Appendix A
EMPLOYMENT AND ECONOMIC CONSEQUENCES OF VISUAL IMPAIRMENT
Andrew J. Houtenville
School of Industrial and Labor Relations, Cornell University
Working-age people with disabilities work less and have less household income than working-age people without disabilities.1 There are also dramatic differences in the kinds and levels of disabilities within the working age population with disabilities.2 Those with severe vision impairments are particularly disadvantaged, for they face many barriers in accessing employment. This paper explores the economic experience and program participation of working-age people with chronic vision-related conditions over the past two decades and compares their experience with those of other working-age people with chronic conditions.
DATA AND IMPORTANT SAMPLING ISSUES
The National Health Interview Survey (NHIS) contains the economic and chronic condition information needed to conduct this study. The NHIS is a complex multistage probability sample of the civilian noninstitutionalized population of the United States.3 The NHIS is collected by the National Center for Health Statistics (NCHS) in the Department of Health and Human Services. The federal government uses data from the NHIS to monitor trends in illness and disability. Researchers use data from the NHIS to analyze access to health care and health insurance and to evaluate federal health programs.
The NHIS collected information on an average of about 60,000 working-age individuals (100,000 individuals in total) annually from 1983 through 1996.4 This paper separates survey participants into subgroups by chronic condition and gender. Some of these subgroups contain very small numbers of individuals, i.e., very small sample sizes. Smaller sample sizes lead to less precise sample estimates. This paper pools multiple years together to boost the sample sizes in these subgroups.
Specific health conditions and impairments are captured in two distinct methods. The differences between these two methods are very important in the interpretation of statistics generated using the NHIS. Very early the survey participants are randomly asked one of six condition checklists. These checklists directly inquire about
specific conditions. Table A-1 contains the checklist relevant to vision and the other conditions addressed in this report. In addition to the checklists, in later parts all survey participants are asked screening questions to reveal general health, doctor’s visits, hospital utilization, sick days, and functioning difficulties. If participants answer yes to these screening questions, they are then asked what conditions caused these issues. The top panel of Table A-2 contains the set of screening questions, while the bottom panel contains an example of the probing questions that follow a screening question.
Thus only one-sixth of NHIS participants are directly asked about blindness and visual impairment and can also reveal blindness or visual impairment if they reveal having general health/functioning difficulties. The remaining five-sixths of NHIS participants reveal blindness or visual impairment only if they reveal having general health/functioning difficulties. As a result, the subsample of NHIS participants reporting blindness in the one-sixth sample is a random subsample of those reporting blindness. The subsample of NHIS participants reporting blindness in the five-sixths sample is a choice-based subsample of those reporting blindness because being in this subsample depends on responses (choices) to the screening questions.5
From this point forward, the term “random sample” refers to the one-sixth of the NHIS sample who were directly asked about their condition, and the term “choice-based” sample is used to refer to those who where asked about their condition after having revealed general health/functioning difficulties.
Prevalence, employment, income, and program participation statistics are calculated separately for random and choice-based samples. There are likely to be important differences between the two samples. The prevalence of blindness should be higher in the random sample than in the choice-based sample, because there are likely to be people reporting blindness who do not have general health/functioning difficulties, i.e., who answer “no” to the question in the bottom panel
TABLE A-1 Condition Checklist Received by the Random Sample
H1.2a. Does anyone in the family {read names} NOW HAVE – |
|
|
If “Yes,” ask 2b and c. b. Who is that? c. Does anyone else now have – |
A. |
Deafness in one or both ears? |
B. |
Any trouble hearing with one or both ears? |
C. |
Tinnitus or ringing in the ears? |
D. |
Blind in one or both eyes? |
E. |
Cataracts? |
F. |
Glaucoma? |
G. |
Color blindness? |
H. |
A detached retina or any other condition of the retina? |
I. |
Any other trouble seeing with one or both eyes EVEN when wearing glasses? |
J. |
A cleft palate or harelip? |
K. |
Stammering or stuttering? |
L. |
Any other speech defect? |
M. |
Loss of taste or smell which has lasted 3 months or more? |
N. |
A missing finger, hand or arm, toe, foot, or leg? |
O. |
A missing joint? |
P. |
A missing breast, kidney, or lung? |
Q. |
Palsy or cerebral palsy? |
R. |
Paralysis of any kind? |
S. |
Curvature of the spine? |
T. |
REPEATED trouble with neck, back, or spine? |
U. |
Any TROUBLE with fallen arches or flatfeet? |
V. |
A clubfoot? |
W. |
A trick knee? |
X. |
PERMANENT stiffness or any deformity of the foot, leg, or back? |
Y. |
PERMANENT stiffness or any deformity of the fingers, arm, or hand? |
Z. |
Mental retardation |
AA. |
Any condition caused by an accident or injury which happened more than 3 months ago? If “Yes,” ask: What is the condition? |
Note: In the NHIS, conditions are determined in two ways. First, participants receive one of six condition lists that ask them if they have a specific condition (this table contains list #2). Second, participants are asked broad questions to reveal general health and functioning (see the top panel of Table A-2). If participants reveal they have health or functioning difficulties, they are then asked what conditions cause these difficulties (see the second |
panel of Table A-2). This method misses those with conditions who have no such difficulties, while the first method captures those with conditions who have no health or functioning difficulties. So only one-sixth of the sample is directly asked about blindness. This one-sixth of the sample is a random sample, because being asked about blindness is not dependent on one’s response to another question. The remaining five-sixths of the sample is choice-based, because revealing blindness is dependent on one’s response (choice) to another question. Source: National Health Interview Survey Core Questionnaire, 1985-94, National Center for Health Statistics, Series 10, No. 199. |
of Table A-2. Similarly, employment rates and mean incomes are likely to be higher and program participation rates are likely to be lower among random sample members reporting blindness than among choice-based sample members reporting blindness. Choice-based members reporting blindness have already revealed health and functioning difficulties and are thus less likely to work or earn income and more likely to participate in government programs.6
TABLE A-2 Broad Health Questions Used to Screen for Condition Information
Screening Questions |
|
1. |
Does any impairment or health problem NOW keep [you] from working at a job or business? |
2. |
Does any impairment or health problem NOW keep [you] from doing any housework at all? |
3. |
Is [you] limited in ANY WAY in any activities because of an impairment or health problem? |
4. |
During those 2 weeks, did [you] miss any time from a job or business because of illness or injury? |
5. |
During those 2 weeks, did [you] miss any time from school because of illness or injury? |
6. |
During those 2 weeks, did [you] stay in bed because of illness or injury? |
7. |
Was there any {OTHER} time during those 2 weeks that [you] cut down on the things [you] usually does because of illness or injury? |
8. |
During those 2 weeks, how many times did [you] see or talk to a medical doctor? {include all types of doctors, such as dermatologists, psychiatrists, and ophthalmologists, as well as general practitioners and osteopaths.} |
9. |
{Besides the time(s) mentioned in [previously]} During those 2 weeks, did anyone in the family receive health care at home or go to a doctor’s office, clinic, hospital or some other place? 2b. Who received this care? |
10. |
{Besides the time(s) you already told me about} During those 2 weeks, did anyone in the family get any medical advice, prescriptions or test results over the phone from a doctor, nurse, or anyone working with or for a medical doctor? |
An Example of Probing Questions (These are the probing questions for the first screening question above.) |
|
A. |
What (other) condition causes this? Ask if injury or operation: When did [the (injury) occur? / [you] have the operation?] Ask if operation over 3 months ago: For what condition did [you] have the operation? |
B. |
Besides (condition) is there any other condition that causes this limitation? |
C. |
Is this limitation caused by any (other) specific condition? |
D. |
Which of these conditions would you say is the MAIN cause of this limitation? |
Source: Design and Estimation of the 1985-94 National Health Interview Survey, Series 2, No. 110, National Center for Health Statistics, Hyattsville, MD, 1989. Note: In the NHIS, conditions are determined in two ways. First, participants receive one of six condition lists that ask them if they have a specific condition (see Table A-1). Second, participants are asked broad questions to reveal general health and functioning (the questions in the top panel of this table). If participants reveal they have health or functioning difficulties, they are then asked what conditions cause these difficulties (for example, the questions in the second panel of this table). This method misses those with conditions who have no such difficulties, while the first method captures those with conditions who have no health or functioning difficulties. So only one-sixth of the sample is directly asked about blindness. This one-sixth of the sample is a random sample, because being asked about blindness is not dependent on one’s response to another question. The remaining five-sixths of the sample is choice-based, because revealing blindness is dependent on one’s response (choice) to another question. |
CONCEPTUAL FRAMEWORK AND DEFINITIONS
Nagi (1965) developed a framework for defining disability, in which diseases/disorders result in the impairment of required functions that then interact with the socioeconomic and physical environment and lead to disability. Using the Nagi framework, this paper distinguishes those with chronic vision-related diseases/disorders, such as cataracts, from those with chronic visual impairments, such as being blind in one eye. Those with cataracts are not necessarily visually impaired.7 This distinction is important in the context of economic experience because impaired function, rather than a specific disease/disorder, is expected to have a greater influence on employment and program participation.
