Next Article in Journal
Thermographic of the Microcirculation in Healthy Children Aged 3–10 Months as an Objective and Noninvasive Method of Assessment
Next Article in Special Issue
Multidimensional Measures and the Extra Costs of Disability: How Are They Related?
Previous Article in Journal
Impact of Government Intervention in Response to Coronavirus Disease 2019
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Estimating Households’ Expenditures on Disability in Africa: The Uses and Limitations of the Standard of Living Method

1
Center for Inclusive Policy, Washington, DC 20005, USA
2
School of Social Welfare, State University of New York-Stony Brook, Suffolk County, NY 11794, USA
3
Prospera, Jakarta 12190, Indonesia
4
Syracuse University College of Law, Burton Blatt Institute, Syracuse University, Syracuse, NY 13244, USA
*
Author to whom correspondence should be addressed.
Int. J. Environ. Res. Public Health 2022, 19(23), 16069; https://doi.org/10.3390/ijerph192316069
Submission received: 31 August 2022 / Revised: 4 November 2022 / Accepted: 26 November 2022 / Published: 1 December 2022
(This article belongs to the Special Issue Costs Incurred by People with Disabilities)

Abstract

:
People with disabilities face extra costs of living to participate in the social and economic lives of their communities on an equal basis with people without disabilities. If these extra costs are not accounted for, then their economic wellbeing will be overestimated. The Standard of Living (SOL) method is a way of generating these estimates and is thus useful for determining the economic impact of those costs in the current environment. However, previous studies have used different indicators for disability and different measures of the standard of living, so it is hard to compare estimates across different countries. This study applies a consistent set of indicators across seven African countries to produce comparable estimates. Our estimates of the extra costs of living in these lower-income countries are much lower than the results produced for higher-income countries in prior work. We argue that this finding highlights the limitations of the SOL method as a useful source of information for developing inclusive systems of social protection in lower-income countries because it captures what households spend but not what the person with a disability needs to fully participate in the social and economic lives of their community. In lower-income countries, people with disabilities are likely to have fewer opportunities to spend on needed items thus resulting in substantial unmet need for disability-related goods and services. Failing to account for these unmet needs can lead to inadequate systems of social protection if they are based solely on SOL estimates.

1. Introduction

Disability is correlated with poverty around the world [1], but that relationship is understated if the measure used is simply based on household income or consumption. When using multi-dimensional measures of poverty that relationship is stronger [2,3], especially when those indices account for the kinds of deprivations which are particularly associated with disability, such as social connections and autonomy [4,5]. However, another factor often left out of the analysis is the extra costs associated with living with disability.
People with disabilities face extra costs of living to participate in the social and economic lives of their communities on an equal basis with people without disabilities [6]. These costs include both those specific to people with disabilities, such as assistive devices and personal assistance, but also increased spending on mainstream needs like transportation and medical care. Thus, to have the same level of economic wellbeing as a household without a person with a disability, a household with a person with a disability must have income or consumption levels that are greater by the amount of those extra costs. A household with a member with a disability who cannot to meet all their basic needs at a particular level of income, may have been able to do so if they did not have a member with a disability who requires extra goods and services.
A household with a person with a disability whose consumption is at the poverty line has one of two options—either live below the poverty line and cover some, if not all, of the extra costs needed by their household member, or live at the poverty line (in terms of household items) but deny their household member with a disability those goods and services they need to live full lives. In either case, the person with a disability living in a household at the poverty line is, in essence, living below it. Additionally, even for families above the poverty line, those additional costs are likely to affect economic wellbeing.
When examining the impact of extra costs on the lives of people with disabilities, it is important to keep in mind the distinction between the extra costs needed for equal participation and the extra expenditures that are made by households with people with disabilities [7]. Household’s may not be able spend the full amount needed for equal participation for several reasons.
  • Families may not be able to afford the needed goods and services
  • The goods and services may not be available, especially in low income countries and in rural or remote areas
  • People may be unaware of goods and services that can help them overcome barriers to participation
  • Discrimination can occur within a household, depriving people with disabilities of things they need
Most of the estimates of the extra costs of disability are estimates of extra expenditures, not the extra costs of goods and services required of equal participation. The estimates are useful in looking at the economic impact of disability on people’s lives, but not for determining people’s needs. Still, those estimates show a significant impact. When adjusting poverty lines by estimated extra expenditures on the needs of people with disabilities, the poverty rate in Cambodia rises from 18% to 34%, from 17.6% to 23% in Vietnam, 21.1% to 30.8% in Bosnia and Herzegovina, 32% to 42% in Mongolia and 38.5% to 52.9% in Ghana [8,9,10]. In the United States, adjusting the standard poverty line by disability-related expenditures would raise the estimated poverty rate of persons with disabilities from 24% to 35% and leave 85% of persons with disabilities in the US living below four times the poverty rate [11].
Most of the estimates of extra expenditures associated with disability rely on a methodology known as the Standard of Living Method (SOL), developed by Zaidi and Burchardt [12]. After briefly explaining this methodology, this paper addresses two challenges to this approach.
First, the SOL method relies on establishing a measure of the standard of living, and it is not clear how sensitive the results are to the chosen measure, and to what extent comparisons across countries are due to actual differences in extra expenditures or differences in how SOL is defined. This is not an indictment of the method, only a caution in comparing results that use different indicators for the standard of living—and comparing results between studies that use different indicators for identifying who has a disability.
Second, studies in South Africa and New Zealand, two very different countries, suggest that the gap between actual expenditures and the costs needed for full participation may be high [13,14]. In both cases, the estimated expenditures required for equal participation are very substantial and often significantly in excess many people’s incomes. As a result, it is not clear whether the SOL method, which does measure the current economic impact of disability expenditures on households, yields estimates that are appropriate for informing the design and budgeting for social protection policies and other schemes aimed at promoting full participation in LMIC.
Other limitations of the SOL are also discussed, for example that it only provides average expenditures, even though the South Africa and New Zealand studies show that these expenditures can vary widely across individuals, and the SOL does not provide information on what is being purchased. These two aspects of the SOL measure also limit its usefulness in designing programs to efficiently and adequately address these costs. Nevertheless, the SOL measure does provide information on the current average economic impact on households with members with disabilities, and sheds light on the inadequacy of poverty measures that do not take this into account. Therefore, it is useful in adjusting measures of disability poverty gaps. Moreover, if the sample size is large enough, SOL can be used to examine the difference in disability expenditures according to personal or household characteristics.

