Labor Force Participation and Disability: Did HomeBased Work Facilitate Labor Force Participation in

1
Labor Force Participation and Disability Did
Home-Based Work Facilitate Labor Force
Participation in the Dawn of the Americans with
Disabilities Act?

• Jennifer Tennant

2
Americans with Disabilities Act

• Provides protection for those in the private and
  public sector
• Signed into law on July 26, 1990 and went into
  effect two years later
• Title I of the ADA addressed the employment
  situation of people with disabilities
• It required that employers take steps to
  reasonably accommodate qualified individuals
  with a disability
• Can accommodate at the onsite worksite or by
  letting workers to work at home
• Home-based work goes against traditional
  constructs of work, so presents challenges in
  implementation

3
Four Sections of Presentation

• Section I Introduction
• Outline the ADA. How does home-based work fit in
  its mission? Does it create difficulties in
  implementation?
• Discuss the literature of Labor Force
  Participation and Disability Status
• Define the various measures of disability in the
  Integrated Public Use Microdata Series (IPUMS)
• Summarize research questions
• Section II summarizes the IPUMS data to compare
  the disabled and the non-disabled in and out of
  the workforce and by worksite choice

4
Four sections (continued)

• Section III Economic theory behind labor force
  participation and disability status
• Modeled after "Home-Based Work and Women's Labor
  Force Decisions" by Linda Edwards and Elizabeth
  Field-Hendrey.
• The presence of a disability alters the fixed
  cost of working, but in various degrees depending
  if one is an onsite worker, a home-based worker,
  self-employed or an employee.
• Section IV Econometric Model
• Multinomial logit coefficients and marginal
  effects for this labor force participation problem

5
Qualified individuals with a disability

• The ADA protects "qualified individuals" with a
  disability.
• Title I of the ADA defines a qualified individual
  as an employee who "with or without reasonable
  accommodation . . . can perform the essential
  functions of the employment position that such
  individual holds or desires" (42 U.S.C.
  12111(8)).
• The concept of which individuals are qualified
  is inextricably bound with the concept of
  reasonable accommodation.
• Employers have a responsibility to try to
  accommodate, but rules do not define any specific
  accommodations

6
Reasonable Accommodation

• An accommodation would be reasonable if it takes
  into account the specific job requirements as
  well as the limitations and skills of the
  employee.
• An accommodation is not considered reasonable if
  an employer has to change the "essential
  functions or components" of the position.
• An employer does not have to offer the employee a
  job that has requirements beyond the employee's
  skill level to reasonably accommodate.
• Acceptable reasonable accommodations include
  creating accessible workplaces, modifying work
  schedules, and altering responsibilities and
  examinations.

7
Burden of Proof

• Employees must show
• she or he is indeed disabled
• with reasonable accommodation, the employee can
  complete the essential functions of the job.
• Employers must show
• - The essential functions of the job are in fact
  essential, and that accommodating the worker
  would create an undue burden on the firm.

8
Working at home a more viable option

• Strides in technology have made working from home
  a more viable option
• Benefits increased flexibility and control,
  lower transportation costs, savings on office
  overhead, lower employee absenteeism, increased
  productivity, improved employee morale, and
  higher employee retention
• Drawbacks some jobs require face-to-face
  contact with the public or clients is necessary,
  performance monitoring and communication delays.

9
The Vande Zande presumption against working from
home

• Vande Zande v. Wisconsin Department of
  Administration addressed whether working from
  home is a reasonable accommodation under the ADA.
  
• Looked at excessive absenteeism cases to build
  its foundation
• Held that a person is not qualified for a
  position if she cannot maintain predictable
  attendance
• This created a presumption that basically said
  that home-based work could never be a reasonable
  accommodation since physical presence was an
  essential component of a job

10
Is this presumption wrong?

• Vande Zande was decided in 1995, right at the
  dawn of the Internet
• Technology has grown drastically since then
• The presumption against home-based work goes
  against the purpose of the ADA, which is to help
  bring disabled workers into the ranks of the
  employed.
• A work-from-home arrangement might be the only
  viable option for a person who cannot leave home
  on a regular basis.

11
Labor Force Participation and Disability Status

• Employment for people with disabilities have been
  getting worse rather than better since the ADA
  took effect. This seems to be counterintuitive
  and goes against the goals of the ADA.
• Individuals with disabilities have lower
  employment rates than the non-disabled
  population.
• Factors that contribute to this are high
  reservation wages and low market wages that
  reflect decreased productivity and
  discrimination.
• Will employers be reluctant to hire workers with
  disabilities after the ADA because of increased
  accommodation and firing costs?

