brandon_Marylandpresentation

Document Sample
brandon_Marylandpresentation Powered By Docstoc
					“Does Disability
among Older
Couples Elicit
Gender-Neutral
Time
Allocations?”



Peter D. Brandon
Carleton College,
Northfield, MN


Paper prepared for the
International Perspectives on
Time Use Conference, June
23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?
                                                Page Title
Introduction:
There are two well-known trends--
     1. People are living longer, getting healthier, and capable of working
          into older ages.
     2. People, especially men, have reduced their labor force
          participation and retiring earlier.
•To sum it up: People are living longer and working less.
•For many older couples, living longer with better health will herald in a
lifestyle characterized by less time spent in paid work and more time spent
in leisure, socializing, volunteering, exercising, and recreation.

•But for a sizeable minority of couples in older ages, a partner with a
disability, or a long-term health condition, or a physical restriction and/or
limitation will profoundly alter the couples’ choice set of activities and how
much time they can devote to those activities either jointly or individually.


.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?
                                                Page Title
Introduction:
•Among the many activities potentially affected by an older
partner having a disability or health condition is housework.
•In other words, among older couples, either retired or moving
towards retirement, the time each partner spends on
housework will be affected by their physical state and health.
•Theoretically, physical restrictions and limitations, or long-
term health conditions among older partners should alter the
balance of time allocations to housework, just like they would
alter the balance to time allocations to paid work.
•This study poses this question: Do disabilities among older
couples elicit more “gender-neutral” time allocations to
housework?


.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?
                                                Page Title
Introduction:
•I argue that we should observe increases in an older partner’s allocations
of time to housework if the other partner is disabled.
•Motives might include--
     •Partner altruism and other socio-emotional attributes
     •Desire to maintain household production at present levels
•However, empirically observing this “partner compensatory time effect”
should depend upon:
     (a)How we think about and measure “disability”;
     (b)Observed configurations of disabilities within households; and,
     (c) Availability & affordability of substitutes for the “healthy” partner.

In this study, I examine (a) and (b) using data from the 2006 Australian
Time Use Survey or “AuTUS”.



.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

AuTus Data:                                     Page Title
Nationally representative sample of households
Data on time allocation among individuals 15 years and older
6,961 individuals responded (82.5% response rate)
Survey conducted over calendar year: covered 4 periods, official
holidays, over-sampled weekends, 2 consecutive diary days
Hierarchical file structure:
        Levels: activity, household, individual, family, episodes.
Time episodes record seven components
Contained 1,169 partnered couples—at least one of them 50 years of
age or older.
        Older persons living alone or without a partner excluded
        Males’ main activity: 46% work, 36% retirement, 7.6% inactivity due to
         illness/disability/injury/handicap, 6.3% home duties, 2.1% volunteering
        Females’ main activities: 28% working still, 17.6% retirement, 4.6% inactivity due to
         illness/disability/injury/handicap, 42.3% home duties, 2.3% volunteering
        42% of couples both 55 years of age or older

.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

                                                Page Title
Data: Measures of disability and health conditions
Asked of individuals 15 years & older
Three measures used in this study:
       1. Disability or health condition that lasted or likely to last
          6 months or more.
                11 possible responses (Eyesight, Hearing, Speech, etc..)
       2. Core activity limitation in one of the three core activity
          areas: self-care, mobility, & communication due to
          long-term health condition.
                Levels: mild, moderate, severe, profound.
       3. Self-assessed current health status.
                Ratings: poor, fair, good, very good, excellent.
Measures used to develop simple typologies of
“conditions” within couple dyads
.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

Data: Time spent on housework                   Page Title
Asked of individuals 15 years & older
Duration of time recorded on housework only
recorded if longer than five minutes
Only analyzed here if recorded the primary activity
Average minutes per day on activity
Committed time
Activities include domestic tasks including: food
and drink preparation/cleanup, laundry and clothes
care, grounds and pet care, sewing and knitting,
dusting, beer making, etc. (See 4150.0 - Time Use Survey:
User Guide, ABS, 2006.)