The central focus of this paper is the economic experience of those who are blind in both eyes, for they are the group among people with vision-related conditions at the greatest risk of economic difficulties. They are also most likely to be eligible for Social Security Disability Insurance (SSDI) payments, Supplemental Security Income (SSI) benefits, and other government programs based on their medical conditions. The economic experience of those who are blind in both eyes is compared with that of those with other visual impairments. Economic statistics are also provided for those with vision-related diseases/disorders, which include glaucoma, cataracts, color blindness, and an “other” category, which consolidates conjunctivitis, disorders of the lacrimal system, disorders of binocular eye movements, and diseases of the retina.8
For comparison purposes the economic experiences of those with some other functional impairments are provided in this report. These functional impairment groups are also seriously at risk of low rates of employment and diminished economic well-being. These categories are deafness in both ears, other hearing impairment, mental retardation, paraplegia, hemiplegia, quadriplegia, and cerebral palsy.
Defining Chronic Conditions
The NHIS provides extensive information on chronic conditions. The term “condition” refers to diseases/disorders and impairments. Chronic conditions are conditions that exist for three or more months, although some conditions are considered chronic regardless of duration.
As mentioned above, the NHIS captures condition information in two ways: (1) checklists of specific conditions and (2) screening questions followed by open-ended probing questions. The next step is to consolidate and classify survey responses from all points in the survey into a set of condition categories based on the International Classification of Diseases. NCHS hires special medical coders to perform this complex task.
According to the NHIS Medical Coding Manual, participants are classified as being blind in both eyes if they describe their condition as blind, no vision, or can’t see. If there is no clear indication that only one eye is involved, it is assumed that both eyes are involved.
The NHIS also provides a category entitled “other visual impairments.” Participants who are blind in one eye are in this category. This category also includes those who describe their eyesight, seeing, sight, or vision as being bad, blurred, defective, limited, poor, double, problem with, trouble with, or who use phrases like partially blind, blind spots, half-blind. Double-vision, color blindness, night blindness, and day blindness are combined into a single “other visual impairments” group. In addition, any active vision-related diseases/ disorders reported by participants are also classified, regardless of whether they cause visual impairment. The NHIS provides the following categories of diseases/disorders: glaucoma, cataracts, color blindness, conjunctivitis, disorders of the lacrimal system, disorders of binocular eye movements, diseases of the retina, and others vision-related of eye and adnexa.
The NHIS defines other chronic impairments in a similar manner to visual impairments. Participants who are reported as being deaf in both ears, having no useful hearing in both ears, or can’t hear in both ears are classified as deaf in both ears. Those reported being partially deaf in both ears or a little deaf in both ears are coded as other hearing impairment. If only one ear is involved, a code of “other hearing impairment” is given. If the medical coder is unable to determine whether one or both ears are involved, the individual is coded as other hearing impairment. Hearing problems relating to allergies or earwax are not classified.
Mental retardation includes mental deficiency or retardation, and those describing themselves as can’t learn, slow learner. Mental retardation is considered chronic regardless of onset.
The NHIS codes paralysis as partial or complete and for various parts or portions of the body. Paraplegia is complete paralysis of the lower body, both legs, or from the waist down. Hemiplegia is complete paralysis of one side of the body, including limbs. Quadriplegia is complete paralysis of the entire body or four limbs. Paralysis must
exist for three or more months to be considered chronic. Cerebral palsy (and its synonyms) is chronic regardless of date of onset and includes those who describe themselves as congenitally “spastic.”
Definitions of Economic Variables
Economic experience is captured via employment rates, mean household size-adjusted income, and receipt of SSDI and SSI payments. The employment rate is based on the following NHIS questions. “During [the past two calendar weeks], did [you] work at any time at a job or business not counting work around the house? (Include unpaid work in family [farm/business].)” Persons not working were asked, “[e]ven though [you] did not work during those 2 weeks, did [you] have a job or business?” Persons who answer “yes” to the first question or “yes” to the second question are considered employed.
Household income is the sum of all income in the household. Households can contain more than one family. The NHIS uses the following questions to determine family income: “Was the total FAMILY income during the past 12 months—that is, yours, [and other family members] more or less than $20,000? Include money from jobs, social security, retirement income, unemployment payments, public assistance, and so forth. Also include income from interest, dividends, net income from businesses, farm, or rent and any other money income received.” And then, “[of the income brackets provided] which [bracket] best represents the total combined FAMILY income during the past 12 months—that is, yours, [and other family members]? Include wages, salaries, and other items we just talked about.” The respondents can choose from 26 income brackets. To obtain a dollar value for family income, family income is assigned the midpoint of the chosen income bracket. Respondents choosing the top bracket ($50,000 and above) are assigned the mean annual family income among those families above $50,000 as estimated from the Current Population Survey.
Household income is adjusted for household size to get a better measure of an individual’s access to household resources. This paper
follows the common practice of dividing household income by the square root of household size. This accounts for the fact that $500 per week provides a higher standard of living for a single-person household than it does for individuals belonging to larger households.9 Because we are comparing income across years, we adjust income using the consumer price index-urban (CPI-U); all income values are in 1998 dollars.
Receipt of SSDI and SSI payments is determined with relatively straightforward questions and refers to receipt of payments in the month prior to the survey. SSDI and SSI recipiency information is available only for 1990-1992, 1994, and 1995.
This paper focuses on working-age men and women (ages 25 to 61). Using this age range avoids confusing reductions in work or economic well-being associated with disability with reductions or declines associated with retirement at older ages or initial transitions in and out of the labor force related to job shopping at younger ages. Men and women are evaluated separately.
RESULTS
To get an idea of the size of populations with the various chronic conditions used in this study, Annex Table A-1 shows the prevalence rates of these chronic conditions in the working-age population in the United States, by gender and the random and choice samples. Annex Tables A-2a through A-2d show the sample sizes used to generate the economic statistics reported below.
Tables A-3 through A-7 compare differences across subgroups. These tables contain employment rates (Table A-3), mean household size-adjusted incomes (Table A-4), the percentages receiving SSDI payments
(Table A-5), SSI benefits (Table A-6), and either SSDI or SSI payments or both (Table A-7). In these tables, the estimates of those who are blind in both eyes are compared with the estimates of the other groups. Asterisks indicate when the difference between those who are blind in both eyes and another group is statistically significant. The remaining tables illustrate changes over time in employment rates (Table A-8) and mean household size-adjusted income (Table A-9) for the choice-based sample. Sample sizes in the random sample are insufficient to accurately measure changes over time. In these tables, asterisks indicate when the difference between 1983-1987 and 1992-1996 is statistically significant.
Table A-10 compares the employment rates for those who are blind in both eyes and visually impaired with the findings of Trupin et al. (1997) and Kirchner et al. (1999).
Prevalence
As expected, the prevalence rates reported in Annex Table A-1 are higher in the random sample than in the choice-based sample for all of the chronic conditions used in this study. This suggests that there are some individuals with chronic conditions who do not have any of the health issues listed in the top panel of Table A-2. This may also suggest that when asked directly about a specific condition, survey participants are more likely to report these conditions.
The relative prevalence rates (Annex Table A-1, columns 3 and 6) reveal that the difference between the random and choice-based samples is least among those with paraplegia/hemiplegia/quadriplegia, mental retardation, cerebral palsy, and those who are blind in both eyes. This suggests individuals with these chronic conditions are more affected by the health issues listed in the top panel of Table A-2. The difference between the random and choice-based samples is dramatically higher among those who report hearing impairments and other visual impairments, diseases, and disorders.
Employment
The discussion of employment, income, and program participation focuses mainly on the random sample. Unless specified otherwise, all results discussed below refer to the random samples.
The first column of Table A-3 shows that 88.8 percent of all working-age men without any visual impairment were employed in the period 1983-1996. Over the same period, the employment rate among men who are blind in both eyes was 49.4 percent. The second column of Table A-3 shows that the relative employment rate of men without visual impairments was 1.80, which means that men without visual impairments were 1.80 times as likely to be employed as men who are blind in both eyes.10 Men with other visual impairments were employed at a rate of 82.3 percent—1.64 times as likely to be employed as men who are blind in both eyes. Similarly, the employment rate of men with vision-related diseases or disorders was 85.0 percent.
Among the other chronic impairments, only men with paraplegia, hemiplegia, or quadriplegia and men with mental retardation were employed at a lower rate than men who are blind in both eyes. Men with paraplegia, hemiplegia, or quadriplegia were about half (0.45) as likely to be employed as men who are blind in both eyes. Similarly, men with mental retardation were about 0.70 times as likely to be employed than men who are blind in both eyes. The employment rate of men with cerebral palsy (58.3 percent) is similar to that of men who are blind in both eyes, and the difference is statistically insignificant. This suggests that there may be no difference or that the sample sizes are insufficient to identify a difference. Men who are deaf in both ears were about one and a half (1.53) times as likely to be employed as men who are blind in both eyes.
In general, the employment rates of working-age women are lower than those of working-age men. However, when comparing across chronic conditions, the employment patterns of women are similar.