2. Standard of Living Method

The basic idea behind the SOL approach is that two families with the same income who are similar in a variety of other ways (e.g., household size, where they live, etc.) are expected to have the same level of wellbeing defined in this study as an asset index. If one of those households has a member with a disability, then any gap in wellbeing is assumed to result from the increased expenditures associated with the needs of the person with disability. In the absence of disability, those expenditures would be used to build up assets but for households with a disability, the expenditures are used to cover disability-related needs.
It is important to note that the SOL method does not address the indirect costs of disability, namely the foregone income of people with disabilities or household members. Those indirect costs result from barriers to employment or the need for household members to forgo paid work in order to provide support. The SOL method is only looking at the direct costs, that is extra expenditures on both disability specific and general items. Comparisons are being made between households that have the same level of income.
The approach developed by Zaidi and Burchardt is shown below in Figure 1. The higher line represents the relationship between income and standard of living for households without members with disabilities, the lower line is for those without members with disabilities. As income increases, the standard of living increases at the same rate for both types of households, but the line for households with members with disabilities is lower by the amount of those extra costs, which are assumed to be fixed. A household with a member with a disability must have an income of “I2” to have the same level of wellbeing as a household without a member with a disability with income “I1”. The line segment “AB” represents the extra cost of disability.
Zaidi and Burchardt formulate the standard of living approach as.
S = αY + βD + γX + k
where S is an indicator of the standard of living, Y is household income, D is the presence of a household member with a disability, and X are other household characteristics. The parameter β is the impact of disability on the standard of living. Zaidi and Burchardt interpret k as the minimum level of standard of living a household needs to survive. The extra cost of disability, E, is given by
E = dY/dD = −β/α
The distance between the lines is CB which is equal to β. The slope of the line is CB/AB equal to α. Thus β/α is CB/(CB/AB) which equals AB, or I2-I1, which is the extra cost of disability.
A common measure of the standard of living in the literature is wealth, represented by an asset index, but other indices could be used. For example, studies have used standard of living indicators based on self-rated financial satisfaction [15,16], the ability to afford different desired goods and services [17,18,19], or subjective assessment of the ability to make ends meet [20,21].
Because of the different definitions of the standard of living, it is unclear how comparable estimates are across countries. Even among studies that have used an asset index, the index has been constructed in different ways. It could be simply an index of the number of assets owned, an index constructed through principal component analysis (PCA), or a polychoric PCA [7,22]. The first of these three methods does not account for the correlation between the ownership of various assets, the second, which identifies a latent underlying variable of assets, does account for this, but the third extends that latent analysis to account for the impact of not owning an asset. For example, if nearly everyone has a particular asset, then having it does not add much information to an index, either based on the number of assets or even a PCA. However, with a polychoric PCA, the fact that one does not have an asset that is owned by most people can contribute to a signal that the person is poor [23].

3. Methodology

The goal of this study was to compare SOL estimates across countries in as comparable a manner as possible. To control for differences in estimates that might come from differences in how SOL is implemented, this study uses countries having data that allow for the similar construction of an asset index, the same covariates, and the same definition of disability. An effort was also made to control for differences that may result because of differences in how people with disabilities are identified across studies.
Survey questions used for identifying people with disabilities can vary significantly, and some types of questions have been shown to produce poor data [24]. The countries selected for this study were chosen because they all use the Washington Group Questions on disability, which have been widely adopted and recommended by many development agencies and international organizations and because they have surveys with data that can be used to construct an asset index in a similar manner. We also focused on lower income countries in one region, Sub-Saharan Africa, to make the results more comparable [25].
The countries with surveys meeting the above criteria were Ethiopia, Tanzania, Liberia, Nigeria, Namibia, Zimbabwe, and Malawi. Descriptive statistics for these countries’ data can be found in Table 1. A description of the survey design, number of observations, response rate, and links to more information about the surveys can be found in Table A1.
In all these surveys, a household was considered to have a member with a disability if at least one household member answered that they had “a lot of difficulty” or “cannot do” to at least one of the six activities in the Washington Group Short Set of Questions. These questions address seeing, hearing, walking or climbing steps, remembering or concentrating, understanding or being understood by others, and self-care (For the exact questions, and documentation on the use and testing of these questions see www.washingtongroup-disability.com (accessed on 29 November 2022)).
The other co-variates used in estimating Equation (1) were the log of consumption, age, age-squared, education of household head, household size, whether the household had health insurance, and whether they lived in an urban area. Regional dummies were also used. Consumption was used, and not income, because that is standard in low income countries where income is more difficult to measure and some households, particularly poor ones, may produce some of what they consume.
The same methodology was used to construct the asset index based on a comparable set of assets across countries, using the tetrachoric PCA command in Stata [26]. Tetrachoric PCA is a kind of polychoric PCA method that is used for estimating the principal component scores for binary variables. The first principal component scores for each country are included in the Appendix A. The mean value of the latent asset index for each country is in Table 1.