12
Definition of Disability According to the ADA

• To be considered as a person with a disability
  under the ADA, at least one of three criteria
  must be met
• a physical or mental impairment that
  substantially limits one or more of the major
  life activities of such individual
• a record of such an impairment or
• being regarded as having such an impairment. (42
  U.S.C. 12102(2))
• To be covered, I must be disabled enough but
  not too disabled (that is, I must still be able
  to work). These are vague guidelines.

13
Definitions of Disability in IPUMS data

• In the 1990 wave of the Census, three different
  types of disabilities were outlined --
  disabilities limiting work, disabilities limiting
  mobility and personal care limitations.
• In 2000, these three disability measures
  remained, and there were three additional
  categories physical difficulty, difficulty
  remembering and vision and/or hearing
  disabilities.

14
Research Questions

• This dissertation will focus on how the option of
  home-based work affected the labor force
  participation of the disabled during the dawn of
  the Americans with Disabilities Act.
• Was the employment landscape for the disabled
  more favorable in 2000 than in 1990, after the
  passage of the ADA?
• Did the reasonable accommodation mandate of the
  ADA and technology improvements make home-based
  work a more viable option for the disabled?
• In sum, did home-based work facilitate labor
  force participation of the disabled?

15
Landscape for the disabled in 1990

• In 1990, there were 6,007,023 people who
  described themselves as having at least one type
  of disability.
• More than half (3,173,684) of the disabled were
  on-site workers approximately 48 as onsite
  employees and 4.9 as the onsite self-employed.
• Only 101,219 disabled people worked at home
  0.5 as employees and 1.1 in a self-employment.
  
• More than 45 of the disabled in 1990 were out of
  the labor force.
• This dearth of home-based workers could be the
  consequence of the lack of telecommuting
  resources and computing power at this time.
• 28.9 of those who were out of the labor force in
  1990 had a disability, compared with 5.5 of
  onsite employees, 5.8 of the onsite
  self-employed, 7.1 of home-based employees and
  6.4 of the home-based self-employed.

16
Landscape for the disabled in 2000

• In 2000, 9,678,458 people identified as having at
  least one type of disability, a 61.1 increase
  from 1990.
• In 2000, 77.6 of the disabled were on-site
  employees, 4.7 were self-employed on-site, .9
  were home-based employees, 1.1 were
  self-employed at home and the remaining 15.6 of
  the disabled were out of the labor force
• There were 195,074 home-based workers and
  7,972,682 onsite workers who had at least one
  disability in 2000.
• People with disabilities comprised 12.1 of
  home-based employees, and 11.4 of the
  self-employed who worked at home. Those with
  disabilities made up 13.6 of on-site employees
  and 13.4 of the self-employed who worked
  on-site. 1,510,702 disabled people were out of
  the labor force in 2000, which accounted for
  24.5 of those not in the labor force.

17
Change from 1990 to 2000

• Between 1990 and 2000, both the employment status
  and worksite location of the disabled had
  undergone important changes.
• A greater number and percentage of the disabled
  were on-site employees and home-based employees
  and fewer were out of the labor force.
• The number of the disabled who were self-employed
  onsite or at home increased, but the percentage
  fell or stayed the same.

18
Change from 1990 to 2000, part 2

• Almost 5 million more disabled people worked in
  2000 than in 1990 and more than 1 million fewer
  identified as being of out of the labor force
  during that time.
• Two things seemed to be happening during this
  period
• disabled people who were out of the labor force
  in 1990 became employed either onsite or at home
  in 2000 and
• those who didnt identify themselves as disabled
  in 1990 did so in 2000.
• The ADA may have had an influence on both of
  these factors.

19
Table 1
20
Table 1, continued
21
Table 2
22
Multiple Disabilities

• There was a large amount of overlap between the
  disability categories of the IPUMS
• almost 20 of those with a work disability also
  had a disability that limited mobility, and more
  than 10 of those identifying a work disability
  also had a personal care limitation.

23
Table 3
24
Wage and Salary Differences

• Regardless of self-employment status or worksite
  choice, the disabled always earned less than
  their non-disabled counterparts.
• In 1990, the average salary of a disabled
  employee was approximately less than half than
  the salary of their non-disabled counterpart.
• The situation had improved for disabled employees
  by 2000. Not only were they making significantly
  more money than in 1990, the income-divide
  between them and their non-disabled cohorts had
  decreased.