TABLE A-3 Employment Rates of Noninstitutionalized Working-Age Civilians (Ages 25 to 61) with Various Chronic Impairments, Diseases, and Disorders, Pooled Over 1983 Through 1996, by Sample and Gender
|
Random Samplea |
|
|
Men |
|
Group |
Employment Rate |
Relative Rateb |
No visual impairments |
88.8 |
1.80*** |
Visual impairments |
81.2 |
1.64*** |
-Blind in both eyes |
49.4 |
1.00 |
-Other visual impairments |
82.3 |
1.67*** |
Vision-related diseases/disorders |
85.0 |
1.72*** |
-Glaucoma |
67.6 |
1.37*** |
-Cataracts |
67.4 |
1.36*** |
-Color blindness |
91.4 |
1.85*** |
-Other vision-related diseases/disordersc |
85.1 |
1.72*** |
Other impairments |
||
-Hearing impairments |
81.6 |
1.65*** |
—Deaf in both ears |
75.4 |
1.53*** |
—Other hearing impairments |
81.9 |
1.66*** |
-Mental retardation |
34.6 |
0.70** |
-Paraplegia, hemiplegia, or quadriplegia |
22.3 |
0.45*** |
-Cerebral palsy |
58.3 |
1.18 |
Note: Asterisks signify when the difference between blind in both eyes and another group is statistically significant at the 99 percent (***), 95 percent (**), and 90 percent (*) levels. NA refers to groups for which sample size is insufficient. a In the NHIS, conditions are determined in two ways. First, participants receive one of six condition lists that ask them if they have a specific condition (see Table A-1). Second, participants are asked broad questions to reveal general health and functioning (see Table A-2, top panel). If participants reveal they have health or functioning difficulties, they are then asked what conditions cause these difficulties (see Table A-2, bottom panel). This method misses those with conditions who have no such difficulties, while the first method captures those with conditions who have no health or functioning difficulties. So only one-sixth of the sample is directly asked |
|
Choice-Based Samplea |
||||
Women |
Men |
Women |
|||
Employment Rate |
Relative Rateb |
Employment Rate |
Relative Rateb |
Employment Rate |
Relative Rateb |
69.2 |
2.30*** |
88.5 |
2.73*** |
68.8 |
2.97*** |
52.9 |
1.76*** |
52.0 |
1.60*** |
36.9 |
1.60*** |
30.0 |
1.00 |
32.4 |
1.00 |
23.1 |
1.00 |
54.7 |
1.82*** |
56.4 |
1.74*** |
40.5 |
1.75*** |
51.3 |
1.71*** |
63.7 |
1.96*** |
46.5 |
2.01*** |
45.5 |
1.51*** |
53.9 |
1.66*** |
40.0 |
1.73*** |
46.0 |
1.53*** |
56.2 |
1.73*** |
35.8 |
1.55*** |
64.7 |
2.15*** |
86.7 |
2.67*** |
NA |
NA |
56.4 |
1.88*** |
69.7 |
2.15*** |
55.5 |
2.40*** |
|
|||||
58.4 |
1.94*** |
77.0 |
2.37*** |
53.5 |
2.31*** |
50.3 |
1.68*** |
64.1 |
1.98*** |
44.6 |
1.93*** |
58.7 |
1.96*** |
78.0 |
2.40*** |
54.5 |
2.36*** |
29.1 |
0.97 |
33.0 |
1.02 |
27.9 |
1.21 |
17.7 |
0.59 |
20.4 |
0.63*** |
19.4 |
0.84 |
27.8 |
0.93 |
43.7 |
1.35** |
32.3 |
1.39* |
about blindness. This one-sixth of the sample is a random sample, because being asked about blindness is not dependent on one’s response to another question. The remaining five-sixths of the sample is choice-based, because revealing blindness is dependent on one’s response (choice) to another question. b The relative employment rate is the employment rate of a given group divided by the employment rate of those who are blind in both eyes. c The category other includes conjunctivitis, disorders of the lacrimal system, disorders of binocular eye movements, and diseases of the retina. Source: Author’s calculations using the National Health Interview Survey, 1983-1996. |
TABLE A-4 Mean Household Size-Adjusted Income (HHSAI) of Noninstitutionalized Working-Age Civilians (Ages 25 to 61) with Various Chronic Impairments, Diseases, and Disorders, Pooled Over 1983 Through 1996, by Sample and Gender
|
Random Samplea |
|
|
Men |
|
Group |
Mean HHSAI |
Relative HHSAIb |
No visual impairments |
31,067 |
1.22*** |
Visual impairments |
29,361 |
1.15* |
-Blind in both eyes |
25,503 |
1.00 |
-Other visual impairments |
29,504 |
1.16* |
Vision-related diseases/disorders |
31,655 |
1.24*** |
-Glaucoma |
28,978 |
1.14 |
-Cataracts |
26,859 |
1.05 |
-Color blindness |
32,991 |
1.29*** |
-Other vision-related diseases/disordersc |
33,155 |
1.30*** |
Other impairments |
||
-Hearing impairments |
30,954 |
1.21** |
—Deaf in both ears |
28,702 |
1.13 |
—Other hearing impairments |
31,070 |
1.22*** |
-Mental retardation |
17,382 |
0.68*** |
-Paraplegia, hemiplegia, or quadriplegia |
20,067 |
0.79* |
-Cerebral palsy |
23,614 |
0.93 |
Note: Asterisks signify when the difference between blind in both eyes and another group is statistically significant at the 99 percent (***), 95 percent (**), and 90 percent (*) levels. NA refers to groups for which sample size is insufficient. a In the NHIS, conditions are determined in two ways. First, participants receive one of six condition lists that ask them if they have a specific condition (see Table A-1). Second, participants are asked broad questions to reveal general health and functioning (see Table A-2, top panel). If participants reveal they have health or functioning difficulties, they are then asked what conditions cause these difficulties (see Table A-2, bottom panel). This method misses those with conditions who have no such difficulties, while the first method captures those with conditions who have no health or functioning difficulties. So only one-sixth of the sample is directly asked about blindness. |
|
Choice-Based Samplea |
||||
Women |
Men |
Women |
|||
Mean HHSAI |
Relative HHSAIb |
Mean HHSAI |
Relative HHSAIb |
Mean HHSAI |
Relative HHSAIb |
28,578 |
1.37*** |
31,110 |
1.70*** |
28,696 |
1.47*** |
22,821 |
1.10 |
21,631 |
1.18*** |
20,624 |
1.05 |
20,837 |
1.00 |
18,348 |
1.00 |
19,567 |
1.00 |
22,975 |
1.10 |
22,363 |
1.22*** |
20,901 |
1.07 |
26,516 |
1.27*** |
26,627 |
1.45*** |
25,149 |
1.29*** |
26,246 |
1.26** |
25,385 |
1.38*** |
22,019 |
1.13 |
24,072 |
1.16 |
23,310 |
1.27*** |
22,786 |
1.16* |
27,692 |
1.33** |
33,194 |
1.81*** |
NA |
NA |
29,422 |
1.41*** |
28,277 |
1.54*** |
28,185 |
1.44*** |
|
|||||
25,641 |
1.23** |
29,434 |
1.60*** |
24,966 |
1.28*** |
23,088 |
1.11 |
26,098 |
1.42*** |
19,888 |
1.02 |
25,758 |
1.24** |
29,690 |
1.62*** |
25,534 |
1.30*** |
18,049 |
0.87 |
16,147 |
0.88*** |
16,927 |
0.87** |
23,304 |
1.12 |
20,245 |
1.10 |
20,362 |
1.04 |
17,677 |
0.85 |
18,438 |
1.00 |
19,147 |
0.98 |
This one-sixth of the sample is a random sample, because being asked about blindness is not dependent on one’s response to another question. The remaining five-sixths of the sample is choice-based, because revealing blindness is dependent on one’s response (choice) to another question. b The relative HHSAI is the mean HHSAI of a given group divided by mean HHSAI of those who are blind in both eyes. c The category other includes conjunctivitis, disorders of the lacrimal system, disorders of binocular eye movements, and diseases of the retina. Source: Author’s calculations using the National Health Interview Survey, 1983-1996. |
TABLE A-5 Percentage Receiving Social Security Disability Insurance (SSDI) Payments among Noninstitutionalized Working-Age Civilians (Ages 25 to 61) with Various Chronic Impairments, Diseases, and Disorders, Pooled Over 1983 Through 1996, by Sample and Gender
|
Random Samplea |
|
|
Men |
|
Group |
Percentage Receiving SSDI |
Relative Recipiencyb |
No visual impairments |
1.87 |
0.08*** |
Visual impairments |
6.61 |
0.28** |
-Blind in both eyes |
23.76 |
1.00 |
-Other visual impairments |
5.79 |
0.24** |
Vision-related diseases/disorders |
4.32 |
0.18** |
-Glaucoma |
11.64 |
0.49 |
-Cataracts |
9.72 |
0.41* |
-Color blindness |
1.97 |
0.08*** |
-Other vision-related diseases/disordersc |
8.84 |
0.37* |
Other impairments |
||
-Hearing impairments |
4.29 |
0.18** |
—Deaf in both ears |
9.80 |
0.41 |
—Other hearing impairments |
4.02 |
0.17** |
-Mental retardation |
37.25 |
1.57 |
-Paraplegia, hemiplegia, or quadriplegia |
55.05 |
2.32** |
-Cerebral palsy |
27.45 |
1.16 |
Note: Asterisks signify when the difference between blind in both eyes and another group is statistically significant at the 99 percent (***), 95 percent (**), and 90 percent (*) levels. NA refers to groups where sample size is insufficient. a In the NHIS, conditions are determined in two ways. First, participants receive one of six condition lists that ask them if they have a specific condition (see Table A-1). Second, participants are asked broad questions to reveal general health and functioning (see Table A-2, top panel). If participants reveal they have health or functioning difficulties, they are then asked what conditions cause these difficulties (see Table A-2, bottom panel). This method misses those with conditions who have no such difficulties, while the first method captures those with conditions that have no health or functioning difficulties. So only one-sixth of the sample is directly asked |