4. Results

The regression results for all seven countries are shown in Table 2, which also lists the names and years of the surveys used. In terms of the covariates other than disability, consumption and education were correlated with higher levels of assets across all the countries, as was household size, except for Namibia where it had a small negative, but statistically significant effect. Having health insurance also was associated with more assets in all countries, except for Ethiopia where there was no statistically significant correlation. Initially an additional year of age is correlated with more assets, but eventually, because of the negative coefficient on age-squared, it starts being associated with decreasing assets. These results are generally as expected.
As for disability, the main variable of interest, it is always negatively associated with the asset index, but is only statistically significant in three of the seven countries: Tanzania, Nigeria, and Zimbabwe. As described above, the estimates for the extra expenditures associated with disability are estimated by way of the ratio of the coefficient for disability and the coefficient for log consumption. Thus, for Ethiopia, we estimate that households including a member with a disability require 6% (−0.012/0.212) more income to maintain their living standards relative to a comparable household without a member with a disability. These estimates are lower than is often seen in higher income countries, as discussed below. It ranges from 4% in Zimbabwe to 10% in Tanzania of household consumption and is not statistically significant in Liberia and Namibia.

5. Discussion

The estimates for the extra expenditures associated with disability contrast sharply with estimates in higher income countries, where the extra expenditures average out at about 43%, as show in Table 3. As stated earlier, this could be because the reasons for the gaps between what is spent and what is needed are much stronger in poorer countries. In fact, Table 4 shows that the two poorest countries analyzed in this study did not show significant increased expenditures. Namibia, however, is the exception. It is by far the country with the highest income among the countries studied and shows no association between disability and extra expenditures. Of course, it is still a relatively poor country. The average per capita GDP in the world in 2020 was $10,961, compared to $4179 in Namibia. However, these numbers also contrast with a study done in Ghana, a similar African country with GDP per capita equal to $2206 using a PCA asset index, though not far from another country with a similar level of income, Vietnam ($2656 per capita), at 11% [10,27].
The SOL is a valuable tool to estimate current average economic impact of the extra expenditures needed for people with disabilities to participate. However, it has significant limitations:
First, it does not account for the indirect costs of disability, that is foregone income, only the direct costs incurred by households including members with a disability. Not accounting for these direct costs is, nevertheless, important as it will overestimate the economic wellbeing of households with members with disabilities. Poverty will be underestimated among households with people with disabilities if the SOL method is not used to adjust poverty rates. However, given the wide range of needs—as demonstrated in the South African and New Zealand studies mentioned above—that average hides a great deal of variance. By using the average cost, we will often under or overstate the economic impact of disability on households. Some households have very high costs and some have very low costs. For example, one person may only need a walker, but another might need a personal assistant and a respirator. Moreover, these differences can very well be correlated with the type and degree of disability. Only looking at the average might unintentionally give the impression that a simple top-up of a cash transfer program can adequately address everyone’s needs.
Second, in low income countries where many people are highly income constrained and are more likely to lack knowledge of or access to goods and services needed for participation, disability-related expenditures may not be substantial. Unable to purchase needed items, people with disabilities in such low-resource environments are likely to go with substantial unmet needs for disability-related goods and services. If designers of social protection policies aimed at equalizing economic wellbeing between households with and without disabilities do not take this into account, they could conclude that a relatively minor top-up to a cash transfer for households with members with disabilities is sufficient. This could be incorrect if estimates of what people are currently spending are small, not because they do not need substantial goods and services but because they cannot afford them or they are not currently available. A cash top-up may—on average—equalize resources available for non-disability related expenditures, but it would be woefully insufficient for providing the resources necessary to purchase the goods and services required for full participation in economic and social life.
The large variance in how costs are incurred could call for a set of programs, for example programs targeted at large expenses, like personal assistance or assistive devices, as well as concessions for areas where costs are higher for people with disabilities (such as transportation costs) as well as cash benefits to cover various idiosyncratic costs.
While the SOL can highlight that disability is leading to extra expenditures and so draw attention to these issues, it is important that policymakers go beyond the SOL to measure the types of goods and services people with different types and degrees of disability require so they can match the structure of government programs to the structure of how those costs are incurred. Work to make such estimates, as described in [7] and drawing upon the methodologies in New Zealand and South Africa described elsewhere [13,14], are currently underway in Georgia, Peru, and Tamil Nadu, and will hopefully be available soon. Further research on estimating the types of goods and services need

6. Conclusions

The SOL method is a powerful approach using widely available data to estimate the economic impact of direct disability costs in the current environment. When making cross-country comparisons, it is important to be cognizant of differences in how both the standard of living and disability are defined. This study provides a cross country comparison in one region that uses common indicators.
The SOL method, however, is not well suited for the design of social protection programs aimed at inclusion. Especially in low income countries, SOL estimates can mistakenly be used to draw the conclusion that only minor benefits are needed because in their current environment, for various reasons, not many expenditures occurring. The design of such program requires information on the types of goods and services that people with disabilities would need to participate on an equal basis with their non-disabled peers.