.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

                                                Page Title
Data: Identifying information & demographic measures
Asked of individuals 15 years & older
Richer array of individual, spousal/partner, income
unit, family, and household identifiers.
       Permits exact matching of couples—de facto & married
       Survey collected data on all persons in household
        Rich amount of information collected on individuals
        time use, labor market participation, income sources,
        household and family composition, labor-saving
        devices, technological aids, attitudes and perceptions.
Missing data on either partner reduced the sample
to 870 older couples. This is the analytical sample
used.
.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

                                                Page Title
Empirical approaches to understanding time allocations
among older couples to housework with 3 alternative
measures of disabilities:
1.Bivariate analyses, Tables 1, 2, and 3.
2.Multivariate analyses, Tables 4, and 5.
3.Graphs of predicted time, Figures 1 and 2
4.Analyses of marginal effects, Table 6.
Statistical model for multivariate analyses:
Tobit models
        Unit of analyses are couples
        Regressions on wives and husbands’ time allocations to housework
                Major interest is estimating effects of disability configurations among couples on
                 housework net of controls
                Contrasting predicted times in housework by disability measures
                Generating estimates of marginal effects for the discrete disability measures.


.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

Tobit models                                   Page Title
In the models, predictors of interest are the three disability
measures
      1. Disabilities or health conditions configurations for older couples
             1.      Neither had disability or health condition
             2.      Male did, female did not
             3.      Female did, male did not
             4.      Both did
      2. Core activity limitations
             1.      Neither had disability or health condition
             2.      Male did, female did not
             3.      Female did, male did not
             4.      Both did
      3. Self-perceived poor health
             1.      Neither reported poor health
             2.      Male did, female did not
             3.      Female did, male did not
             4.      Both did
.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

Tobit models                                   Page Title
        Insufficient time today to discuss other covariates, but those results
         are interesting.

        Covariates included: age dummies, educational attainment
         categories, retirement configurations among couples, labor force
         attachment, housework income, dummy for geographic region, hire
         home help, and other survey design controls.




.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

               Page Title
Findings: Bivariate analyses
Results differ depending upon measure used.
        Table 1:
                Couples jointly disabled (or non-disabled) varies greatly by measure.
                Essentially among equal proportions of couples, one or the other has a core
                 limitation or self-assessed poor health, not so for disability or a long-term health
                 condition.
                Older males getting up to twice as likely to state have disability/LT health
                 condition.




.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Table 1. Distribution of “Disability” among older couples by three alternative measures
                                                        Measure of “Disability”
                                    Disability or Long-       Core Activity         Self-assessed
                                       term Health             Limitationb         “poor” healthc
Proportion of couples:                  Conditiona
Neither partners have, No                   0.28                   0.71                  0.90
Male Yes/Female No                          0.24                   0.12                  0.06
Male No/Female Yes                          0.14                   0.12                  0.04
Both partners have, Yes                     0.34                   0.05                  0.01
                                                1.00                     1.00                    1.01
Source: ATUS 2006, ABS. Notes: N = 870 couples with one partner at least 50 years of age; aA disability
or long-term health condition exists if a limitation, restriction, impairment, disease or disorder, had
lasted, or was likely to last for at least six months, and which restricted everyday activities; bCore
activity limitation is based on whether a person needs help, has difficulty, or uses aids or equipment
with any of these core activities: self-care, mobility, or communication. A person's overall level of
core activity limitation is determined by their highest level of limitation in these activities;
cRespondent's perception of their current health status.
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

               Page Title
Findings: Bivariate analyses
Results in Table 2 again show differences depending upon
measure used. In Table 2,
                Male partners allocate more time to housework when wives have disabilities &
                 they don’t compared with other older males, but estimated amount differs by
                 measure adopted.
                Female partners allocate more time to housework when husbands have
                 disabilities & they don’t compared with other older females, but estimated
                 amount differs by measure adopted.
                        Female partner increase in time to housework appears proportionately more than males, at least
                         from the back of an envelope calculation
                When both partners are disabled, males still do much less housework compared
                 with their wives. The amount of difference depends upon the measure.
                Two instances where male allocates equal or more time to housework is when
                 female partner has a core activity limitation or self-perceived poor health.
                        Related, if measure is female has disability or a long-term health condition, she still does more
                         household compared with her male partner.
                Worth noting results go in opposite directions in an important instance: for
                 couples both with disabilities or long-term health conditions—that’s the first
                 measure, they both spend more time on housework, but not so if core
                 limitations or poor health used.
.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Table 2. Average daily time in minutes spent on housework among older couples by three
alternative measures of disabilitya
                                Disability or Long- Core Activity Limitation Self-assessed “poor”
Measure used:
                              term Health Condition                                  health