|
Choice-Based Samplea |
||||
Women |
Men |
Women |
|||
Percentage Receiving SSDI |
Relative Recipiencyb |
Percentage Receiving SSDI |
Relative Recipiencyb |
Percentage Receiving SSDI |
Relative Recipiencyb |
1.24 |
0.03*** |
2.37 |
0.05*** |
1.35 |
0.04*** |
8.43 |
0.21*** |
23.39 |
0.49*** |
14.75 |
0.38*** |
39.31 |
1.00 |
47.43 |
1.00 |
38.66 |
1.00 |
6.24 |
0.16*** |
16.86 |
0.36*** |
8.55 |
0.22*** |
6.12 |
0.16*** |
16.68 |
0.35*** |
9.79 |
0.25*** |
5.59 |
0.14*** |
22.81 |
0.48*** |
10.77 |
0.28*** |
6.08 |
0.15*** |
12.43 |
0.26*** |
11.24 |
0.29*** |
2.47 |
0.06*** |
NA |
NA |
NA |
NA |
11.90 |
0.30** |
22.16 |
0.47*** |
9.37 |
0.24*** |
|
|||||
3.98 |
0.10*** |
5.87 |
0.12*** |
3.68 |
0.10*** |
23.50 |
0.60 |
15.55 |
0.33*** |
5.96 |
0.15*** |
3.36 |
0.09** |
5.45 |
0.11*** |
3.55 |
0.09*** |
29.52 |
0.75 |
49.31 |
1.04 |
40.82 |
1.06 |
28.61 |
0.73 |
69.07 |
1.46*** |
33.37 |
0.86 |
11.37 |
0.29* |
40.52 |
0.85 |
48.59 |
1.26 |
about blindness. This one-sixth of the sample is a random sample because being asked about blindness is not dependent one’s response to another question. The remaining five-sixths of the sample is choice-based because revealing blindness is dependent one’s response (choice) to another question. b The relative recipiency is the percentage of a given group divided by the recipiency of those who are blind in both eyes. c The category other includes conjunctivitis, disorders of the lacrimal system, disorders of binocular eye movements, and diseases of the retina. Source: Author’s calculations using the National Health Interview Survey, 1983-1996. |
TABLE A-6 Percentage Receiving Supplemental Security Income (SSI) among Noninstitutionalized Working-Age Civilians (Ages 25 to 61) with Various Chronic Impairments, Diseases, and Disorders, Pooled Over 1990-1992, 1994, and 1995, by Sample and Gender
|
Random Samplea |
||||
|
Men |
||||
Group |
Percentage Receiving SSI |
Relative Recipiencyb |
|||
No visual impairments |
0.99 |
0.04*** |
|||
Visual impairments |
3.97 |
0.16*** |
|||
-Blind in both eyes |
24.69 |
1.00 |
|||
-Other visual impairments |
2.97 |
0.12*** |
|||
Vision-related diseases/disorders |
2.28 |
0.09*** |
|||
-Glaucoma |
9.40 |
0.38* |
|||
-Cataracts |
4.97 |
0.20*** |
|||
-Color blindness |
0.36 |
0.01*** |
|||
-Other vision-related diseases/disordersc |
5.91 |
0.24** |
|||
Other impairments |
|||||
-Hearing impairments |
1.86 |
0.08*** |
|||
—Deaf in both ears |
4.41 |
0.18*** |
|||
—Other hearing impairments |
1.73 |
0.07*** |
|||
-Mental retardation |
42.73 |
1.73* |
|||
-Paraplegia, hemiplegia, or quadriplegia |
34.64 |
1.40 |
|||
-Cerebral palsy |
31.07 |
1.26 |
|||
Note: Asterisks signify when the difference between blind in both eyes and another group is statistically significant at the 99 percent (***), 95 percent (**), and 90 percent (*) levels. NA refers to groups where sample size is insufficient. a In the NHIS, conditions are determined in two ways. First, participants receive one of six condition lists that ask them if they have a specific condition (see Table A-1). Second, participants are asked broad questions to reveal general health and functioning (see Table A-2, top panel). If participants reveal they have health or functioning difficulties, they are then asked what conditions cause these difficulties (see Table A-2, bottom panel). This method misses those with conditions who have no such difficulties, while the first method captures those with conditions that have no health or functioning difficulties. So only one-sixth of the sample is directly asked |
|
Choice-Based Samplea |
||||
Women |
Men |
Women |
|||
Percentage Receiving SSI |
Relative Recipiencyb |
Percentage Receiving SSI |
Relative Recipiencyb |
Percentage Receiving SSI |
Relative Recipiencyb |
1.46 |
0.06*** |
1.08 |
0.04*** |
1.39 |
0.04*** |
10.38 |
0.42 |
9.97 |
0.36*** |
15.76 |
0.46*** |
24.74 |
1.00 |
27.43 |
1.00 |
34.57 |
1.00 |
9.36 |
0.38* |
5.23 |
0.19*** |
10.88 |
0.31*** |
8.44 |
0.34* |
7.45 |
0.27*** |
9.46 |
0.27*** |
8.89 |
0.36* |
9.22 |
0.34*** |
12.76 |
0.37*** |
10.59 |
0.43 |
10.74 |
0.39** |
10.82 |
0.31*** |
1.23 |
0.05*** |
NA |
NA |
NA |
NA |
12.57 |
0.51 |
6.03 |
0.22*** |
7.50 |
0.22*** |
|
|||||
3.35 |
0.14** |
2.34 |
0.09*** |
5.93 |
0.17*** |
9.19 |
0.37 |
4.12 |
0.15*** |
12.84 |
0.37*** |
3.16 |
0.13*** |
2.26 |
0.08*** |
5.55 |
0.16*** |
52.36 |
2.12*** |
45.18 |
1.65*** |
40.71 |
1.18 |
24.03 |
0.97 |
26.51 |
0.97 |
17.28 |
0.50** |
40.26 |
1.63 |
34.00 |
1.24 |
46.27 |
1.34 |
about blindness. This one-sixth of the sample is a random sample because being asked about blindness is not dependent one’s response to another question. The remaining five-sixths of the sample is choice-based because revealing blindness is dependent one’s response (choice) to another question. b The relative recipiency is the percentage of a given group divided by the recipiency of those who are blind in both eyes. c The category other includes conjunctivitis, disorders of the lacrimal system, disorders of binocular eye movements, and diseases of the retina. Source: Author’s calculations using the National Health Interview Survey, 1983-1996. |
TABLE A-7 Percentage Receiving Supplemental Security Income (SSI) and/or Social Security Disability Insurance (SSDI) Payments among Noninstitutionalized Working-Age Civilians (Ages 25 to 61) with Various Chronic Impairments, Diseases, and Disorders, Pooled Over 1990-1992, 1994, and 1995, by Sample and Gender
|
Random Samplea |
|
|
Men |
|
Group |
Percentage Receiving SSDI, SSI |
Relative Recipiencyb |
No visual impairments |
2.62 |
0.07*** |
Visual impairments |
9.08 |
0.24*** |
-Blind in both eyes |
37.38 |
1.00 |
-Other visual impairments |
7.72 |
0.21*** |
Vision-related diseases/disorders |
5.88 |
0.16*** |
-Glaucoma |
15.56 |
0.42** |
-Cataracts |
13.55 |
0.36*** |
-Color blindness |
2.33 |
0.06*** |
-Other vision-related diseases/disordersc |
14.75 |
0.39** |
Other impairments |
||
-Hearing impairments |
5.97 |
0.16*** |
—Deaf in both ears |
14.21 |
0.38** |
—Other hearing impairments |
5.57 |
0.15*** |
-Mental retardation |
63.56 |
1.70** |
-Paraplegia, hemiplegia, or quadriplegia |
79.19 |
2.12*** |
-Cerebral palsy |
51.34 |
1.37 |
Note: Asterisks signify when the difference between blind in both eyes and another group is statistically significant at the 99 percent (***), 95 percent (**), and 90 percent (*) levels. NA refers to groups where sample size is insufficient. a In the NHIS, conditions are determined in two ways. First, participants receive one of six condition lists that ask them if they have a specific condition (see Table A-1). Second, participants are asked broad questions to reveal general health and functioning (see Table A-2, top panel). If participants reveal they have health or functioning difficulties, they are then asked what conditions cause these difficulties (see Table A-2, bottom panel). This method misses those with conditions who have no such difficulties, while the first method captures those with conditions that have no health or functioning difficulties. So only one-sixth of the sample is directly asked |
|
Choice-Based Samplea |
||||
Women |
Men |
Women |
|||
Percentage Receiving SSDI, SSI |
Relative Recipiencyb |
Percentage Receiving SSDI, SSI |
Relative Recipiencyb |
Percentage Receiving SSDI, SSI |
Relative Recipiencyb |
2.47 |
0.05*** |
3.08 |
0.05*** |
2.44 |
0.04*** |
15.74 |
0.33*** |
28.49 |
0.48*** |
27.27 |
0.45*** |
47.87 |
1.00 |
59.75 |
1.00 |
60.87 |
1.00 |
13.46 |
0.28*** |
20.00 |
0.33*** |
18.56 |
0.30*** |
12.14 |
0.25*** |
20.18 |
0.34*** |
16.14 |
0.27*** |
13.92 |
0.29*** |
28.04 |
0.47*** |
19.46 |
0.32*** |
13.41 |
0.28*** |
18.45 |
0.31*** |
17.35 |
0.29*** |
3.69 |
0.08*** |
NA |
NA |
NA |
NA |
19.26 |
0.40** |
22.16 |
0.37*** |
13.52 |
0.22*** |
|
|||||
6.84 |
0.14*** |
7.38 |
0.12*** |
8.84 |
0.15*** |
26.89 |
0.56 |
16.98 |
0.28*** |
15.97 |
0.26*** |
6.20 |
0.13*** |
6.96 |
0.12*** |
8.45 |
0.14*** |
70.29 |
1.47* |
72.75 |
1.22** |
67.72 |
1.11 |
52.64 |
1.10 |
84.37 |
1.41*** |
41.68 |
0.68** |
51.63 |
1.08 |
63.01 |
1.05 |
77.28 |
1.27* |
about blindness. This one-sixth of the sample is a random sample because being asked about blindness is not dependent one’s response to another question. The remaining five-sixths of the sample is choice-based because revealing blindness is dependent one’s response (choice) to another question. b The relative recipiency is the percentage of a given group divided by the recipiency of those who are blind in both eyes. c The category other includes conjunctivitis, disorders of the lacrimal system, disorders of binocular eye movements, and diseases of the retina. Source: Author’s calculations using the National Health Interview Survey, 1983-1996. |