Author Contributions

Conceptualization, D.M.; methodology, D.M., N.G., Z.M. and M.N.; validation, N.G., Z.M. and M.N., formal analysis, N.G., M.N. and Z.M.; resources, D.M.; data curation, N.G., Z.M. and M.N.; writing—original draft preparation, D.M.; writing—review and editing, Z.M., N.G. and M.N.; supervision, D.M.; funding acquisition, D.M. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Namibia Household Income and Expenditure Survey, 2015/2016: https://nsa.org.na/microdata1/index.php/catalog/28 (accessed on 29 November 2022); Nigeria Living Standards Survey 2018–2019: https://microdata.worldbank.org/index.php/catalog/3827 (accessed on 29 November 2022); Ethiopia Socioeconomic Survey 2018/2019: https://microdata.worldbank.org/index.php/catalog/3823 (accessed on 29 November 2022)); Tanzania National Panel Survey 2019/2020: https://microdata.worldbank.org/index.php/catalog/3885 (accessed on 29 November 2022)); Liberia Household Income and Expenditure Survey 2016: https://ghdx.healthdata.org/record/liberia-household-income-and-expenditure-survey-2016-2017 (accessed on 29 November 2022)); Zimbabwe2017 Poverty Income Consumption Survey: https://catalog.ihsn.org/catalog/9250 (accessed on 29 November 2022)); Malawi Fifth Integrated Household Survey 2019–2020 Malawi, 2019–2020: https://microdata.worldbank.org/index.php/catalog/3818 (accessed on 29 November 2022)).

Acknowledgments

We’d like to acknowledge the comments of Alex Cote, Morgon Banks, Jill Hannass-Hancock, Vlad Grigorus, Ludovico Carrera, Sophie Mitra, and Monica Pinilla-Roncancio.

Conflicts of Interest

The authors declare no conflict of interest.