25
Wage and Salary Differences, Pt 2

• A disabled onsite workers income was
  approximately 47 of their counterparts income
  in 1990, but had risen to 77 of the income of
  the non-disabled onsite worker in 2000.
• In 1990, the income of a home-based disabled
  employee was approximately 51 of the income of
  the home-based non-disabled employee. In 2000,
  this proportion had risen to 73.
• On average, it is more lucrative to be an
  employee than self-employed, regardless of
  disability status, worksite choice or year. This
  became even more striking in 2000 than in 1990,
  since incomes for the self-employed dropped
  during the 1990s, while incomes for employees
  rose.

26
Table 4
27
Public Assistance

• The disabled were much more likely to be on
  public assistance than their non-disabled
  counterparts in 1990.
• Those who are out of the labor force are markedly
  more likely to be on public assistance than those
  who are onsite workers and home-based workers,
  regardless of disability status.
• Onsite workers are the least likely to be on any
  sort of public assistance like welfare or Social
  Security.

28
The Change from AFDC to TANF

• The presence of the ADA was not the only change
  that happened in the 1990s. In 1996, Congress
  passed the Personal Responsibility and Work
  Opportunity Reconciliation Act that radically
  changed the welfare system in the United States.
  Welfare changed from Aid for Families with
  Dependent Children (AFDC) to Temporary Assistance
  for Needy Families (TANF).
• TANF is much more restrictive than AFDC, and
  includes work requirements and time limits. It
  also decreases the implicit tax on non-welfare
  earnings to incentivize work.
• As the cushion of disability insurance lessened
  for those with disabilities in the 1990s, they
  may have been more likely to enter the labor
  force.
• In 2000, the number and percentage of those on
  welfare were both much lower than in 1990. The
  disabled who were out of the labor force were
  still more likely in percentage terms to receive
  welfare benefits than the non-disabled and the
  disabled in the workforce, but the divide had
  markedly decreased.

29
Table 5
30
Theory Behind Home Based Work Model

• Based on Home-Based Work and Womens Labor Force
  Decisions by Linda Edwards and Betsy
  Field-Hendrey, which is based on previous work
  done by John Cogan on fixed costs and labor force
  decisions.
• They outline two differences between home-based
  and onsite work.
• First, the fixed costs associated with working
  (e.g. time costs associated with commuting,
  out-of-pocket commuting expenditures and clothing
  costs) are greatly reduced for home-based
  workers. Second, home-based workers may be able
  to engage in some joint production of income and
  household commodities. (Edwards and
  Field-Hendrey, 174)

31
Applying the Model to the Disabled

• Edwards and Field-Hendreys analysis focuses on
  the labor force participation of married women.
• Married women often do the majority of the
  child-rearing and household maintenance, so they
  have a different decision making process than men
  when considering whether or not to enter the
  labor force and for how many hours.
• Women have different reservation wages, that is
  the minimum wage level to induce a person into
  the market, and this reservation wage is
  dependent on many factors, including other income
  and presence of children in the family.
• Since the option of home-based work reduces many
  of the costs of working, Edwards and
  Field-Hendrey states that the presence of the
  home-based work option leads some women who would
  have chosen to be out of the labor force to enter
  as a home-based worker (176).
• The theoretical basis for the disabled is similar
  to that of married women.
• The disabled, too, have greater costs of working
  than their non-disabled counterparts.
  Transportation, mobility and accommodation issues
  affect the disabled more than the non-disabled.
  Also, a household commodity for a disabled
  person could be having a doctors or therapy
  appointment, or taking care of oneself when
  symptoms flare up. Home-based work could provide
  the flexibility and lower costs of work that
  could induce those with disabilities to enter the
  labor force.