TABLE A-8 Employment Rates of Noninstitutionalized Working-Age Civilians (Ages 25 to 61) in the Choice-Based Sample with Various Chronic Impairments, Diseases, and Disorders, Pooled Over 1983-1987 and 1992-1996 and Percentage Changes, by Gender
|
Group |
No visual impairments |
Visual impairments |
-Blind in both eyes |
-Other visual impairments |
Vision-related diseases/disorders |
-Glaucoma |
-Cataracts |
-Color blindness |
-Other vision-related diseases/disordersb |
Other impairments |
-Hearing impairments |
—Deaf in both ears |
—Other hearing impairments |
-Mental retardation |
-Paraplegia, hemiplegia, or quadriplegia |
-Cerebral palsy |
Note: Asterisks signify when the difference between 1983-1987 and 1992-1996 is statistically significant at the 99 percent (***), 95 percent (**), and 90 percent (*) levels. NA refers to groups where sample size is insufficient. In the NHIS, conditions are determined in two ways. First, participants receive one of six condition lists that ask them if they have a specific condition (see Table A-1). Second, participants are asked broad questions to reveal general health and functioning (see Table A-2, top panel). If participants reveal they have health or functioning difficulties, they are then asked what conditions cause these difficulties (see Table A-2, bottom panel). This method misses those with conditions who have no such difficulties, while the first method captures those with conditions that have no health or functioning difficulties. So only one-sixth of the sample is directly asked |
Men |
Women |
||||
1983-87 |
1992-96 |
Percentage Changea |
1983-87 |
1992-96 |
Percentage Changea |
88.6 |
87.9 |
−0.79*** |
65.2 |
71.1 |
8.66*** |
54.5 |
48.5 |
−11.65* |
34.5 |
39.9 |
14.52 |
34.8 |
27.3 |
−24.15 |
19.5 |
27.8 |
35.10 |
58.7 |
53.2 |
−9.83 |
38.1 |
43.6 |
13.46 |
62.7 |
63.3 |
0.95 |
41.1 |
48.8 |
17.13 |
50.6 |
52.0 |
2.73 |
33.3 |
40.5 |
19.51 |
53.3 |
59.4 |
10.83 |
33.3 |
34.2 |
2.67 |
92.0 |
77.2 |
−17.49 |
NA |
NA |
NA |
71.5 |
69.6 |
−2.69 |
50.1 |
59.8 |
17.65 |
|
|||||
65.9 |
65.2 |
−1.07 |
43.6 |
48.2 |
10.02 |
61.1 |
58.8 |
−3.84 |
37.6 |
44.4 |
16.59 |
66.8 |
66.1 |
−1.05 |
45.0 |
49.1 |
8.71 |
32.9 |
34.3 |
4.17 |
23.6 |
31.5 |
28.68** |
19.1 |
22.7 |
17.22 |
7.2 |
29.5 |
121.53*** |
44.2 |
33.8 |
−26.67 |
35.6 |
27.7 |
−24.96 |
about blindness. This one-sixth of the sample is a random sample because being asked about blindness is not dependent one’s response to another question. The remaining five-sixths of the sample is choice-based because revealing blindness is dependent one’s response (choice) to another question. a The percentage change is the difference between the two periods divided by the average of the two periods multiplied by 100. b The category other includes conjunctivitis, disorders of the lacrimal system, disorders of binocular eye movements, and diseases of the retina. Source: Author’s calculations using the National Health Interview Survey, 1983-1996. |
TABLE A-9 Mean Household Size-Adjusted Income of Noninstitutionalized Working-Age Civilians (Ages 25 to 61) in the Choice-Based Sample with Various Chronic Impairments, Diseases, and Disorders, Pooled Over 1983-1987 and 1992-1996 and Percentage Changes, by Gender
|
||
Group |
||
No visual impairments |
||
Visual impairments |
||
-Blind in both eyes |
||
-Other visual impairments |
||
Vision-related diseases/disorders |
||
-Glaucoma |
||
-Cataracts |
||
-Color blindness |
||
-Other vision-related diseases/disordersb |
||
Other impairments |
||
-Hearing impairments |
||
—Deaf in both ears |
||
—Other hearing impairments |
||
-Mental retardation |
||
-Paraplegia, hemiplegia, or quadriplegia |
||
-Cerebral palsy |
||
Note: Asterisks signify when the difference between 1983-87 and 1992-96 is statistically significant at the 99 percent (***), 95 percent (**), and 90 percent (*) levels. NA refers to groups where sample size is insufficient. All dollar values are adjusted for inflation to 1998 dollar values. In the NHIS, conditions are determined in two ways. First, participants receive one of six condition lists that ask them if they have a specific condition (see Table A-1). Second, participants are asked broad questions to reveal general health and functioning (see Table A-2, top panel). If participants reveal they have health or functioning difficulties, they are then asked what conditions cause these difficulties (see Table A-2, bottom panel). This method misses those with conditions who have no such difficulties, while the first method captures those with conditions that have no health or functioning difficulties. So only one-sixth of the sample is directly asked about blindness. This one- |
Men |
Women |
||||
1983-87 |
1992-96 |
Percentage Changea |
1983-87 |
1992-96 |
Percentage Changea |
29,574 |
31,885 |
7.52*** |
27,082 |
29,608 |
8.91*** |
21,110 |
21,349 |
1.13 |
19,339 |
21,152 |
8.96 |
16,969 |
17,741 |
4.45 |
20,925 |
18,613 |
−11.70 |
21,989 |
22,149 |
0.72 |
18,952 |
21,947 |
14.65** |
24,270 |
28,104 |
14.64** |
24,321 |
24,432 |
0.46 |
21,423 |
27,657 |
25.40** |
21,143 |
20,775 |
−1.76 |
23,668 |
24,375 |
2.94 |
20,663 |
24,464 |
16.85 |
34,189 |
32,155 |
−6.13 |
NA |
NA |
NA |
24,484 |
29,204 |
17.58* |
28,234 |
27,058 |
−4.25 |
24,027 |
26,532 |
9.91* |
22,321 |
22,580 |
1.15 |
26,077 |
22,847 |
−13.20 |
19,832 |
18,181 |
−8.69 |
23,682 |
27,102 |
13.47** |
22,911 |
23,597 |
2.95 |
15,593 |
15,234 |
−2.33 |
15,451 |
17,102 |
10.14 |
17,702 |
22,264 |
22.83** |
17,844 |
21,849 |
20.18 |
19,116 |
17,698 |
−7.70 |
20,477 |
16,655 |
−20.59 |
sixth of the sample is a random sample because being asked about blindness is not dependent one’s response to another question. The remaining five-sixths of the sample is choice-based because revealing blindness is dependent one’s response (choice) to another question. a The percentage change is the difference between the two periods divided by the average of the two periods multiplied by 100. b The category other includes conjunctivitis, disorders of the lacrimal system, disorders of binocular eye movements, and diseases of the retina. Source: Author’s calculations using the National Health Interview Survey, 1983-1996. |
TABLE A-10 Comparison with Results from Other Study of Vision-Related Conditions and Labor Market Attachment
|
Study of Vision-Related Conditions and Labor Market Attachment |
|||
Descriptions |
This Study |
|||
Dataset |
National Health Interview Survey |
|||
Time period |
Pooled over 1983-96 |
|||
Age group |
Aged 25 to 61 |
|||
Measure of labor market attachmentb |
Have a job in the previous two weeks |
|||
Secondary restrictions |
Random samplec |
Choice-based samplec |
||
Measures of vision-related conditions |
Blind in both eyes |
Other visual impairments |
Blind in both eyes |
Other visual impairments |
Percentage attached to the labor market |
39.5 |
73.1 |
28.2 |
49.8 |
a Trupin et al. (1997) report annual estimates from 1983-1994. The pooled results below are my calculations using their annual estimates. b All three studies are based on the same set of questions. Trupin et al. (1997) include those who are looking for work but mention similar estimates are obtained when excluding those who are looking for work. c In the NHIS, conditions are determined in two ways. First, participants receive one of six condition lists that ask them if they have a specific condition (see Table A-1). Second, participants are asked broad questions to reveal general health and functioning (see Table A-2, top panel). If participants reveal they have health or functioning difficulties, they are then asked what conditions cause these difficulties (see Table A-2, bottom panel). This method misses those with conditions who have no such difficulties, while the first method captures those with conditions that have no health or functioning difficulties. So only one-sixth of the sample is directly asked |
|
||
Trupin et al. (1997) |
Kirchner et al. (1999) |
|
National Health Interview Survey on Disability, Phase I |
National Health Interview Survey |
|
Pooled over 1983-94a |
Pooled over 1994-95 |
|
Aged 18 to 64 |
Aged 18 to 54 |
|
Have a job or looking for work in the previous two weeks |
Have a job in the previous two weeks |
|
Condition is the main cause of work and/or activity limitationsd |
None |
|
Blind in both eyes |
Other visual impairments |