Appendix A

Table A1. Description of surveys used in the research.
Table A1. Description of surveys used in the research.
Name of SurveySurvey Design Response RatesAdditional InformationObs.
EthiopiaEthiopia Socioeconomic Survey 2018/2019Two-stage stratified probability sample providing nationally representative samplePlanned interview = 7527 households from 565 enumeration areas. Actually interviewed: 6770 households from 535 enumeration areas with a response rate of 90 percenthttps://microdata.worldbank.org/index.php/catalog/3823 (accessed on 29 November 2022)5846
TanzaniaThe National Panel Survey 2019/2020Multi-stage clustered sample design providing nationally representative sample.974 of the 989 households had been located and 908 households were successfully re-interviewed for a total household attrition rate of 9.2 percent. https://microdata.worldbank.org/index.php/catalog/3885 (accessed on 29 November 2022))1080
LiberiaHousehold income and expenditure survey (2016)Two-phased clustered sampling methods providing nationally representative sample.n/ahttps://microdata.worldbank.org/index.php/catalog/2986 (accessed on 29 November 2022))8287
NigeriaNigeria Living Standards Survey 2018/2019Random systematic sampling across enumeration areas with limited sample from the state of Borno due to conflict. n/ahttps://microdata.worldbank.org/index.php/catalog/3827 (accessed on 29 November 2022))20,750
NamibiaNamibia Household Income and Expenditure Survey, 2015/2016Random stratified two-stage cluster sample providing a nationally representative sample.10,090 out of 10,368 sampled households were successfully interviewed, resulting in a 97.3 percent response rate.https://nsa.org.na/microdata1/index.php/catalog/28/related-materials (accessed on 29 November 2022))10,090
ZimbabwePoverty, Income, Consumption and Expenditure Survey Questionnaire 2017Stratified two-stage sample design providing nationally representative sample.Of 32,256 sampled households, a total of 31,195 households successfully completed interviews providing a response rate of 96.7 percent.https://catalog.ihsn.org/catalog/9250 (accessed on 29 November 2022)26,431
MalawiIntegrated Household Survey and Integrated Household Panel Survey 2019Stratified two-stage sample design providing nationally representative sample.12,288 households from 768 enumerated areas were selected with 51 areas unable to be visited due to COVID. Final response rate was 93 percent.https://microdata.worldbank.org/index.php/catalog/3818 (accessed on 29 November 2022))11,404
Note: N/A not available. See [32] for more information on these surveys.
Table A2. Asset index for Malawi- Results from tetrachoric Principal Components Analysis.
Table A2. Asset index for Malawi- Results from tetrachoric Principal Components Analysis.
VariableFirst Principal Component Score
Radio with flash drive/micro CD0.1296
Air conditioner0.0882
Bed0.1938
Bicycle0.039
Beer-brewing drum0.0034
Car0.1959
Chair0.1128
Iron (for pressing clothes)0.1906
Clock0.1836
Computer equipment & accessories0.2029
Coffee table (for sitting room)0.1754
Desk0.1307
Cupboard, drawers, bureau0.178
Tape or CD/DVD player; HiFi0.1925
Electric or gas stove; hot plate, cooker0.2099
Fan0.205
Refrigerator0.2194
Generator0.1574
Electric Kettle0.2057
Kerosene/paraffin stove0.024
Lantern (paraffin)0.0184
Lorry0.1465
Mini-bus0.1302
Motorcycle/scooter0.0655
Mortar/pestle 0.0402
Radio (‘wireless’)0.0313
Sewing machine0.1171
Upholstered chair, sofa set0.2029
Solar panel0.0268
Satellite dish0.2114
Table0.1396
Television0.2158
VCR0.1225
Washing machine0.1372
House−0.0837
Toilet0.1098
Wall0.1287
Roof0.1849
Floor0.2
Water0.1646
Electricity0.211
Mobile0.1684
Table A3. Asset index for Liberia—Results from Tetrachoric Principal Components Analysis.
Table A3. Asset index for Liberia—Results from Tetrachoric Principal Components Analysis.
VariableFirst Principal Component Score
Refrigerator0.254
Electric Fan0.253
TV 0.242
Electricity (community generator, own generator, power supplier, solar pane, car motorcycle battery, other)0.242
Car0.232
Satellite0.225
Floor (CONCRETE, CEMENT, TILES, TIMBER/Not earth)0.213
Generator0.211
Flush toilet 0.210
Walls (zinc, iron, tin; stone clay bricks; concrete, cement, blocks, wood or timber, poles)0.207
Computer0.203
Iron0.202
Cupboards0.194
Telephone0.191
Air conditioning0.181
Water heater0.174
Calculator 0.164