                                Female        Male        Female         Male        Female        Male
Neither partners have, No
                                 28.5         16.7          31.4         20.5         32.7         21.6
Male Yes/Female No
                                 36.0         21.3          41.3         17.8         36.3         15.1
Male No/Female Yes
                                 35.6         23.9          32.0         32.0         22.3         31.0
Both partners have, Yes
                                  32.0            24.7         27.2        20.5          30.8       13.0
Source: ATUS 2006, ABS. Notes: N = 870 couples with one partner at least 50 years of age; aSee
definitions of alternative measures of disability in Table 1 and methods section of paper; housework is
primary time activity.
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

               Page Title
Findings: Bivariate analyses
Table 3:
       Older couples with no disabilities or health conditions,
        core activity limitations or perceived poor health were
               relatively younger
               Better educated
               Had more income, less government income transfers
               Less likely to live by themselves
       When males only partner disabled those couples had
        lower household income compared with couples where
        female only partner disabled
               Perhaps related to male’s higher past earnings, present
                earnings, greater attachment to labor force over life course,
                wealth accumulation.

.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Table 3. Characteristics of older couples by gender, disabilities among couples, and three alternative measures of disabilitya
                 Disability or Long-term Health Condition                     Core Limitation                              Self-assessed “poor” health
               Neither     Male          Male         Both        Neither Male          Male           Both       Neither Male             Male           Both
               partners    Yes/Female No/Female partners partners Yes/Female No/Female partners partners Yes/Female No/Female partners
               have, No No               Yes          have,       have,   No            Yes            have,      have,     No             Yes            have,
                                                      Yes         No                                   Yes        No                                      Yes
Females
Age 50-59        0.67        0.42          0.45         0.32        0.52   0.33            0.29          0.26       0.47      0.44            0.21          0.25
Age 60-69        0.26        0.41          0.34         0.32        0.32   0.31            0.39          0.37       0.33      0.25            0.32          0.50
Age 70 plus      0.08        0.17          0.20         0.36        0.16   0.36            0.32          0.37       0.20      0.31            0.47          0.25
Less than        0.51        0.57          0.57         0.65        0.56   0.62            0.70          0.60       0.56      0.67            0.82          0.88
H.S
H.S. only        0.16        0.14          0.21         0.19        0.17   0.17            0.14          0.21       0.18      0.19            0.06          0.13
Some college     0.14        0.11          0.11         0.08        0.11   0.09            0.08          0.12       0.11      0.08            0.09          0.00
College plus     0.19        0.18          0.11         0.08        0.16   0.11            0.09          0.07       0.15      0.06            0.03          0.00
Weekly           5.08        4.63          4.15         3.84        4.64   4.17            3.68          3.49       4.50      3.93            3.18          3.00
income.
Carer’s          0.02        0.10          0.02         0.10        0.05   0.17            0.05          0.14       0.06      0.23            0.03          0.00
allowance
Pension          0.14        0.30          0.25         0.41        0.24   0.38            0.39          0.44       0.27      0.29            0.53          0.50
Males
Age 50-59        0.53        0.29          0.38         0.20        0.40   0.24            0.19          0.09       0.36      0.27            0.03          0.38
Age 60-69        0.34        0.38          0.37         0.35        0.36   0.31            0.38          0.33       0.36      0.33            0.32          0.38
Age 70 plus      0.13        0.33          0.25         0.45        0.24   0.45            0.43          0.58       0.29      0.40            0.65          0.25
Less than        0.38        0.45          0.43         0.43        0.40   0.44            0.47          0.51       0.41      0.52            0.50          0.50
H.S
H.S. only        0.31        0.35          0.30         0.38        0.33   0.41            0.31          0.37       0.34      0.33            0.38          0.50
Some college     0.09        0.07          0.09         0.06        0.09   0.05            0.06          0.02       0.07      0.15            0.03          0.00
College plus     0.22        0.12          0.18         0.13        0.18   0.10            0.16          0.09       0.18      0.00            0.09          0.00
Weekly           6.77        5.33          6.02         4.75        6.01   4.67            4.77          4.48       5.75      4.52            4.67          3.38
income.
Carer’s          0.02        0.01          0.03         0.02        0.01   0.00            0.08          0.00       0.02      0.02            0.09          0.00
allowance
Pension          0.14        0.33          0.24         0.43        0.26   0.38            0.40          0.47       0.29      0.33            0.53          0.25
Couple
Live in rural    0.17        0.12          0.12         0.15        0.14   0.20            0.09          0.26       0.15      0.17            0.09          0.00
Couple only      0.70        0.78          0.70         0.86        0.74   0.86            0.83          0.93       0.77      0.79            0.85          0.88
Source: ATUS 2006, ABS. Notes: N = 870 couples with one partner at least 50 years of age; aSee definitions of alternative measures of disability in Table 1 and
methods section of paper.
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

