Page 54 - IJPS-3-1

Disability policies and public views on work disability...

disability system (index = 21) and Sweden ranks the most generous (index = 34).
However, countries rank differently in each specific policy dimension. Some policy
dimensions, such as Policies 2, 8, 9 and 10, are highly correlated, as evidenced
by the magnitudes of the correlation coefficients (0.7 or more between any two).
Moreover, these four policy dimensions affect the vignettes’ ratings in a similar
fashion, as indicated by the correlation coefficients between each policy dimension
and the vignettes classifying (Appendix Table A1). In our estimation and policy
simulation shown later, we group these highly correlated policy dimensions to reduce
collinearity. More importantly, these four aspects likely reflect how a country treats
milder disability cases, as the policies regarding the minimum disability level (Policy
2), sickness benefits (Policies 8 and 9), and unemployment benefits (Policy 10) target
partial or temporary disability cases.
The remainder of the paper is structured as follows: In Section 2, we describe the
data and the anchoring vignette approach. In Section 3, we present estimation results
and policy simulation results. Section 4 provides further discussion about the results
and concludes.

2 Data and Method

2.1 Study Sample

To study individuals’ perceptions, we look at how differently people characterize a
given level of work disability across countries. Measures of self-assessed disability
status, which are commonly available in survey data, are insufficient to conduct
this type of analysis because they will reflect both the true level of work disability
and reporting styles. We take advantage of unique data on disability vignettes from
comparable U.S. and European surveys, and use vignette data to study reporting
heterogeneity. A vignette describes the work limitation of a hypothetical person. A
respondent is asked to evaluate the severity of the vignette work limitation on the
same five-point scale used for their own health assessment. Since the vignettes are
identical for all the respondents, the differences in respondents’ evaluations must be
due to different reporting styles. We hypothesize that the scale that the respondents use
to classify the severity of a given vignette character’s work limitation is a function of
the country’s disability policy, particularly a severity classification scale used by their
country’s disability system.
We use the 2004 wave of the Health and Retirement Study (HRS), a bi-annual panel
with a representative sample of the U.S. population aged over 50 and their spouses.
It has been conducted by the University of Michigan since 1992. The information
collected includes health, socio-economic status, and social program participation.
We use a subsample of respondents who first completed a face-to-face interview and
later completed a leave-behind questionnaire that consists of a series of work disability
vignettes. We use the 2004 wave because it is the only year in the panel when a
vignette questionnaire was given to a random subgroup of respondents.
The Survey of Health, Ageing and Retirement in Europe (SHARE) is a longitudinal
dataset on European citizens of aged 50 and older and their spouses. We use the
2004 wave of the survey. SHARE was purposely modeled after the HRS and follows
a common set-up across all countries with the goal of facilitating cross-country
research. For a subset of countries that agreed to participate, SHARE included a set
of self-assessments and vignette questions on work limitations as part of a drop-
off questionnaire. The eight countries that participated in this vignette experiment
were Belgium, France, Germany, Greece, Italy, the Netherlands, Spain, and Sweden.
The work disability vignettes were identical to the work disability vignettes in the
HRS leave-behind questionnaire. In our analysis, we exclude Greece because the
comparable index for the disability policy generosity is not available in the OECD
report.

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