Serious visual impairment even when wearing glassese |
30.1 |
61.0 |
59.0 |
about blindness. This one-sixth of the sample is a random sample because being asked about blindness is not dependent one’s response to another question. The remaining five-sixths of the sample is choice-based because revealing blindness is dependent one’s response (choice) to another question. d Trupin et al. (1997) restrict their sample to only those who report blindness as the main source of work and/or activity limitation, which is a subset of the prompting questions referred to in Table A-2. This allows them to disregard the distinction between recipients and nonrecipients of List #2, yet they ignore those with conditions who are not limited. e In Kirchner et al. (1999) individuals are considered to have a “serious visual impairment” if they have “SERIOUS difficulty seeing even when wearing glasses or contact lenses” and then identify themselves as “legally blind” or expect themselves “to have SERIOUS difficulty seeing, for at least the next 12 months.” |
The third column of Table A-3 shows that 69.2 percent of working-age women without visual impairments were employed in the period 1983-1996. The employment rate among women who are visually impaired is 52.9 percent and the rate for women who are blind in both eyes is 30.0 percent. The fourth column of Table A-3 reveals that women without visual impairments were over twice (2.30) as likely to be employed as women who are blind in both eyes. Women with other visual impairments were employed at a rate of 54.7 percent. The employment rate of women with vision-related diseases or disorders is 51.3 percent.
Among the other chronic conditions, the employment rates of women with cerebral palsy, paraplegia/hemiplegia/quadriplegia, and mental retardation are statistically indistinguishable from the employment rates of women who are blind in both eyes. This suggests that there may be no difference or that the sample sizes are insufficient to identify a difference. All other groups were employed at higher rates than women who are blind in both eyes.
As expected, among all groups, the employment rate of men in the random sample is higher than the employment rate of men in the choice-based sample. This is also true among women, except for women with cerebral palsy and paraplegia/hemiplegia/quadriplegia.11 Comparing the employment rates of the random sample and the choice-based sample, it appears that four groups always had the lowest employment rate: (1) those who are blind in both eyes and those with (2) cerebral palsy, (3) paraplegia/hemiplegia/quadriplegia, and (4) mental retardation. This is consistent with the finding that these groups are more affected by the health issues listed in Table A-2.
Household Size-Adjusted Income
Unless specified otherwise, all results discussed below refer to the random samples. The mean household size-adjusted income among working-age men without visual impairments was $31,067 (Table A-4, column 1) for the period 1983-1996. This figure is 1.22 times higher than the mean household size-adjusted income among men who are blind in both eyes ($25,503). Men with other visual impairments had a mean household size-adjusted income of $29,504, which is 1.16 times the mean household size-adjusted income of men who are blind in both eyes. The mean household size-adjusted income of men with vision-related diseases or disorders was $31,655.
Among the other chronic conditions, mean household size-adjusted incomes lower than that of men who are blind in both eyes are found among men with mental retardation and paraplegia/hemiplegia/ quadriplegia. Men with cerebral palsy and men who are deaf in both ears have similar mean household size-adjusted income; the differences are statistically insignificant.
The cross-condition comparisons are fairly similar for working-age women. The third column of Table A-4 reveals that the mean household size-adjusted income among working-age women without visual impairments was $28,578, compared with $20,837 among working-age women who are blind in both eyes. Mean household size-adjusted income of women with other visual impairments ($22,975) was slightly higher than that of women who are blind in both eyes. Interestingly, the mean household size-adjusted income of women who are deaf in both ears is not statistically different than that of women who are blind in both eyes.
As expected, among all groups, the mean household size-adjusted income of those in the random sample was higher than the mean household size-adjusted income of those in the choice-based sample, except for men with paraplegia/hemiplegia/quadriplegia and women with cerebral palsy.
Program Participation
Unless specified otherwise, all results discussed below refer to the random samples. Table A-5 shows the percentage of men and women receiving SSDI payments and their rates relative to those who are blind in both eyes. (Note that SSDI eligibility must be established by working for a prescribed length of time.) These results are based on data pooled over the years 1990-1992, 1994, and 1995. Few working-age men without visual impairments received SSDI, 1.87 percent (Table A-5, column 1). Among men who are blind in both eyes, only about a quarter (23.76 percent) received SSDI payments. Only 5.79 percent of men with other visual impairments received SSDI payments, which may reflect the severity of these impairments.
Among other groups, only men with paraplegia/hemiplegia/ quadriplegia received SSDI payments at a higher rate (55.05 percent); they were 2.32 times as likely to receive SSDI payments. The difference in SSDI recipiency between men who are blind in both eyes (23.76 percent) and men with mental retardation (37.25) is not statistically significant. Men who are deaf in both ears received SSDI payments at a rate of 9.80 percent and were thus 0.41 times as likely to receive SSDI payments as men who are blind in both eyes, although the difference between the two groups is statistically insignificant.
The third column of Table A-5 shows that 39.31 percent of women who are blind in both eyes received SSDI payments. No other group has a higher recipiency, although the difference is statistically insignificant for women who are deaf in both ears and women with mental retardation or paraplegia/hemiplegia/quadriplegia.
Table A-6 reveals similar patterns for SSI recipiency. Among men who are blind in both eyes, 24.69 percent received SSI. Men with other visual impairments participated at a rate of 3.97 percent, which again may reflect the severity of other visual impairments. Men with mental retardation were 1.73 times as likely to receive SSI benefits as men who are blind in both eyes. Interestingly, men who are deaf in both ears were much less likely (0.18 times) to receive SSI benefits than men who are blind in both eyes. SSI recipiency is generally higher among women than among men (Table A-6, columns 3 and 4).
The patterns across conditions are similar to the patterns shown for men.
Dual eligibility for SSDI and SSI is possible, as long as SSDI payments do not cause an individual to go over the SSI means test. Table A-7 shows that 37.38 percent of men who are blind in both eyes participated in SSDI and/or SSI, which indicates that 11.07 percent participated in both programs.
Consistent with previous findings, the recipiency rates for SSDI and SSI are generally higher for those in the choice-based sample than for those in the random sample. Again, this may reflect the severity of the disability.
Changes Over Time
Due to sample size constraints, the over-time results were generated only for the choice-based sample. Tables A-8 and A-9 illustrate changes in employment rates and mean household size-adjusted income over time for the choice-based sample. Pooled results from 1983-1987 are compared with pooled results from 1992-1996. These years are chosen because they represent similar phases of the business cycle. Both represent recovery periods in the U.S. economy. The NHIS began collecting SSDI and SSI information in 1990; therefore changes in SSDI and SSI participation over time cannot be measured.
Those in the choice-based sample who are blind in both eyes can be thought of as those who are blind in both eyes and report being constrained by their condition. Recall from Table A-5 that when comparing the employment rates of the random sample and the choice-based sample, it appears that four groups always have the lowest employment rate: (1) those who are blind in both eyes and those with (2) cerebral palsy, (3) paraplegia/hemiplegia/quadriplegia, and (4) mental retardation.