Roof (CONCRETE, CEMENT, roofing tiles, iron sheets)0.157
Sofas0.147
Books0.130
Bicycle0.124
Radio0.121
Electric gas stove0.119
Watches0.117
Tables 0.113
Water for handwashing laundry0.106
Chairs0.105
Beds 0.103
Motorcycle0.083
Bus0.083
Sewing machine0.074
Piped water0.067
Lanterns0.057
Cooking equipment0.044
Utensils0.031
Lamp0.018
Own land−0.062
Leather −0.163
Table A4. Asset index for Ethiopia—Results from Tetrachoric Principal Components Analysis.
Table A4. Asset index for Ethiopia—Results from Tetrachoric Principal Components Analysis.
VariableFirst Principal Component Score
house−0.1153
toilet0.1679
wall0.186
roof0.1704
floor0.2138
wash basin0.1707
water0.1599
electricity0.1816
axe−0.1278
bicycle0.1175
blanket0.1008
bstove0.0516
clock0.106
cstove0.1067
dresser0.2075
emitad0.2084
estove0.2173
fridge0.2132
gold0.1541
hncart0.007
kstove0.1043
landline0.0561
mattress bed0.151
motorcycle0.0451
mplogh−0.0836
pick axe−0.1435
private car0.152
radio0.068
sewing0.0459
shelf0.1799
sickle−0.1727
silver0.1376
sofast0.2051
solard−0.0664
sstove0.1467
satellite dish0.2268
television0.2277
plogh−0.1634
videod0.1885
weavng−0.0238
water pump−0.0311
wtstor0.0111
ownland−0.1213
mobilehh0.1691
livestock−0.1929
Table A5. Asset index for Tanzania—Results from Tetrachoric Principal Components Analysis.
Table A5. Asset index for Tanzania—Results from Tetrachoric Principal Components Analysis.
VariableFirst Principal Component Score
toilet0.208
water0.1299
electricity0.2075
walls0.1967
floor0.232
wash basin0.1371
beds0.1963
books0.1313
car0.2023
chairs0.0604
computers0.19
cooking0.0188
cupboards0.1774
dish0.2265
electricity0.2264
handmi~h0.085
iron0.2241
lanterns0.0514
mosquito0.081
motorcycle0.1068
musics~m0.13
others~e0.1551
radio0.1404
refrigerator0.2261
sewing machine0.1147
sofas0.2286
tables0.1567
telephone0.1922
tv0.2549
video0.2248
watches0.1833
waterh~r0.1725
waterpump0.1096
wheelbarrow0.0838
Table A6. Asset index for Nigeria—Results from tetrachoric Principal Components Analysis.
Table A6. Asset index for Nigeria—Results from tetrachoric Principal Components Analysis.
VariableFirst Principal Component Score
Toilet0.1961
Water0.1473
Electricity0.2142
Walls (Stone, bricks, concrete, cement, wood)0.1825
Roof (iron sheets, concrete)0.0292
Floor (cement, concrete, wood, tile, terrazzo, marble)0.1806
Land−0.0539
Livestock−0.1024
Sofa0.163
Chairs0.0923
Table0.1473
Mattress0.1234
Bed0.0559
Mat−0.0884
Sewing Machine0.0718
Gas Cooking0.1791
Electric Stove0.1527
Gas Stove0.1349
Kerosene cooking0.1146
Refrigerator0.195
Freezer0.1887
Airconditioning0.1934
Washing Machine0.1918
Clothes Dryer0.1554
Bicycle−0.0053
Motorbike0.0007
Car0.1718
Generator0.1813
Fan0.2157
Radio0.0543
Recorder0.093
Hi-fi0.1745
Microwave0.1961
Iron0.2032
TV0.2205
Computer0.1832
DVD0.1872
Satellite Dish0.2039
Music0.1323
Inverter0.1241
Chairs0.0899
Smart Phone0.1821
Mobile Phone0.0585
Table A7. Asset Index for Namibia-Results from tetrachoric Principal Components Analysis.
Table A7. Asset Index for Namibia-Results from tetrachoric Principal Components Analysis.
VariableFirst Principal Component Score
Toilet0.2044
Water0.1777
Electricity0.2045
Walls (concrete, cement, stones, baked/burnt bricks0.1872
Roof (cement, brick, corrugated iron/zinc, tiles, slate)0.1112
Floor (concrete, wood, tiles)0.1911
Land−0.007
Livestock−0.0875
Car0.1771
Bus0.1266
Bakkie0.159
Motorbike0.1512
Bicycle0.0949
Electric Stove0.1901
Gas Stove0.0768
Microwave0.2186
Refrigerator0.2184
Freezer0.1808
Washing Machine0.2156
Sewing Machine0.0991
Radio−0.01
Stereo0.1612
TV0.2178
Satellite Dish0.1905
DVD player0.1661
Phone0.1904
Cell Phone0.1508
Computer0.197
Tablet0.1686
Camera0.1772
Generator0.1062
Livingroom Furniture0.212
Bedroom Furniture0.1746
Dining room Furniture0.1871
Donkey/ox−0.0267
Plough−0.0918
Tractor0.1199
Wheelbarrow0.0728
Grindmill0.0803
Boat0.0933
Tent−0.0384
Table A8. Asset index for Zimbabwe—Results from tetrachoric Principal Components Analysis.
Table A8. Asset index for Zimbabwe—Results from tetrachoric Principal Components Analysis.
VariableFirst Principal Component Score
Television0.2779
DVD Player0.2276
Satellite Dish & components0.2741
Computer0.2238
Refrigerator0.2729
Deep–freezer0.2231
Toaster ordinary0.2505
Sandwich Toaster0.2421
Microwave0.2614
Stove (gas or electric)0.2814
Electric Heater0.2393
Lounge suite0.2094
Dining room suite0.1992
Carpets0.2239
own_power0.2457
own_water0.1711
own_sanit0.2684