                  Page Title
Findings: Multivariate analyses
Contrast group: couples in which neither partner reports a disability.
Couples in which only males have disabilities:
             o He spends between 6.7 & 14.6 fewer av. minutes/day on housework,
             based upon 2 of the 3 measures
             o She spends between 6.2 & 8.6 more av. minutes/day on housework,
             based upon 2 of the 3 measures
Couples in which only females have disabilities:
             oHe spends between 6.6 & 9.5 more av. minutes/day on housework,
             based upon 2 of the 3 measures
             o She spends either more time, 6.6, or less time, 13.5, av. minutes/day on
             housework, based upon 2 of the 3 measures
Couples in which both partners have disabilities: Mixed results
             o He spends 10.8 & she 11.8 fewer av. minutes/day on housework, based
             upon one measure
             o Though statistically insignificant, if use 1st measure in Table 4 both
             spend more average daily minutes, if use 2nd measure both spend less.
.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Table 4: Tobit regressions estimating effects of alternative measures of disability and health on time allocated to housework among older couples.
Measure:                         A disability or health condition                      A core limitation                         Perceived poor health
Time spent on housework:         Husband                    Wife              Husband                     Wife           Husband                      Wife
Husband only with trait              0.42                  6.20**              -6.79**                   8.64**           -14.66***                   4.79
                                   [0.14]                 [2.12]               [-2.29]                  [2.01]             [-3.14]                   [0.83]
Wife only with trait                6.60*                  6.56*                9.46**                   -2.14                3.27                -13.52***
                                   [1.85]                 [1.65]                [2.44]                  [-0.63]             [0.70]                  [-2.86]
Both with trait                      2.86                   2.49                 -4.61                   -5.55            -10.18***                -11.86**
                                   [1.05]                 [0.83]               [-1.10]                  [-1.46]            [-2.62]                  [-2.17]
Couple household only                3.10                 -8.13**                 3.26                 -8.54***               2.93                 -8.52***
                                   [1.29]                 [-2.45]               [1.38]                  [-2.62]             [1.27]                  [-2.59]
He's retired                        5.72*                   1.65                 5.40*                    1.75               5.22*                    1.53
                                   [1.82]                 [0.54]                [1.67]                  [0.57]              [1.62]                   [0.50]
She's retired                       -2.53                  -1.61                 -2.47                   -0.33               -0.84                   -1.95
                                  [-0.44]                 [-0.21]              [-0.44]                  [-0.04]            [-0.15]                  [-0.25]
Both retired                        -0.85                  -3.77                 -1.19                   -3.85               -1.88                   -5.05
                                  [-0.22]                 [-1.05]              [-0.31]                  [-1.06]            [-0.48]                  [-1.39]
60-69 years of age                  -1.43                  6.49*                 -1.91                   7.15**              -1.86                   7.17**
                                  [-0.54]                 [1.88]               [-0.73]                  [2.01]             [-0.71]                   [2.03]
70 plus years of age                -2.40                  8.43*                 -2.43                   8.44*               -3.05                   9.26**
                                  [-0.64]                 [1.84]               [-0.65]                  [1.87]             [-0.82]                   [2.02]
High school only                     1.39                  -2.63                  1.66                   -2.74                1.35                   -3.10
                                   [0.54]                 [-0.86]               [0.65]                  [-0.93]             [0.53]                  [-1.03]
Trade or certificate qual.          -3.17                  -3.22                 -2.94                   -3.16               -2.34                   -3.26
                                  [-0.97]                 [-0.95]              [-0.91]                  [-0.96]            [-0.72]                  [-0.98]
University or more                   0.35                  -0.40                  0.15                   -0.17               -0.48                   -0.28
                                   [0.12]                 [-0.10]               [0.05]                  [-0.04]            [-0.16]                  [-0.07]
Total weekly income                 0.88*                 -1.66**                0.86*                  -1.77**              0.81*                  -1.70**
                                   [1.79]                 [-2.42]               [1.79]                  [-2.53]             [1.68]                  [-2.44]
Working full-time                -13.89***                 -1.93              -14.73***                  -1.53            -16.37***                  -2.72
                                  [-4.06]                 [-0.34]              [-4.10]                  [-0.27]            [-4.35]                  [-0.49]
Working part-time                -10.69***                 -3.42              -11.94***                  -3.24            -12.48***                  -4.30
                                  [-3.23]                 [-0.88]              [-3.39]                  [-0.84]            [-3.44]                  [-1.16]
Old age pension                    8.55***                 -1.47                8.30***                  -1.49              7.85***                  -1.24
                                   [2.85]                 [-0.43]               [2.82]                  [-0.44]             [2.59]                  [-0.36]
Used housework help in             -5.54*                 -7.65**               -5.20*                  -7.71**             -4.83*                  -7.17**
last two weeks                    [-1.91]                 [-2.40]              [-1.82]                  [-2.44]            [-1.69]                  [-2.23]
Rural                                5.28                  7.21**                6.43*                   6.34*                5.31                   6.40*
                                   [1.44]                 [2.22]                [1.77]                  [1.91]              [1.45]                   [1.95]
Constant                          12.77***               42.29***              15.31***                44.50***            17.99***                45.39***
                                   [2.92]                 [6.95]                [3.72]                  [8.19]              [4.41]                   [8.63]
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