The employment rate of working-age men in the choice-based sample without visual impairments was 88.6 percent in the period 1983-1987, and 87.9 percent in the period 1992-1996, which represents a
statistically significant decline of 0.79 percent (Table A-8, columns 1, 2, and 3). The employment rate of men in the choice-based sample with visual impairments declined by much more (11.65 percent), from 54.5 percent in the period 1983-1987 to 48.5 percent in the period 1992-1996. The changes in the employment rates for all other groups of men are not statistically significantly different from zero, which suggests that there may be little change or the sample sizes are insufficient to identify a change.12
The last three columns of Table A-8 show that the employment rates of working-age women in the choice-based sample without visual impairments rose by 8.66 percent, from 65.2 percent in 1983-1987 to 71.1 percent in 1992-1996. However, the change in the employment rates among women in the choice-based sample who are blind in both eyes is not statistically different from zero. Interestingly, the employment rates of women in the choice-based sample increased for those with mental retardation (23.6 to 31.5 percent) and paraplegia/ hemiplegia/quadriplegia (7.2 to 29.5 percent).
The third and sixth columns of Table A-9 show that the mean household size-adjusted income increased between the periods 1983-1987 and 1992-1996 by 7.52 percent among men in the choice-based sample without visual impairments and 8.91 percent among women in the choice-based sample without visual impairments. Changes in the mean household size-adjusted income among men and women in the choice-based sample who are blind in both eyes are not significantly different from zero.
COMPARISON WITH OTHER STUDIES
Table A-10 compares the employment results in this paper to the results of two other studies. The results are remarkably close. Trupin et al. (1997) used the NHIS over the period 1983-1994; thus they should find lower rates of labor force attachment than the ones reported here—1995 and 1996 were growth years in the overall economy. Trupin et al. (1997) report labor force participation rates, which include people who are actively looking for a job; thus their estimates should be higher than the employment rates presented in this paper. They analyzed the population for whom being blind in both eyes is the main cause of activity limitations; thus their sample is similar to the results from the choice-based sample for which main cause of activity limitations is one way of revealing blindness (see Table A-2). They evaluated those ages 18 to 64 and thus should capture lower rates of labor force attachment than those reported here. They combined women and men. Table A-10 adjusts the results of Table A-5 to pull together the results for men and women.
As is shown in the fifth column of Table A-10, Trupin et al. (1997) found a 1983-1994 labor force participation rate of 30.1 percent among those ages 18 to 64 for whom being blind in both eyes is the main cause of activity limitations. This is remarkably similar to the 1983-1996 employment rate of 28.2 percent for those age 25 to 61 in the choice-based sample who are blind in both eyes.
Kirchner et al. (1999) used the National Health Interview on Disability pooled over 1994-1995; thus they should find higher rates of labor force attachment than the ones reported here—they did not cover the recession of the early 1990s. They use exactly the same definition of employment used here. In their study, individuals were considered to have a “serious visual impairment,” if they have “SERIOUS difficulty seeing even when wearing glasses or contact lenses” and then identify themselves as “legally blind” or expect themselves “to have SERIOUS difficulty seeing, for at least the next 12 months.” Their results should fall between the random sample results for those who are blind in both eyes and those with other visual impairments. They evaluated those ages 18 to 54 and thus should also find lower rates of labor force
attachment than those reported here. They also combined women and men.
Kirchner et al. (1999) found a 1994-1995 employment rate of 59.9 percent among those ages 18 to 54 who are severely visually impaired. This is between the 1983-1996 employment rates of 39.5 and 73.1 percent for those ages 25 to 61 in the random sample who are blind in both eyes and who have other visual impairments, respectively.
A Canadian study by Fawcett (1996) estimated labor force participation (employed or actively looking for work) rates for the population with disabilities in Canada. Among working-age persons with “seeing disabilities,” 45.6 percent participated in the labor force in 1991. Similar rates were found among those with disabilities related to mobility (43.3), agility (46.0), speaking (41.8), and mental function/ learning (47.5). The labor force participation of those with hearing disabilities in Canada (62.7 percent) was higher than that of those with seeing disabilities in Canada. These cross-disability patterns are similar to employment patterns shown in Table A-3 for the United States.
CONCLUSION
This paper provides a statistical description of the economic experience of working-age individuals with chronic vision-related conditions over the period 1983-1996 using the NHIS. The economic experience of individuals who are blind in both eyes is compared with the economic experience of those with other chronic conditions. The economic experience of those who are blind in both eyes is worse than those with less severe visual impairments but similar to those with other serious chronic conditions (e.g., paraplegia/hemiplegia/ quadriplegia, mental retardation, and cerebral palsy).
ACKNOWLEDGMENT
The author thanks Richard V. Burkhauser and Corrine Kirchner for their helpful comments.
REFERENCES
Burkhauser, R. V., Daly, M. C., & Houtenville, A. J. (2000). How working age people with disabilities fared over the 1990s business cycle. In P. P. Budetti, R. V. Burkhauser, J. M. Gregory, & H. A. Hunt (eds.), Ensuring health and income security for an aging workforce . Kalamazoo, MI: W.E. UpJohn Institute for Employment Research.
Fawcett, G. M. (1996). Living with disability in Canada: An economic portrait. Hull, Quebec: Human Resources Development Canada, Office for Disability Issues.
Kirchner, C., Schmeidler, E., & Todorov, A. (1999). Looking at employment through a lifespan telescope: Age, health, and employment status of people with serious visual impairment. New York, NY: American Foundation for the Blind (subcontractor to Rehabilitation Research and Training Center on Bindness and Low Vision, Mississippi State University).
Massey, J. T., Moore, T. F., Parsons, V. L., & Tadros, W. (1989). Design and estimation for the National Health Interview Survey, 1985-1994. (Report No. PHS 89-1384). Hyattsville, MD: U.S. Department of Health and Human Services.
Nagi, S. Z. (1965). Some conceptual issues in disability and rehabilitation. In M. B. Sussman (ed.), Sociology and rehabilitation. Washington, DC: American Sociological Association.
Trupin, L., Sebesta, D. S., Yelin, E., & LaPlante, M. P. (1997). Disability statistics report: Trends in labor force participation among persons with disabilities, 1983-1994. (Report No. 10). San Francisco, CA: Disability Statistics Rehabilitation Research and Training Center, Institute for Health and Aging, University of California.
Annex Table A-1 Prevalence Rates of Various Chronic Impairments, Diseases, and Disorders Among Noninstitutionalized Working-Age Civilians (Ages 25 to 61) Pooled Over 1983 Through 1996, by Sample and Gender
|
Group |
Visual impairments |
-Blind in both eyes |
-Other visual impairments |
Vision-related diseases/disorders |
-Glaucoma |
-Cataracts |
-Color blindness |
-Other vision-related diseases/disordersc |
Other impairments |
-Hearing impairments |
—Deaf in both ears |
—Other hearing impairments |
-Mental retardation |
-Paraplegia, hemiplegia, or quadriplegia |
-Cerebral palsy |
Note: Asterisks signify when the difference between the random sample and choice-based sample is statistically significant at the 99 percent (***), 95 percent (**), and 90 percent (*) levels. NA refers to groups where sample size is insufficient. a In the NHIS, conditions are determined in two ways. First, participants receive one of six condition lists that ask them if they have a specific condition (see Table A-1). Second, participants are asked broad questions to reveal general health and functioning (see Table A-2, top panel). If participants reveal they have health or functioning difficulties, they are then asked what conditions cause these difficulties (see Table A-2, bottom panel). This method misses those with conditions who have no such difficulties, while the first method captures those with conditions that have no health or functioning difficulties. So only one-sixth of the sample is directly asked |