32
The Econometric Model -- Data

• The data used in this article are from the
  Integrated Public Use Microdata Series (IPUMS) of
  the Census for the years 1990 and 2000. The
  sample used is the 5 sample of housing units,
  and includes all people aged 25-55 who did not
  live in group quarters, were not in school or in
  the military. This sample also does not include
  the unemployed it only includes those who are
  either employed or out of the labor force.
• Home-based work status is taken from the
  "transportation to work" question of the survey,
  question 23a.
• "How did this person usually get to work LAST
  WEEK?"
• If the respondent's answer was "worked at home,"
  then he or she is considered a home-based worker
  in this study.
• People who work at home occasionally are not
  considered home-based workers for the purpose of
  this study since their motivations for working at
  home may differ with those who have chosen it as
  their primary worksite.
• The sample only includes those aged 25-55 to
  separate the worksite decision from the school or
  retirement decision.
• Special weighting to make the samples of onsite
  and homebased workers approximately equal

33
The Econometric Model

• There are three separate, but related questions
  that have to be answered by an econometric model
  and our data. First, are those with disabilities
  more or less likely to out of the labor force?
  Second, for those who are in the labor force, are
  the disabled more or less likely to be a
  home-based worker? Third, was there any change
  in the first two questions between 1990 and 2000?
  
• To answer these questions, I used a multinomial
  logistic regression model (mlogit) and the IPUMS
  data for 1990 and 2000.
• The mlogit regression is used because the
  dependent variable is a polytomous variable.
• The dependent variable, workstate, has only 5
  choices out of the labor force, on-site
  employee, on-site self-employed work, home-based
  employee and home-based self-employed worker.

34
Marginal Effects

• After calculating the mlogit coefficients, I will
  determine the marginal effects that is, the
  change in the predicted probabilities of being in
  each category connected to changes in the
  explanatory variables.
• Marginal effects look at a one unit change in the
  explanatory variables when the variables are
  continuous, and the change from 0 to 1 in
  dichotomous variables.
• Focus on the marginal effects related to
  disability type to answer questions like
• What is the change in the predicted probability
  of being an on-site employee when a person has a
  work disability compared with not having a work
  disability?
• What is the change in the predicted probability
  of being a home-based employee when a person has
  a personal care limitation compared with those
  who do not?

35
Independent Variables

• Using the data from 1990, the independent
  variables of the regression were age, age
  squared, education level, education level
  squared, age multiplied by education level,
  married, white, black, hispanic, presence of
  children in the household in the following age
  categories under one, one to two, three to
  five, six to twelve, and thirteen to seventeen,
  rural status, welfare recipient status, Social
  Security recipient status, other income and one
  of the three different disability measures a
  disability that limited work, a mobility
  disability and a personal care limitation.
• The regression was the same for 2000, but
  included SSI recipient status and three other
  disability measures physical difficulty,
  difficulty remembering and vision and/or hearing
  difficulty.
• These four variables were not added to the IPUMS
  data until 2000.
• The logit regressions were weighted with the
  home-based vs. onsite weighting described
  previously.

36
How Can We Separate the Effect of the ADA from
other Effects?

• The protections of the ADA only apply to
  employees those who choose or are forced to
  choose to work for themselves do not get the
  protection of reasonable accommodation.
• We can look at how the marginal effects of the
  home-based and onsite employees changed relative
  to the self-employed workers to see the impact of
  this legislation on the labor force participation
  of the disabled.

37
Table 6
38
Table 7
39
Table 8
40
Table 8, continued
41
Table 9
42
Table 9, continued
43
Table 10
44
Results

• In 1990, having a disability drastically
  decreased the probability of working as an
  on-site employee.
• The disability limiting work category,
  understandably, decreased the probability the
  most by 24 percent.
• A person who had a mobility disability in 1990
  was almost 13 percent less likely to be an
  on-site employee.
• In 2000, these effects changed in a remarkable
  way. Now, an employee with a work disability was
  1.1 percent more likely to work on-site.

45
Results, continued

• Home-based employees
• In 1990, having a work disability or a mobility
  disability slightly increased the probability of
  being a home-based employee.
• Having a personal care limitation slightly
  decreased that probability, by .1 percent.
• For the most part in 2000, disability status
  slightly increased the probability of being a
  home-based employee.

46
Results, continued

• The self employed
• In general, the changes in probabilities for
  self-employed workers were either negative or
  much more muted in a positive way, than for those
  who were employees.
• This lends credence to the fact that the ADA may
  have had a positive influence on the employment
  situation of those with disabilities, giving them
  more options to work.

47
Conclusions

• Between 1990 and 2000, the disabled were more
  likely to be working and less likely to be out of
  the labor force.
• The vast majority of those who entered the
  workplace went to onsite jobs, but home-based
  work provided an important place as well for
  increased labor force participation of the
  disabled.
• As the digital age continues to thrive, there is
  greater and greater place for home-based work
  opportunities.
• The onsite world of work will always be dominant,
  but the reasonableness of working from home
  increases with every technological advance.