                  Page Title
Findings: Multivariate analyses continued
•Instead of examining three measures in the multivariate contexts,
examine levels or degrees of core activity limitations among older
couples.
•Trade off: Meaningful, less-broad categories that more accurately map
couples into configurations of limitations, but reduced cell sizes

* Drop the measures of disability or long-term health condition and
perceived self-reported poor health and instead look at:

Configurations of degrees of core activity limitations among older
         couples
         7 categories: contrast group is still neither partner has a disability,
utilized configurations of mild, moderate, severe and profound core activity
limitations. Again test to see if either partner has core limitation and the
other does not, whether the compensatory effect is detected.

.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

                  Page Title
Findings: Multivariate analyses, Table 5
Contrast: couples in which neither partner reports a core activity limitation.
Couples in which only males have:
      mild/moderate core limitations and females have none:
             o She spends about 10 more av. minutes/day on housework
      Severe/profound core limitations and females have none:
             oHe spends nearly 14 fewer av. minutes/day on housework

Couples in which only females have:
      mild/moderate core limitations and males have none:
             o No statistically discernible effects on housework time
      Severe/profound core limitations and males have none:
             oHe spends about 10.4 more av. minutes/day on housework

Couples in which both partners have:
      Moderate/profound core limitations:
             o He spends about 9.5 and she about 12 av. minutes/day on housework
Though statistically insignificant, most other measures in the predicted directions.