References

  1. Banks, L.M.; Kuper, H.; Polack, S. Poverty and disability in low- and middle-income countries: A systematic review. PLoS ONE 2017, 12, e0189996. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  2. Mitra, S.; Posarac, A.; Vick, B. Disability and Poverty in Developing Countries: A Multidimensional Study. World Dev. 2013, 41, 1–18. [Google Scholar] [CrossRef]
  3. Pinilla-Roncancio, M. The reality of disability: Multidimensional poverty of people with disability and their families in Latin America. Disabil. Health J. 2018, 11, 398–404. [Google Scholar] [CrossRef] [PubMed]
  4. Trani, J.F.; Bakhshi, P.; Myers Tlapek, S.; Lopez, D.; Gall, F. Disability and poverty in Morocco and Tunisia: A multidimensional approach. J. Hum. Dev. Capab. 2015, 16, 518–548. [Google Scholar] [CrossRef]
  5. Clausen, J.; Barrantes, N. Implementing a Group-Specific Multidimensional Poverty Measure: The Case of Persons with Disabilities in Peru. J. Hum. Dev. Capab. 2020, 21, 355–388. [Google Scholar] [CrossRef]
  6. Mitra, S.; Palmer, M.; Kim, H.; Mont, D.; Groce, N. Extra costs of living with a disability: A review and agenda for research. Disabil. Health J. 2017, 10, 475–484. [Google Scholar] [CrossRef]
  7. Mont, D.; Cote, A.; Hanass-Hancock, J.; Banks, L.M.; Grigorus, V.; Carraro, L.; Morris, Z.; Pinilla-Roncancio, M. Estimating the Extra Costs for Disability for Social Protection Programs. UNPRPD, August 2022. Available online: https://www.social-protection.org/gimi/RessourcePDF.action?id=57850 (accessed on 29 November 2022).
  8. Palmer, M.; Williams, J.; McPake, B. Standard of Living and Disability in Cambodia. J. Dev. Stud. 2019, 55, 2382–2402. [Google Scholar] [CrossRef]
  9. Braithwaite, J.; Mont, D. Disability and poverty: A survey of World Bank poverty assessments and implications. Alter 2009, 3, 219–232. Available online: https://www.adb.org/publications/living-disability-mongolia-progress-toward-inclusion (accessed on 29 November 2022). [CrossRef] [Green Version]
  10. Asuman, D.; Ackah, C.G.; Agyire-Tettey, F. Disability and Household Welfare in Ghana: Costs and Correlates. J. Fam. Econ. Issues 2021, 42, 633–649. [Google Scholar] [CrossRef]
  11. Morris, Z.A.; McGarity, S.V.; Goodman, N.; Zaidi, A. The Extra Costs Associated With Living With a Disability in the United States. J. Disabil. Policy Stud. 2022, 33, 158–167. [Google Scholar] [CrossRef]
  12. Zaidi, A.; Burchardt, T. Comparing incomes when needs differ: Equivalization for the extra costs of disability in the U.K. Rev. Income Wealth 2005, 51, 89–114. [Google Scholar] [CrossRef]
  13. Hanass-Hancock, J.; Nene, S.; Deghaye, N.; Pillay, S. These are not luxuries, it is essential for access to life: Disability related out-of-pocket costs as a driver of economic vulnerability in South Africa. Afr. J. Disabil. 2017, 6, 280. [Google Scholar] [CrossRef] [PubMed]
  14. Wilkinson-Meyers, L.; Brown, P.; McNeill, R.; Patston, P.; Dylan, S.; Baker, R. Estimating the additional cost of disability: Beyond budget standards. Soc. Sci. Med. 2010, 71, 1882–1889. [Google Scholar] [CrossRef] [PubMed]
  15. Nghiem, S.; Khanam, R.; Vu, X.-B.; Tran, B.X. Implicitly Estimating the Cost of Mental Illness in Australia: A Standard-of-Living Approach. Appl. Heal. Econ. Heal. Policy 2020, 18, 261–270. [Google Scholar] [CrossRef] [Green Version]
  16. Vu, B.; Khanam, R.; Rahman, M.; Nghiem, S. The costs of disability in Australia: A hybrid panel-data examination. Health Econ. Rev. 2020, 10, 6. [Google Scholar] [CrossRef] [PubMed]
  17. Li, J.; Brown, L.; La, H.A.; Miranti, R.; Vidyattama, Y. Inequalities in Standards of Living: Evidence for Improved Income Support for People with Disability. National Centre for Economic and Social Monitoring, Institute for Governance and Policy Analysis, 2019. Available online: https://apo.org.au/node/259261 (accessed on 29 November 2022).
  18. Solmi, F.; Melnychuk, M.; Morris, S. The cost of mental and physical health disability in childhood and adolescence to families in the UK: Findings from a repeated cross-sectional survey using propensity score matching. BMJ Open 2018, 8, e018729. [Google Scholar] [CrossRef]
  19. Morciano, M.; Hancock, R.; Pudney, S. Disability Costs and Equivalence Scales in the Older Population in Great Britain. Rev. Income Wealth 2015, 61, 494–514. [Google Scholar] [CrossRef] [Green Version]
  20. Antón, J.I.; Braña, F.J.; de Bustillo, R.M. An analysis of the cost of disability across Europe using the standard of living approach. SERIEs 2016, 7, 281–306. [Google Scholar] [CrossRef] [Green Version]
  21. Morris, Z.A.; Zaidi, A. Estimating the extra costs of disability in European countries: Implications for poverty measurement and disability-related decommodification. J. Eur. Soc. Policy 2020, 30, 339–354. [Google Scholar] [CrossRef] [Green Version]
  22. Filmer, D.; Pritchett, L.H. Estimating wealth effects without expenditure data—Or tears: An application to educational enrollments in states of India. Demography 2001, 38, 115–132. [Google Scholar]
  23. Moser, C.; Felton, A. The construction of an asset index. Poverty Dyn. Interdiscip. Perspect. 2009, 10, 102–127. [Google Scholar]
  24. Mont, D. Measuring Disability Prevalence, Social Protection Discussion Paper; World Bank: Washington, DC, USA, 2007; Volume 706. [Google Scholar]
  25. Groce, N.E.; Mont, D. Counting disability: Emerging consensus on the Washington Group questionnaire. Lancet Glob. Health 2017, 5, e649–e650. [Google Scholar] [CrossRef] [PubMed]
  26. StataCorp. Stata Statistical Software: Release 15; StataCorp LLC: College Station, TX, USA, 2017. [Google Scholar]
  27. Mont, D.; Cuong, N.V. Disability and poverty in Vietnam. World Bank Econ. Rev. 2011, 25, 323–359. [Google Scholar] [CrossRef]
  28. Ozdamar, O.; Giovanis, E.; Samuk, S. Household disability costs and living standards in Turkey: Evidence from a health reform. Int. J. Soc. Econ. 2020, 48. [Google Scholar]
  29. Touchet, A.; Morciano, M. The Disability Price Tag 2019. Technical Report. 2019. Available online: https://www.scope.org.uk/scope/media/files/campaigns/disability-price-tag-technical-report-2019.pdf (accessed on 30 August 2022).
  30. Carraro, L.; Cumpa, M.C. Accounting for different needs when identifying the poor and targeting social assistance. In Proceedings of the IARIW 33rd General Conference, Rotterdam, The Netherlands, 21 August 2014; pp. 24–30. [Google Scholar]
  31. Amin, R.M.; Adros, N.S.M. The Extra Costs of Having a Disability: The Case of IIUM. Intellect. Discourse 2019, 27, 829–854. [Google Scholar]
  32. Mitra, S.; Yap, J. The Disability Data Report. Disability Data Initiative; Fordham Research Consortium on Disability: New York, NY, USA, 2021. [Google Scholar]
