Page 62 - IJPS-3-1

Disability policies and public views on work disability...

concentrated in the middle of the work limitation distribution (cut-points 2 and 3).
However, the sign of the effects seems counterintuitive.
While looking at the effects of disability policy generosity and other country-level
factor, we also control for a series of respondents’ characteristics in the estimation
because reporting heterogeneity may be related to differences in population
demographics, socioeconomic characteristics, and health distribution in each country.
We test whether the predictive power of disability policy generosity is robust to
including these respondents’ individual-level factors.
Our model (in Table 4) controls for standard respondents’ demographic covariates:
gender, age (in groups), education (in years), and a series of health indicators (high
blood pressure, diabetes, cancer, lung problems, heart problems, arthritis, obesity, and
number of Activity of Daily Living or ADL limitations, and number of Instrumental
Activity of Daily Living or IADL limitations). Work disability is not purely a health
indicator but also depends on the work context. So we also include seven dummies
for occupation (technical/sales/administrative support, service, farming/fishing/
forestry, precision production/craft/repair, operators/fabricators/laborers, elementary
occupation, and managerial/professional specialty as omitted/reference occupation).
Some of these individual-level factors significantly predict vignette rating styles, but
they do not reduce the predictive power of disability policy generosity.
3.3 Counterfactual Simulations of the Effects of Disability Policy
Generosity
Having estimated the cut-points adjusted for country-level and individual-level factors,
we impose the estimated response scales on the model for self-reported work disability.
We want to see whether allowing reporting heterogeneity affects the distribution of
self-reported work disability. Specifically we focus on the disability policy effects on
the response scale which in turn shifts the distribution of self-reported disability in the
country. We illustrate such an impact with policy simulations and demonstrate whether
changing a country’s disability institutional environment would affect what the
population’s (self-reported) disability distribution looks like. The results are presented
in Figure 3. In particular, we show how the disability severity distribution in the U.S.
would look like if the U.S. were to adopt other European countries’ disability policy.
We carry out such simulations for every policy dimension, except for policies 2 and
8–10 which we combine since they are so highly correlated and reflect a country’s
overall policy towards milder work limitations. All the disability severity distributions
in the graphs are adjusted by demographics, health conditions and occupations in the
U.S. The exercise performed here is more of a comparative analysis than a simulation.
We do not consider any feed-back effect if different policies could change people’s
behavior.
The first graph depicts the policy simulation results for the first policy dimension,
disability policy coverage. The U.S. disability programs cover the labor force plus
a means-tested program. The U.S programs score 3 in terms of their generosity
according to OECD and rank in the middle compared to the other countries. If the U.S.
were to adopt Sweden’s coverage policy, which is the most generous, and covers not
only the labor force but the whole population, the simulation shows significant increase
in the likelihood of Americans reporting disability. The most pronounced change is an
increase of more than 10 percentage points (from 33.5 percent to 44.2 percent) in the
proportion of Americans reporting mild work disability, and a drop of more than 10
percentage points in the proportion reporting no disability.
The reporting pattern changes in a similar fashion except with smaller magnitude if
we applied the more generous coverage in the Netherlands, where more than the labor
force is covered but not the whole population, as compared to the U.S. Tightening
the U.S. disability policy to cover only the part of the labor force that is self-insured,
similar to the German policy, would make fewer people covered and possibly less
likely to characterize their health conditions as work disabling, at least according to the
coverage standard of the public disability programs. As shown in the graph, about 5

56 International Journal of Population Studies 2017, Volume 3, Issue 1