.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Table 5: Tobit regressions estimating effects of core limitation configurations on time allocated
to housework among older couples

Measure:                                                     Type of core limitation configuration


Time spent on housework:                                        Husband                Wife
Male partner has mild to moderate limitations, her none           -2.88                9.91*
                                                                 [-0.95]              [1.72]

Male partner has severe to profound limitations, her            -13.73***              4.55
none                                                             [-3.61]              [1.07]

Female partner has mild to moderate limitations, him              6.96                 2.49
none                                                             [1.53]               [0.56]

Female partner has severe to profound limitations, him           10.41**              -5.00
none                                                             [2.01]              [-1.17]

Both have mild limitations                                        2.07                 4.25
                                                                 [0.39]               [0.99]

Both partners have moderate to profound limitations              -9.47**             -11.96**
                                                                 [-2.28]              [-2.49]
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

Post-estimation:Page Title 1.
                Part 1, Figure
From sample on which models were fitted, predicted values for average
minutes/ day in housework were obtained, graphed, and compared with
raw values calculated from times reported in diaries:

Older female partners (left of center of graph):
Net of controls, predicted values for time in housework for
the 3 disability measures when,
       neither has disability,        predicted values ≥ reported values
       male partner has disability, predicted values ≥ reported values
       female partner has disability, predicted values ≤ reported values
       both have disability,          predicted values ≤ reported values




.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
          Figure 1. Couples' Time Allocated to Housework
             By reported & predicted values for measures & Gender

40
35
30
25
20
15
10
5
0
      ReportedDis/hlth      Limit Poor hlth   ReportedDis/hlth    Limit Poor hlth
                Female partner                          Male partner
                                  Neither              Him only
                                  Her only             Both
     Sources: ATUS (2006), ABS.
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

Post-estimation:Page Title 1.
                 Part 1, Figure
Older male partners (right of center of graph):
Net of controls, predicted values for time in housework for
the 3 disability measures when,
       neither has disability,        predicted cf. reported => MIXED?
       male partner has disability, predicted values < reported values
       female partner has disability, predicted cf. reported => MIXED?
       both have disability,          predicted values ≤ reported values

      Predicted values relative to reported values depend on the measure.




.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

                Part I, Title
Post-estimation,Page Figure 2:
Rather than 3 measures, Figure 2 shows predicted values for degrees of
core limitations and compared with values calculated from time diaries:

Older female partners (left of center of graph):
       Highest average daily amount of time spent on housework done by
      females with male partner with disability
       Females with severe to profound disabilities (with partners without
      core limitations) do less housework than any other group of older
      partnered females, except for those females who live with a partner
      with severe to profound core activity limitation




.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
          Figure 2. Couples' Time Allocated to Housework
          By reported & predicted values for core limitations & Gender
45
42
39
36
33
30
27
24
21
18
15
12
 9
 6
 3
 0
           Reported           Predicted       Reported       Predicted
                Female partner                     Male partner
                                  Neither         HeMMherNo
                                  HeSvherNo       SheMMheNo
                                  SheSvheNo       BothMod
                                  BothSv
     Sources: ATUS (2006), ABS.
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

                Part I, Title
Post-estimation,Page Figure 2:
Rather than 3 measures, Figure 2 shows predicted values for degrees of
core limitations and compared with values calculated from time diaries:

Older male partners (left of center of graph):
       Highest average daily amount of time spent on housework done by
      males with female partner with disability
       Older males with severe to profound disabilities (with partners
      without core limitations) do less housework than any other group of
      older partnered males—there is no exception like there was for
      females.
       Whereas females with mild to moderate core activity limitations with
      a husband with none spent more time on housework than couples with
      neither partner having core limitations, males with the same traits
      spent less time on housework than the equivalent comparison group.



.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

                 Page Title
Post-estimation Part II: Marginal effects
McDonald-Moffitt decomposition exploited to yield marginal effects for:
      1.       latent dependent variable
      2.       expected value of the dependent variable conditional on being uncensored
      3.       unconditional expected value of the dependent variable, and,
      4.       probability of being uncensored.
Results in Table 6 indicate that we would expect more time spent
on housework among older couples with neither partner having a
core activity limitation given that other covariates are taken at their
means. A couple in which the male only has a core limitation with
other covariates evaluated at their means would have a decrease
of 5.44 average minutes per day in unconditional expected time
spent on housework and a 3.86 decrease in conditional expected
housework time. Moreover, a typical older couple with no core
limitations has a .086 higher probability of being uncensored than
that of a couple with a male with a core limitation.