Figure 1. Relation Between Standard of Living and Income.
Figure 1. Relation Between Standard of Living and Income.
Ijerph 19 16069 g001
Table 1. Descriptive statistics by country and according to households with and without a person with a disability (unweighted).
Table 1. Descriptive statistics by country and according to households with and without a person with a disability (unweighted).
Disability (Mean)SDNo Disability (Mean)SD
Ethiopia
Consumption (log)7.010.737.230.79
Age of household head54.8716.9741.7514.47
Education of household head (years)8.074.228.844.54
Urban0.500.500.590.49
Household size5.412.354.342.23
Health insurance0.180.380.170.38
Asset Index0.771.281.041.28
Observations 633 6137
Tanzania
Consumption (log)11.330.9011.530.94
Age of household head57.3016.3442.5714.97
Education of household head (years)0.720.450.870.34
Urban1.380.491.440.50
Household size3.421.182.921.06
Health insurance0.080.270.070.26
Asset Index1.751.021.911.08
Observations 234 950
Liberia
Consumption (log)7.651.027.720.95
Age of household head47.8114.7141.7320.67
Education of household head (years)0.140.190.200.26
Urban0.280.450.340.47
Household size3.080.992.760.97
Health insurance0.020.140.020.13
Asset Index0.920.670.920.68
Observations 814 7536
Nigeria
Consumption (log)11.930.6312.110.68
Age of household head55.3317.7047.8815.26
Education of household head (years)0.730.440.820.39
Urban1.760.431.680.46
Household size6.083.785.083.20
Health insurance0.020.140.040.20
Asset Index1.000.991.241.07
Observations 2993 19130
Namibia
Consumption (log)10.880.9811.001.00
Age of household head59.2818.2044.8716.21
Education of household head (years)0.710.460.840.37
Urban1.670.471.530.50
Household size5.753.543.902.73
Health insurance0.020.150.090.29
Asset Index1.111.141.381.23
Observations 1222 8868
Note: SD = Standard deviation. Disability defined as moderate-severe.
Table 2. Regression Results for SOL.
Table 2. Regression Results for SOL.
Dependent Variable: Latent Asset IndexEthiopiaTanzaniaLiberiaNigeriaNamibiaZimbabweMalawi
SurveyEthiopia Socioeconomic Survey 2018/2019The National Panel Survey 2019/2020Household income and expenditure survey (2016)Nigeria Living Standards Survey 2018/2019Namibia Household Income and Expenditure Survey, 2015/2016Poverty, Income, Consumption and Expenditure Survey Questionnaire 2017Integrated Household Survey and Integrated Household Panel Survey 2019
Extra expenditure estimate6%10%Not significant8%Not significant4%Not significant
Person with moderate-severe disability in household−0.012−0.056 **−0.003−0.053 ***−0.007−0.014 ***−0.009
Consumption (log)0.212 ***0.578 ***0.261 ***0.692 ***0.555 ***0.382 ***0.337 ***
Age 0.305 ***0.0080.122 ***0.012 ***0.018 ***0.132 ***0.435 ***
Age squared−0.290 ***0.048−0.094 ***−0.000 ***−0.000 ***−0.126 ***−0.233 ***
Education of household head0.237 ***0.122 ***0.172 ***0.367 ***0.308 ***0.145 ***0.337 ***
Urban−0.598 ***0.1570.267 ***−0.763 ***−0.711 ***0.431 ***0.248 ***
Regional dummiesIncludedIncluded IncludedIncludedIncludedIncludedIncluded
Household size0.035 ***0.219 ***0.253 ***0.072 ***−0.018 ***0.215 ***0.171 ***
Health insurance−0.0050.048 *0.084 ***0.679 ***0.753 ***0.031 ***0.107 ***
Observations 58461080828720,75010,09026,43111,404
* is significant at the 90% confidence level; ** is significant at the 95% confidence level; *** is significant at the 99% confidence level.
Table 3. Extra expenditures as a proportion of household consumption in high income countries.
Table 3. Extra expenditures as a proportion of household consumption in high income countries.
CountryPercentCountryPercent
USA29Ireland41
UK51Iceland77
Switzerland54Hungary16
Spain41Germany35
Slovenia52Greece32
Slovakia25France29
Romania40Finland78
Portugal38Estonia27
Poland16Denmark56
Luxembourg36Czech Republic36
Norway89Cyprus17
Netherlands63Croatia27
Malta53Bulgaria21
Lithuania30Belgium36
Latvia37Austria54
Italy45Australia50
AVERAGE43
Sources: Antón, J. I. et al. (2016). An analysis of the cost of disability across Europe using the standard of living approach. SERIEs, 7(3), 281–306 [20]. Ozdamar et al. (2020) [28]; Touchet, A., & Morciano, M. (2019). The Disability Price Tag 2019. Technical Report [29]; Morris, Z.A. et al. (2022). The extra costs associated with living with a disability in the United States. Journal of Disability Policy Studies, 10442073211043521 [11]; Vu, B. et al. (2020). The costs of disability in Australia: a hybrid panel-data examination. Health Economics Review, 10(1), 1–10 [16]; Palmer, Carraro, L. & Cumpa, M.C. (2014). Accounting for different needs when identifying the poor and targeting social assistance. Paper prepared for the IARIW 33rd General Conference, Rotterdam, 24–30 [30]; Amin, R.M., & Adros, N.S.M. (2019). The Extra Costs of Having a Disability: The Case of IIUM. Intellectual Discourse, 27(SI# 2), 829–854 [31].
Table 4. Estimates of Extra Expenditures by Per Capita GDP.
Table 4. Estimates of Extra Expenditures by Per Capita GDP.
Per Capita GDP in 2020 aEstimate of Extra Expenditures
Liberia632.9Not significant
Malawi636.8Not significant
Ethiopia963.36 percent
Tanzania1076.510 percent
Zimbabwe1214.54 percent
Nigeria2097.18 percent
Namibia4179.3Not significant
a Source: https://data.worldbank.org/indicator/NY.GDP.PCAP.CD (accessed on 29 November 2022).
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Mont, D.; Morris, Z.; Nasiir, M.; Goodman, N. Estimating Households’ Expenditures on Disability in Africa: The Uses and Limitations of the Standard of Living Method. Int. J. Environ. Res. Public Health 2022, 19, 16069. https://doi.org/10.3390/ijerph192316069

AMA Style

Mont D, Morris Z, Nasiir M, Goodman N. Estimating Households’ Expenditures on Disability in Africa: The Uses and Limitations of the Standard of Living Method. International Journal of Environmental Research and Public Health. 2022; 19(23):16069. https://doi.org/10.3390/ijerph192316069

Chicago/Turabian Style

Mont, Daniel, Zachary Morris, Mercoledi Nasiir, and Nanette Goodman. 2022. "Estimating Households’ Expenditures on Disability in Africa: The Uses and Limitations of the Standard of Living Method" International Journal of Environmental Research and Public Health 19, no. 23: 16069. https://doi.org/10.3390/ijerph192316069

APA Style

Mont, D., Morris, Z., Nasiir, M., & Goodman, N. (2022). Estimating Households’ Expenditures on Disability in Africa: The Uses and Limitations of the Standard of Living Method. International Journal of Environmental Research and Public Health, 19(23), 16069. https://doi.org/10.3390/ijerph192316069

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop