MS ritgerð Heilsuhagfræði The effect of children´s disability and … · Leiðbeinandi minn er Tinna Laufey Ásgeirsdóttir, doktor í hagfræði og prófessor við hagfræðideild

MS ritgerð

Heilsuhagfræði

The effect of children´s disability and

long-term illness on parent´s income

Elísabet Eggertsdóttir

Leiðbeinandi Tinna Laufey Ásgeirsdóttir

Hagfræðideild

Júní 2016

The effect of children´s disability and

long-term illness on parent´s income

Elísabet Eggertsdóttir

Lokaverkefni til MS -gráðu í heilsuhagfræði

Leiðbeinandi: Tinna Laufey Ásgeirsdóttir

Hagfræðideild

Félagsvísindasvið Háskóla Íslands

Júní 2016

3

The effect of children´s disability and long-term illness

on parent´s income.

Ritgerð þessi er 30 eininga lokaverkefni til MS prófs við Hagfræðideild,

Félagsvísindasvið Háskóla Íslands.

© 2016 Elísabet Eggertsdóttir

Ritgerðina má ekki afrita nema með leyfi höfundar.

Prentun: Prentsmiðja Háskólaprent ehf

Reykjavík, 2016

4

Formáli

Þessi rannsokn er lokaverkefni til meistaragraðu i heilsuhagfræði við Haskola Islands og

er metin til 30 ETCS eininga. Leiðbeinandi minn er Tinna Laufey Asgeirsdottir, doktor i

hagfræði og professor við hagfræðideild Haskola Islands. Eg vil þakka henni kærlega fyrir

goða leiðsogn.

5

Abstract

Objectives: We aim to study the effects of having a severely ill or disabled child on the

parent´s finances. Firstly, we examine how the onset of a child’s illness or disability affects

the parent´s income. Secondly, we show how income progresses following a period of

caring for an ill or disabled child. The broader objective is to examine to what extent

parents adapt and if social services meet the needs of individuals who face these

circumstances.

Data and Methods: Data from Statistics Iceland and The Icelandic Social Insurance

Administration is merged on personal identification numbers for the purposes of this

study. The treatment group includes beneficiaries who temporarily received home-care

allowance during the period from 1997 to 2011. Those treated are matched with up to

four control observations on gender, age, pre-treatment income level, marital status and

number of dependent children. Generalized least square regression with random effects

is used for the statistical examination.

Results: The treatment group has a lower average total income than the control group in

the time interval of the research. The period of receiving home-care allowance shows the

treated group to have higher income than their counterparts. In the period following the

termination of the allowance, income for the treated group is however lower than for

those untreated.

Conclusion and discussion: Although labour-market income is reduced, parents do not

appear to face a decrease in total income while receiving home-care allowance. However

parents do not adjust well financially after the allowance is terminated. This indicates that

parents might need greater support after a period of taking care of an ill or disabled child.

6

Table of content

Formáli ....................................................................................................................... 4

Abstract ...................................................................................................................... 5

Table of content ......................................................................................................... 6

Pictures ...................................................................................................................... 7

Tables ......................................................................................................................... 8

1 Introduction ......................................................................................................... 9

2 Background ........................................................................................................ 11

2.1 Social services ............................................................................................. 11

2.2 Literature .................................................................................................... 12

3 Data .................................................................................................................... 14

4 Methods ............................................................................................................. 20

5 Results ................................................................................................................ 22

6 Conclusion and discussion ................................................................................. 33

7 References ......................................................................................................... 35

Appendix A ............................................................................................................... 38

Appendix B ............................................................................................................... 40

Appendix C ............................................................................................................... 41

Appendix D ............................................................................................................... 42

Appendix E ............................................................................................................... 46

7

Pictures

Picture 1 Wage index for years 1995-2012…………………………………………………………………….15 Picture 2 Average total income for the research group and the control group

from the original data………………………….........…….……………………………………..………16 Picture 3 Average total income for the treatment group and the control group from the adjusted data…………………………………………………………………………………………………16 Picture 4 Average HCA for different groups of individuals………………………..……………………17 Picture 5 Income during the first 10 years of receiving HCA for all, female and male

(model 2)…………………………………………………………………….………………………………….. 26 Picture 6 Income during the first 10 years after HCA termination for all, female

and male (models 3)………………………............……………………………………………………..26 Picture 7 Income during the first 10 years of receiving HCA for all, partner and

no partner (model 2)……………………………………....……………………………………………….27 Picture 8 Income during the first 10 years after HCA termination for all, partner

and no partner (model 3)…………………………………………………………………………………27 Picture 9 Income during the first 10 years of receiving HCA for child´s different

diagnosis (model 2)…………………………………………………………………………………….……29 Picture 10 Income during the first 10 years after HCA termination for child´s

different diagnosis (model 3)…………………………………………………………………………..29 Picture 11 Income during the first 10 years after HCA termination for different

HCA duration (model 3)………………………………………………………….…………………………30 Picture 12 Income during the first 10 years after HCA termination for child´s different

age at HCA onset (model 3)………………………………………………………………………………31 Picture 13 Income during the first 10 years after HCA termination for child´s

different age at HCA ending (model 3)……………………………………………………………….32 Picture 14 Total number of HCA beneficiaries in December each year in years

2003 to 2013…………………………………………………………………………………………………….38 Picture 15 Annual expenditure on HCA in ISK millions in years 2003 to 2013

(price index 2013)……………………………………………………………………………………………..39

8

Tables

Table 1 Summary statistics ............................................................................................... 19

Table 2 Results for income for all, female, male, partner and no partner (models

1, 2 and 3) ............................................................................................................. 25

Table 3 Results for income for child´s different diagnosis (models 2 and 3) ................... 28

Table 4 Results for income after HCA termination for different HCA duration

(model 3) .............................................................................................................. 30

Table 5 Results for income after HCA termination for child´s different age at HCA

onset (model 3) .................................................................................................... 31

Table 6 Results for income after HCA termination for child´s different age at HCA

ending (model 3) .................................................................................................. 32

Table 7 Monthly Home-care allowance in ISK in January 2015 ....................................... 40

Table 8 Results for income for all, female, male, partner and no partner for the

original data, see table 2 for the adjusted data (models 1, 2 and 3) ................... 42

Table 9 Results for income for child´s different diagnosis for the original data, see

table 3 for the adjusted data (models 2 and 3) .................................................... 43

Table 10 Results for income after HCA termination for different HCA duration for

the original data, see table 4 for the adjusted data (model 3) ............................ 44

Table 11 Results for income after HCA termination for child´s different age at

HCA onset for the original data, see table 5 for the adjusted data (model

3) ........................................................................................................................... 44

Table 12 Results for income after HCA termination for child´s different age at

HCA onset for the original data, see table 6 for the adjusted data (model

3) ........................................................................................................................... 45

9

1 Introduction

When children become severely ill or disabled (SI/D), their condition can have multiple

effects on their own and their family´s lifestyle. The parents may reduce labour supply

and thereby negatively affect the family´s finances. The Icelandic government provides

various types of support to these families, including financial support in the form of direct

tax-free caregiver payments known as home-care allowance (HCA). This allowance is

provided when care is demanding and the parent´s cost of health-care service, treatment

and training has become considerable (Tryggingastofnun 2015a). HCA does not serve as

an income replacement for parents, it is only intended to pay for added expenses. No

information on the samples health-care spending is provided in this research. HCA can be

applied for and is available to those fulfilling certain conditions, evaluated by the Icelandic

Social Insurance Administration (i. Tryggingastofnun Íslands) with a so-called home-care

assessment (Tryggingastofnun, 2015b).

Starting to receive HCA is a turning point indicating that authorities have recognized a

child as being an SI/D. It can be assumed that this support does not start immediately at

the onset of the child´s condition, the preface can be time consuming due to unclear

diagnosis, different evaluations, assessments and application processes. When HCA

terminates it has been concluded that the criteria for the allowance is no longer met. The

payments usually cease due to a child´s recovery, becoming an adult or passing away.

This makes it possible to use the onset and termination of HCA payments as junctures in

parent´s income development. These two turning points are used as breaks on which this

research is built. The Icelandic Social Insurance Administration provided individual level

information on HCA beneficiaries for the current research and Statistics Iceland (i.

Hagstofa Íslands) supplies information in the beneficiarie´s income, before, during and

after the period of receiving HCA.