.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Table 6: Marginal effects of core limitations and a disability or health condition on older couples’ housework time evaluated at means of other covariates.
                                 Latent variablea         Unconditional              Conditional      Probability of           Marginals
                                                              E(Value)             uncensored value     uncensored            evaluated at:
Core limitation
Husband:
      Male Yes/Female No              -7.503**                 -5.446**                -3.860**           -0.086**               0.132
                                       (-2.26)                  (-2.36)                 (-2.38)            (-2.14)
      Male No/Female Yes              9.968***                 7.922***                5.771***           0.097***               0.122
                                       (2.97)                   (2.85)                   (2.78)            (3.32)
   Both partners have, Yes             -5.060                   -3.703                  -2.627             -0.057                0.054
                                       (-1.04)                  (-1.08)                 (-1.09)            (-0.99)
      Mean predicted time               20.09                    24.41                   32.24               0.75
Wife:
      Male Yes/Female No              8.723***                 7.769***                6.092***           0.057***
                                       (2.77)                   (2.71)                   (2.65)            (3.17)
      Male No/Female Yes               -0.900                   -0.779                  -0.596             -0.007
                                       (-0.28)                  (-0.28)                 (-0.28)            (-0.27)
   Both partners have, Yes             -7.062                   -5.955                  -4.468*            -0.060
                                       (-1.53)                  (-1.58)                 (-1.61)            (-1.37)
      Mean predicted time               31.69                    33.65                   38.73               0.86
Disability or Long-term
Health Condition
Husband:
      Male Yes/Female No               -0.401                   -0.302                  -0.216             0.0004                0.236
                                       (-0.13)                  (-0.13)                 (-0.13)           (0.103)
      Male No/Female Yes               7.000**                  5.464*                   3.948*            0.071**               0.146
                                       (1.99)                   (1.93)                   (1.90)            (2.13)
   Both partners have, Yes              2.240                    1.700                   1.216              0.024                0.360
                                       (0.75)                   (0.75)                   (0.74)            (0.75)
      Mean predicted time               20.06                    24.48                   32.42               0.75
Wife:
      Male Yes/Female No               6.415**                  5.645**                 4.374**            0.045**
                                       (2.17)                   (2.15)                   (2.12)            (2.32)
      Male No/Female Yes               7.086**                  6.275**                 4.892**            0.048**
                                       (2.08)                   (2.05)                   (2.01)            (2.31)
   Both partners have, Yes              2.020                    1.758                   1.349              0.015
                                       (0.71)                   (0.71)                   (0.71)            (0.72)
      Mean predicted time               31.68                    33.67                   38.78               0.86
Observations                             791                      791                     791                791
Source: ATUS, 2006; Notes: standard errors in parentheses; a Latent variable coefficients not robust; discrete change of dichotomous variable from 0 to 1; see
Tables 4 and 5 for listing of other covariates; * p < 0.11, ** p < 0.05, *** p < 0.01.
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

                 Page Title
Post-estimation Part II: Marginal effects
Pervious interpretation is okay, but under a different measure no
such substantive interpretation might happen.
For example, if use same category—male has core limitation
only—but measure is disability or long-term health condition then
results lead to no such discussion.
Even if results are significant under two measures, interpretations
of marginal effects can differ. See for older females for example.




.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?

Conclusions:                                    Page Title
oResults differ by measure so interpretations can differ
depending upon measure and magnitudes of differences
across measures.
oSome evidence here that disabilities do elicit gender neutral
time allocations to housework.
      o        Possible the observed results represent an earlier move towards
               more equal time allocated to housework (household production?)
oStill, even in this case, it seems older females do more
housework and change only occurs when a notable
detrimental physical/mental change occurs in the male
partner of both
oNot such a good result that equality in housework comes
when both disabled!
oLife course for disabled older couples differs from AARP
.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011
Does Disability among Older Couples Elicit Gender-Neutral Time Allocations?


 Caveats:                                            Title
                                                Page Future research:
 Measurement of disability                                              Examine time allocation to other
 Sample sizes                                                           activities, such as:
 Modeling strategy needs further                                            • Employment
 refinement.                                                                 • Volunteering
                                                                             • Recreation & leisure
                                                                             • Socializing
                                                                         Improve modeling:
                                                                                    endogeneity of income
                                                                                    Different metric for dependent variable




.Brandon: Paper prepared for the International Perspectives on Time Use Conference, June 23-24, 2011

				
DOCUMENT INFO