination of all the characteristics of working poor and near-poor Kentuckians reveals that having a job is the single most important characteristic of adults in determining whether their families live above or below poverty -- regardless of the income from that job or the number of hours worked. When considering the situation of those who work more than 75 percent of full-time hours, however, it is clear that the number of children in the family, the number of workers in the family, and the income from the job all have a significant effect on whether that family earns an income above 200 percent of poverty or whether they remain officially out of poverty, but at risk of falling into it.

[Insert Table 2.8: Characteristics of Adults Who Worked More Than 75 Percent of Full-Time Hours]

[Insert Table 2.9: Kentucky Population]

Several important demographic trends relating to total population changes and to various subgroups of the poor have occurred in Kentucky. In reference to the total population, Table 2.9 lists the number of Kentuckians in various subgroups in 30-year increments since 1900. As the table shows, the long-term trend in Kentucky (as in the U.S.) is that the number of young has declined both in absolute number and relative to other age groups, while the number of elderly has done just the opposite.

In 1979 the number of related children under 18 years old in Kentucky was 1,063,000. By 1989 that number had declined by 12 percent to 938,000. According to population projections by the Kentucky Data Center, this trend is expected to continue at least through the year 2020.

While the total number of youth is on the decline in Kentucky, the percentage of children who are poor is growing. According to Census data, the poverty rate for the subgroup of related children under age 18 in the state was 24.8 percent (229,530 children) in 1989. This is 4.9 percentage points greater than the comparable national rate for the same year, and 3.2 percentage points greater than the 1979 rate for Kentucky.

[Insert Table 2.10: Change in Poverty Status of Selected Demographic Subgroups]

What accounts for the rise in the percentage of children who are poor? From Table 2.10, note that the increase in the poverty rate among children between 1979 and 1989 occurred in tandem with a slight decrease in the number of children in poverty. This is because there was a much greater decline in the number of nonpoor children relative to the decline in the number of poor children. Several factors could lead to this situation: birth rates could be higher for those below poverty; children previously classified as above poverty could have moved into poverty; and families with children above poverty could have moved out of the state at a greater rate than those below poverty.

The total number of families (with or without children under 18) increased by 4 percent between 1979 and 1989, while those classified as below poverty increased by 14 percent, compared to 2 percent for families classified as above poverty. This stronger growth in the number of families below poverty could also be due to several factors: families could be forming at a greater rate below poverty; families previously classified as above poverty could have fallen into poverty; and families above poverty could have moved out of the state at a greater rate than those below poverty.

Note that the previous two references to time-series Census data limit the possible reasons for change to three factors: creation of families, change in classification of families from above to below poverty, and migration of families. While the change attributable to migration can be measured somewhat through the use of available Census data, the remaining two (creation and change in classification) are not measurable because the Bureau of the Census does not track particular individuals or families over time. Consequently, it is difficult to say which is the predominant cause of the observed changes.

In the case of Kentucky's elderly, their numbers are on the rise. According to Table 2.9, in 1960 there were 292,000 elderly, comprising 9.6 percent of the total population. This number grew to 464,999 in 1990 (12.6 percent of the total). Based on Kentucky Data Center population projections, Kentucky will have over 650,000 elderly in 2020 (16.7 percent of the total). The 1989 state poverty rate for the elderly was 20.6 percent. Although this represents a 2.7 percentage-point decline from 1979, Kentucky's poverty rate for the elderly still remains well above the national average of 12 percent for this group.

The foregoing chapter presented a brief description of poverty in Kentucky. Included was a discussion of how many poor there are, where they live, who they are, and how their numbers have changed over time. The attempt was to provide information that goes beyond a simple presentation of the poverty rate by county and a summary of numbers of people in various categories. Disaggregation of the numbers by subcounty regions and examination of the characteristics of individuals and families who are poor allowed for a more complete picture of poverty in Kentucky. The benefit of a more complete picture is that it helps challenge some common misconceptions about the nature of poverty in the state. Such misconceptions are the focus of the following chapter.