Men |
Women |
||||
Random Samplea |
Choice-Based Samplea |
Relative Prevalenceb |
Random Samplea |
Choice-Based Samplea |
Relative Prevalenceb |
4.89 |
0.57 |
0.12*** |
2.38 |
0.39 |
0.17*** |
0.17 |
0.10 |
0.60*** |
0.17 |
0.08 |
0.48*** |
4.71 |
0.47 |
0.10*** |
2.21 |
0.31 |
0.14*** |
4.16 |
0.31 |
0.08*** |
1.97 |
0.36 |
0.18*** |
0.50 |
0.09 |
0.17*** |
0.47 |
0.11 |
0.24*** |
0.62 |
0.10 |
0.16*** |
0.82 |
0.10 |
0.12*** |
2.68 |
0.02 |
0.01*** |
0.27 |
0.00 |
0.00*** |
0.48 |
0.12 |
0.25*** |
0.51 |
0.16 |
0.32*** |
|
|||||
10.75 |
1.14 |
0.11*** |
5.94 |
0.62 |
0.10*** |
0.53 |
0.08 |
0.15*** |
0.26 |
0.06 |
0.24*** |
10.22 |
1.06 |
0.10*** |
5.68 |
0.55 |
0.10*** |
0.46 |
0.35 |
0.76*** |
0.35 |
0.25 |
0.71*** |
0.20 |
0.16 |
0.81* |
0.09 |
0.06 |
0.66** |
0.11 |
0.07 |
0.64** |
0.09 |
0.06 |
0.63** |
about blindness. This one-sixth of the sample is a random sample because being asked about blindness is not dependent one’s response to another question. The remaining five-sixths of the sample is choice-based because revealing blindness is dependent one’s response (choice) to another question. b The relative prevalence is the prevalence in the random sample divided by the prevalence in the choice-based sample. c The category other includes conjunctivitis, disorders of the lacrimal system, disorders of binocular eye movements, and diseases of the retina. Source: Author’s calculations using the National Health Interview Survey, 1983-1996. |
Annex Table A-2a Sample Size of Noninstitutionalized Working-Age Civilian Men (Ages 25 to 61) in the Random Sample with Various Chronic Impairments, Diseases, and Disorders, 1983-1996
|
Year |
|||||
Group |
1983 |
1984 |
1985 |
1986 |
1987 |
1988 |
No visual impairments |
3,652 |
3,599 |
3,130 |
2,128 |
4,344 |
4,281 |
Visual impairments |
204 |
187 |
180 |
113 |
219 |
233 |
Blind in both eyes |
2 |
5 |
4 |
3 |
8 |
11 |
Other visual impairments |
202 |
182 |
176 |
110 |
211 |
222 |
Vision-related diseases/ disorders |
184 |
148 |
150 |
93 |
219 |
193 |
Glaucoma |
14 |
20 |
22 |
8 |
23 |
24 |
Cataracts |
29 |
27 |
32 |
11 |
28 |
27 |
Color blindness |
126 |
90 |
83 |
67 |
144 |
126 |
Other vision-related diseases/disordersa |
19 |
16 |
17 |
8 |
28 |
26 |
Other impairments |
||||||
Hearing impairments |
454 |
412 |
389 |
237 |
494 |
483 |
Deaf in both ears |
29 |
29 |
21 |
12 |
21 |
23 |
Other hearing impairments |
425 |
383 |
368 |
225 |
473 |
460 |
Mental retardation |
15 |
12 |
14 |
16 |
14 |
21 |
Paraplegia, hemiplegia, or quadriplegia |
6 |
5 |
8 |
7 |
10 |
3 |
Cerebral palsy |
3 |
6 |
1 |
5 |
3 |
6 |
Note: In the NHIS, conditions are determined in two ways. First, participants receive one of six condition lists that ask them if they have a specific condition (see Table A-1). Second, participants are asked broad questions to reveal general health and functioning (see Table A-2, top panel). If participants reveal they have health or functioning difficulties, they are then asked what conditions cause these difficulties (see Table A-2, bottom panel). This method misses those with conditions who have no such difficulties, while the first method captures those with conditions that have no health or functioning difficulties. So only one-sixth of the sample is directly asked about blindness. This one-sixth of the sample is a random sample because |
Annex Table A-2b Sample Size of Noninstitutionalized Working-Age Civilian Men (Ages 25 to 61) in the Choice-Based Sample with Various Chronic Impairments, Diseases, and Disorders, 1983-1996
|
Year |
|||||
Group |
1983 |
1984 |
1985 |
1986 |
1987 |
1988 |
No visual impairments |
19,018 |
19,154 |
16,561 |
11,286 |
22,453 |
22,731 |
Visual impairments |
120 |
122 |
104 |
66 |
127 |
135 |
Blind in both eyes |
27 |
16 |
20 |
11 |
18 |
22 |
Other visual impairments |
93 |
106 |
84 |
55 |
109 |
113 |
Vision-related diseases/ disorders |
87 |
58 |
61 |
30 |
51 |
71 |
Glaucoma |
25 |
12 |
17 |
6 |
14 |
22 |
Cataracts |
32 |
15 |
23 |
13 |
22 |
21 |
Color blindness |
7 |
6 |
1 |
3 |
3 |
2 |
Other vision-related diseases/disordersa |
27 |
25 |
24 |
11 |
12 |
29 |
Other impairments |
||||||
Hearing impairments |
100 |
87 |
74 |
50 |
116 |
107 |
Deaf in both ears |
21 |
12 |
17 |
5 |
7 |
9 |
Other hearing impairments |
79 |
75 |
57 |
45 |
109 |
98 |
Mental retardation |
65 |
41 |
45 |
42 |
64 |
90 |
Paraplegia, hemiplegia, or quadriplegia |
40 |
38 |
29 |
16 |
36 |
30 |
Cerebral palsy |
12 |
14 |
10 |
6 |
10 |
10 |
Note: In the NHIS, conditions are determined in two ways. First, participants receive one of six condition lists that ask them if they have a specific condition (see Table A-1). Second, participants are asked broad questions to reveal general health and functioning (see Table A-2, top panel). If participants reveal they have health or functioning difficulties, they are then asked what conditions cause these difficulties (see Table A-2, bottom panel). This method misses those with conditions who have no such difficulties, while the first method captures those with conditions that have no health or functioning difficulties. So only one-sixth of the sample is directly asked about blindness. This one-sixth of the sample is a random sample because |
Annex Table A-2c Sample Size of Noninstitutionalized Working-Age Civilian Women (Ages 25 to 61) in the Random Sample with Various Chronic Impairments, Diseases, and Disorders, 1983-1996a
|
Year |
|||||
Group |
1983 |
1984 |
1985 |
1986 |
1987 |
1988 |
No visual impairments |
4,219 |
4,211 |
3,657 |
2,491 |
5,020 |
4,929 |
Visual impairments |
118 |
89 |
81 |
65 |
146 |
121 |
Blind in both eyes |
5 |
6 |
1 |
4 |
15 |
6 |
Other visual impairments |
113 |
83 |
80 |
61 |
131 |
115 |
Vision-related diseases/ disorders |
89 |
75 |
66 |
52 |
101 |
93 |
Glaucoma |
15 |
14 |
21 |
9 |
29 |
25 |
Cataracts |
30 |
32 |
30 |
23 |
33 |
44 |
Color blindness |
12 |
9 |
4 |
9 |
13 |
11 |
Other vision-related diseases/disordersa |
36 |
23 |
14 |
12 |
31 |
22 |
Other impairments |
||||||
Hearing impairments |
275 |
263 |
246 |
149 |
304 |
294 |
Deaf in both ears |
23 |
11 |
10 |
10 |
18 |
10 |
Other hearing impairments |
252 |
252 |
236 |
139 |
286 |
284 |
Mental retardation |
9 |
13 |
6 |
9 |
15 |
15 |
Paraplegia, hemiplegia, or quadriplegia |
1 |
10 |
2 |
3 |
2 |
7 |
Cerebral palsy |
2 |
4 |
1 |
2 |
6 |
3 |
Note: In the NHIS, conditions are determined in two ways. First, participants receive one of six condition lists that ask them if they have a specific condition (see Table A-1). Second, participants are asked broad questions to reveal general health and functioning (see Table A-2, top panel). If participants reveal they have health or functioning difficulties, they are then asked what conditions cause these difficulties (see Table A-2, bottom panel). This method misses those with conditions who have no such difficulties, while the first method captures those with conditions that have no health or functioning difficulties. So only one-sixth of the sample is directly asked about blindness. This one-sixth of the sample is a random sample because |
Annex Table A-2d Sample Size of Noninstitutionalized Working-Age Civilian Women (Ages 25 to 61) in the Choice-Based Sample with Various Chronic Impairments, Diseases, and Disorders, 1983-1996
|
Year |
|||||
Group |
1983 |
1984 |
1985 |
1986 |
1987 |
1988 |
No visual impairments |
20,934 |
21,214 |
18,429 |
12,658 |
25,058 |
25,278 |
Visual impairments |
91 |
100 |
85 |
47 |
99 |
84 |
Blind in both eyes |
16 |
16 |
14 |
12 |
23 |
11 |
Other visual impairments |
75 |
84 |
71 |
35 |
76 |
73 |
Vision-related diseases/ disorders |
98 |
66 |
83 |
45 |
103 |
82 |
Glaucoma |
31 |
17 |
23 |
15 |
39 |
22 |
Cataracts |
36 |
22 |
28 |
16 |
28 |
27 |
Color blindness |
0 |
0 |
0 |
0 |
0 |
0 |
Other vision-related Diseases/disordersa |
34 |
29 |
37 |
20 |
47 |
39 |
Other impairments |
||||||
Hearing impairments |
63 |
58 |
63 |
33 |
77 |
60 |
Deaf in both ears |
15 |
6 |
11 |
7 |
18 |
4 |
Other hearing impairments |
48 |
52 |
52 |
26 |
59 |
56 |
Mental retardation |
44 |
51 |
37 |
40 |
57 |
66 |
Paraplegia, hemiplegia, or quadriplegia |
12 |
19 |
16 |
5 |
16 |
20 |
Cerebral palsy |
8 |
12 |
9 |
5 |
14 |
13 |
Note: In the NHIS, conditions are determined in two ways. First, participants receive one of six condition lists that ask them if they have a specific condition (see Table A-1). Second, participants are asked broad questions to reveal general health and functioning (see Table A-2, top panel). If participants reveal they have health or functioning difficulties, they are then asked what conditions cause these difficulties (see Table A-2, bottom panel). This method misses those with conditions who have no such difficulties, while the first method captures those with conditions that have no health or functioning difficulties. So only one-sixth of the sample is directly asked about blindness. This one-sixth of the sample is a random sample because |