R. Mark Wilson, University of South Florida

This study describes the effects of the symptoms of multiple sclerosis on the labor market activities of an individual and her/his family and on the household activities of the individual. Seven hundred eleven individuals in the Tampa Bay area of Florida who have multiple sclerosis completed a lengthy questionnaire about their use of time in the labor market and at home. Sample information shows that MS symptoms (a) lower the probability of being employed, (b) reduce the hours of work, alter the occupational distribution and increase the absenteeism of those who are employed, and (c) hasten retirement. Almost 88% of the respondents say that having MS causes them to accomplish less in a typical hour spent in a variety of non-work activities; they also spend fewer hours performing this variety of household activities. Logit estimations reveal that functional limitations are correlated with both a loss of labor market production and with a reduction in the effectiveness of household time.

Multiple sclerosis is the number one disabling neurological illness of young adults (Bourdette et al., 1993; Gronning, Hannisdal and Mellgren, 1990). Onset of MS among individuals in the United States occurs most frequently, but not exclusively, from ages 20 to 40 (Baum and Rothschild, 1981). Prevalence estimates put the U.S. multiple sclerosis population at roughly 300,000, with 4.2 cases per 100,000 people newly diagnosed each year (Anderson et al., 1992; Baum and Rothschild, 1981). The disease is progressive, and its course within each individual can be predicted with only modest success (Bourdette et al., 1993; Kelly, 1985; Rodriguez, Siva and Ward, 1994; Warren, Cockerill and Warren, 1991). Multiple sclerosis occurs when the myelin coating surrounding an individual's spinal cord deteriorates and scar tissue forms, preventing nerve messages from the brain from reaching the appropriate parts of the body. The result can be a wide variety of neuromuscular symptoms including, but not limited to, pain, tremors, numbness, fatigue, and paralysis. Many symptoms are treatable, but no cure has been found for MS. The economic effect of these symptoms on an individual with MS and on her/his family can be quite large; not only might there be extra expenses, but MS in a family also will change the ways in which individuals use their labor market and their non-labor market (household) time.

Only a few studies have examined the changes in her/his activities and the reduction in an individual's productivity due to having MS. Those that focus on the effect of having MS on labor market time find disabling symptoms inhibit or prevent employment (Catanzaro and Weinert, 1992; Inman, 1987; Kornblith, LaRocca and Baum, 1986; LaRocca, Kalb, Scheinberg and Kendall, 1985) and lower the value of family earnings (Catanzaro and Weinert, 1992; Inman, 1987). The variety of symptoms make general statements about the effect of multiple sclerosis difficult and more information is needed. A similar statement can be made about the effect on the use of non-labor market, or household, time. The same symptoms that reduce labor market activity also cause individuals to "remap" their non-labor market time (Dyck, 1995), presumably reflecting a reduction in the ability to perform household activities. Economists lack data on this reduction in household production and omit this loss in cost-of-illness studies, thereby undervaluing the costs of multiple sclerosis.

The purpose of this study is to demonstrate the importance of both foregone labor market and household production due to MS, and to explore the connections between these losses and an individual's functional limitations. Seven hundred eleven individuals in the Tampa Bay area of Florida with multiple sclerosis completed a lengthy questionnaire about medical costs and time spent in labor market and non-market activities. The results provide insight into the size and nature of the changes in the productivity of market and household time due to having MS.

The Survey of Individuals with Multiple Sclerosis (SIMS) was created to gather information about individuals living with MS: its effect on their purchases of medical care, on their work experiences and on their home activities. The SIMS questionnaire was mailed to all members of the Florida Gulf Coast Chapter of the National Multiple Sclerosis Society in mid-March of 1992. By July, 1992, 711 complete and usable questionnaires had been returned, many with additional hand-written notes about the quality of their life with MS. These returns provided a response rate of 29.34%. Of the 711, 72.2% are from females. The youngest sample member is 22 and the oldest 83; the median age is 51. One hundred two (14.4%) live alone and 12 (1.7%) live in group facilities. Almost half, 48.5%, live with one other person. Two-thirds are married, and one-quarter have children under the age of 18 living with them. Ninety-four percent of the sample is White. Ninety-five percent have high school degrees and 25.5% have at least one college degree. The median age at onset of the multiple sclerosis was 30.

The twelve-page SIMS questionnaire consists of four parts in addition to the preceding general information: employment experiences, lifestyle (i.e. non-work) changes, use of medical care and functional limitations due having MS. Information from questions concerning symptoms preventing work-for-pay, employment experience, part-time versus full-time employment, occupation, absenteeism and retirement provide indications of losses in labor market production due to having MS. Individuals also state whether they spend more, the same, or less time performing several non-work activities: volunteering; shopping, running errands, and housekeeping; exercising and recreating; visiting with friends; or participating in outdoor events. Functional limitations are measured on a slightly modified version of the health state classification system for the role, physical, and social emotional functions listed in Torrance, Boyle and Horwood (1982). The three limitations used in the following analysis are "needed help to eat, dress, bathe or go to the toilet", "needed help from another person and/or mechanical aids to get around" and "had few friends and little contact with others". Responses to the questionnaire are summarized and the effect of functional limitations caused by MS on market and non-market activities are examined in the following section.

Labor Market Time

Almost 65% of the respondents state that MS symptoms currently prevent them from working for pay. Many retired early in their careers due to having MS. Two-hundred thirty-five of the 711 respondents, 33%, classify themselves as retired, and of those, over 83% say having MS was a factor in their retirement decision. The median age at retirement among members of the sample is 50, whereas the median retirement age of the U.S. population is currently 62.7 years (Gendell and Seigel, 1992).

Thirty percent of the sample is currently employed, 70% work full-time, 35 or more hours per week. The percentages are approximately the same for females and for males in the sample. Nationally, however, 74.4% of employed females and 89.6% of employed males work 35 or more hours per week (Bureau of Labor Statistics, 1992). Almost the entire sample, 98%, has labor market experience. The median number of years of work experience is 19; 17 years for the female sample and 25 years for the male sample.

Table 1 lists four occupational distributions: that of those employed, of those occupations from which people retired, of those occupations most recently lost or changed due to having MS and of individuals in the Tampa Bay area. The employed portion of the sample currently is much more concentrated in the white collar categories than is the Tampa Bay area population. It appears that the highly educated and predominantly female sample held previous jobs that were slightly

skewed toward white collar employment. Then MS symptoms caused individuals to lose jobs, to change jobs and/or to retire. The result is a distribution of current occupations with very little blue collar employment. Also, one-quarter of the employed sample are self-employed, 22.3% of the employed women and 34.9% of the employed men. This is almost three times the national rate of 9% self-employment (Bureau of Labor Statistics, 1992).

The average loss of workdays due to any illness among the employed in the SIMS sample is almost eleven days, over twice the national average of 4.8 days (U.S. Department of Health and Human Services, 1989). However, the median days of work loss per year in the sample is four; a quarter of the sample missed no days, half the sample missed four or fewer days, and 76.5% missed ten or fewer days. The high average days of absenteeism is due to a relatively few individuals missing literally months of work. The figures for males and females are quite similar.

Having MS not only affects the careers of those individuals, but it also alters the work-life of other family members (see also Catanzaro and Weinert, 1992). Sixty spouses stopped working (32 men and 28 women) and 33 spouses now work shorter hours; seventeen started working and 35 now work longer hours; 47 changed jobs, ten did not change jobs, and eleven relocated their jobs, all due to their spouse's having multiple sclerosis. Sixteen people each said their parents or their children changed their work-lives as a result of having MS in the family.

Household Time

Table 2 indicates the extent of the reallocation of household time by sample subsets that reflect three levels of attachment to the labor force: those who work full-time, those who work part-time and individuals who are of working age but whose MS symptoms prevent them from working for pay. As a general statement, respondents tend to devote less time to volunteer activities, shopping, running errands and housekeeping, exercising and recreating, visiting with friends, and participating in outdoor activities because of having MS. Those who are employed

full-time reduce their time spent in these activities the least and individuals whose symptoms prevent working reduce it the most among the three groups. Almost 94% of people in the symptoms-prevent-working group report accomplishing less in a typical hour spent at activities other than working for pay; the percentage is lower, yet still large, for those who work full-time. When asked whether changes in the listed non-work activities made them worse off, nearly twice the percentage of the symptoms-prevent-working sample selected the extreme category, "much worse," than did the full-time employed. The percentage of part-time employed respondents fell between that of the other two groups for all of the question responses.

Functional Limitations, Labor Market Time and Household Time

Functional limitations influence an individual's use of both labor market and household time. Table 2 shows the inverse relationship between employment status and self-help, mobility and social limitations. Further evidence of the detrimental effect of these limitations on the use of labor market time, as proxied by employment status, comes from three multinomial logit estimations of the probability of being in each of three employment categories, "full-time employment," "part-time employment" and "MS symptoms prevent employment (age less than 63)". In each estimation the probability is hypothesized to be a function of one of the three types of functional limitations, while holding constant gender, age, marital status, education and duration of illness. Self-care and mobility limitations significantly lower labor force participation at = .01; social limitations do not. Mobility restrictions reduce the probability of full-time work by 32% and of part-time work by 10%. In the multinomial logit estimation that includes the mobility restrictions variable, having more than a college degree and being married increase both the probability of full-time employment and the probability of part-time employment, and age reduces the probability of full-time work, all at = .10.

Fifteen logit estimations of the probability that an individual spends less time due to having MS, rather than spending the same or more time, in the five household activities listed in Table 2 provides more information about the use of household time. The existence of each of the self-care, mobility and social functional limitations, as well as gender, age, marital status, education and duration of illness are the independent variables. Mobility and social limitations significantly raise the probability of spending less time in each of the five activities at = .01 and/or .05; self-care does not. Mobility restrictions raise the probability of spending less time from nine to twenty-two percent, depending upon the activity. None of the other variables have significant coefficients at conventional levels across all the activities.

MS symptoms affect every aspect of employment; including the probability of being employed, the number of hours of work, the occupation, the tendency to be self-employed, the rate of absenteeism and the decision to retire. MS symptoms can also influence the work-lives of family members. The statistics tell a convincing story that the sample is a highly educated, experienced group of individuals, most of whom are prevented from working due to their MS symptoms. Labor market attachment could be prolonged and income loss inhibited with effective job counseling to individuals with MS and to their families, redesign of work tasks to ease the burden of mobility limitations and the availability of flextime. Well-intentioned efforts to extend and to enrich the work lives of individuals with multiple sclerosis have had only modest success, and more labor market information and a greater policy focus on these employment problems could be helpful (Rumrill, Jr. (1996)).

Respondents accomplish less per hour and spend fewer hours performing a variety of non-work activities. These losses make them worse off yet are omitted from cost-of-illness studies (Johannesson and Jonsson (1991); Posnett and Jan (1996)). Mobility limitations reduce the effective use of household time simultaneously with decreasing labor market activity. The disease captures progressively more of all time, both labor market and household. Indeed, the positive effect of marriage on employment may arise from the fact that someone else helps with household chores, enabling people with MS to devote more time and energy to the workplace. The detrimental effects of mobility and social limitations could be investigated with regard to constructive roles for home health care workers and for volunteers.

Kornblith, et al. (1986) and others noted the effect of mobility restrictions on labor market status. This study confirms those results on a recent sample and adds two points about functional limitations: (1) mobility restrictions also reduce the effectiveness of non-labor market time and (2) although other functional limitations significantly effect the use of time, mobility restrictions are the most pervasive across the activities in this study. Reducing the problems of mobility limitations would improve the quality of life of the people surveyed.

This study was supported by a grant from the National Multiple Sclerosis Society. I appreciate the excellent help and advice from Joselin Martin and every one of the respondents. I am responsible for any remaining errors.

Anderson D.W., Ellenberg J.H., Leventhal C.M., Reingold S.C., Rodriguez M. and Silberberg D.H. (1992) Revised estimate of the prevalence of multiple sclerosis in the United States. Annals of Neurology 31, 333-336.

Baum H.M. and Rothschild B.A. (1981) The incidence and prevalence of reported multiple sclerosis. Annals of Neurology 10, 420-428.

Bourdette D.N., Prochazka A.V., Mitchell W., Licari P., Burks J., and the VA Multiple Sclerosis Rehabilitation Study Group. (1993) Health care costs of veterans with multiple sclerosis: implications for the rehabilitation of MS. Archives of Physical Medicine and Rehabilitation 74, 26-31.

Bureau of Labor Statistics. (1992) Employment and Earnings 39, Washington, D.C., U.S. Department of Labor, June.

Catanzaro, M. and C. Weinert. (1992) Economic status of families living with multiple sclerosis. International Journal of Rehabilitation Research 15, 209-218.

Dyck, I. (1995) Hidden geographies: the changing lifeworlds of women with multiple sclerosis. Social Science and Medicine 40, 307-320.

Gendell, M. and J.S. Siegel. (1992) Trends in retirement age by sex, 1950-2005. Monthly Labor Review 115, 22-29, July.

Gronning M., Hannisdal E. and Mellgren S.I. (1990) Multivariate analysis of factors associated with unemployment in people with multiple sclerosis. Journal of Neurology, Neurosurgery, and Psychiatry 53, 388-390.

Inman, R.P. (1987) The economic consequences of debilitating illness: the case of multiple sclerosis. The Review of Economics and Statistics LXIX, 651-660.

Johannesson, M. and B. Jonsson. (1991) Economic evaluation in health care: Is there a role for cost-benefit analysis? Health Policy 17, 1-23.

Kelly, R. (1985) Clinical aspects of multiple sclerosis. In Vinken P.M., Bruyn G.W., Klawans H.S. and Koetsier J.C. eds. Demyelinating Diseases. Amsterdam: Elsevier, 49-78. (Handbook of clinical neurology; vol. 47.)

Kornblith A.B., LaRocca N.G. and Baum H.M. (1986) Employment in individuals with multiple sclerosis. International Journal of Rehabilitation Research 9, 155-165.

LaRocca, N.G. et al. (1985) Factors associated with unemployment of patients with multiple sclerosis. Journal of Chronic Diseases 38, 203-210.

Posnett, J. and S. Jan. (1996) Indirect cost in economic evaluation: the opportunity cost of unpaid inputs. Health Economics 5, 13-23.

Rodriguez M., Siva A., Ward J., Stolp-Smith K., O'Brien P., and Kurland L. (1994) Impairment, disability, and handicap in multiple sclerosis: a population-based study in Olmsted County, Minnesota Neurology 44, 28-33.

Rumrill, Jr., P.D. (1996) Employment Issues and Multiple Sclerosis. New York: Demos Vermande.

Torrance, G.W., M.H. Boyle and S.P. Horwood. (1982) Applications of multi-attribute utility theory to measure social preferences for health states. Operations Research 30, 1043-1069, November-December.

U.S. Department of Health and Human Services. (1989) Vital and Health Statistics: Health Characteristics by Occupation and Industry: United States, 1983-85. Series 10, No. 170, National Center for Health Statistics, Hyattsville, Maryland. DHHS Publication No. (PHS) 90-1598.

Warren S., Cockerill R. and Warren K.G. (1991) Risk factors by onsetage in multiple sclerosis. Neuroepidemiology 10, 9-17.