The aim of this research is to add to the current knowledge on how a family´s finances

are impacted when a child become SI/D and if a potential shock to finances is limited to

the period of receiving HCA or if it is sustained into the future. The reason Iceland is a

feasible country for this project is the access to centralized data, including personal

10

identification numbers. This makes individual-level data mergers possible and provides a

complete data set of all beneficiaries for accurate results. The study has not only

scientific, but also practical value that relates to a service that is of current interest. The

results could potentially give reason to revise the amount of the allowance or to put focus

on alternative measures to support families currently caring for SI/D children and families

who are adjusting to the labour marked after a period of such care. This project will thus

provide informative policy-relevant information to policy-makers and social insurance

administrators, as well as the personnel working directly with these families. It also

enables an opportunity to see where Iceland stands in comparison with other countries

when it comes to supporting families with SI/D children. The research has been granted

permission from The Data Protection Authority (2014111532TS/--).

11

2 Background

2.1 Social services

Since 1997 home-care allowance has been available to parents from the point when

parental leave benefits end until the age of 18. If the child lives at home past that age, it

is possible to extend the allowance until the age of 20 (Tryggingastofnun, 2015e). If a

parent needs to quit paid work when a child becomes SI/D, he or she may be entitled to

joint income related payments with the other parent for up to three months. When it

comes to the death of an SI/D child, parents are entitled to HCA for up to six months after

the child passes away. Another form of benefits for parents with SI/D children are so

called basic payments, which is also a joint entitlement with the other parent. These basic

payments can go on until the child becomes 18 years old (Alþingi, 2015). One more type

of allowance that is intended for people with disabilities is User directed personal

assistance (Notendastýrð persónuleg aðstoð, NPA). These are direct payments provided

by local authorities for parent to hire assistants. The goal of NPA is to supply people with

maximum control over shaping their own lifestyle and provide them with the same

opportunities as non-disabled people (NPA miðstöðin, 2016). It is possible for parents

with disabled children to apply for this type of allowance but very few such contracts have

been made.

The Icelandic Social Insurance Administration provides information on number of

beneficiaries and total expenditure on home-care allowance back to the year 2003

(Tryggingastofnun, 2015c). In years 2003 to 2013 the annual number of beneficiaries was

mostly between 2.000 and 2.200 (Appendix A). The total expenditure on home-care

allowance in the same time interval is also very stable, between ISK 1,5 and 1,8 billion

each year (Appendix A). In January 2015 the amount of the allowance that came with

each child varied from ISK 36.338 to 145.351 depending on the child’s condition

(Tryggingastofnun, 2015d, Appendix B).

This research is based on individuals who received home-care allowance during the

period from September 1st 1997 to December 31st 2011. The reason why the period

starts in 1997 is a change in the terms of the allowance, which was enacted on September

1st (Alþingi, 1997). The period ends in 2011 because Statistic Iceland didn´t have more

recent data on parent´s income when the data was requested.

12

2.2 Literature

Literature on how parent´s income is impacted by children’s SI/D in general is scarce.

Research on how a family´s finances evolve, following a temporary period of caring for a

child with SI/D, is similarly limited. Still considerable research exists on how parent´s

finances are affected by children’s specific diseases or disabilities.

A study on the economic impact on families when a child is diagnosed with cancer,

conducted in Canada and based on semi-structured interviews including specific

questions about the economic effects of cancer, finds that mothers typically terminate or

reduce work hours, which affects the entire family´s financial well-being (Miedema,

Easley, Fortin, Hamilton, Mathews, 2008). It also concludes that this economic burden

can have long-term effects on the financial security of the family, but in particular the

mother. This view is supported by a second study on costs of caring for a child with cancer

in England, which shows that many parents, mainly mothers, give up or reduce labour-

market employment in order to care for their child (Eiser, Upton, 2006). This British study

is based on questionnaires on income, expenditure, employment and financial support.

It concludes that the financial burden associated with caring for an ill child has acute and

long-term implications for parents. Another examination, conducted in Norway using

regression models on data from national registries, comes to a different conclusion as it

finds that parent´s employment is not adversely affected by a child´s cancer (Syse, Larsen,

Tretli, 2011). It is concluded that cancer in children is associated with only minor

reductions in parent´s earnings and found that the reason may be the extensive welfare

options in the Nordic countries in cases of illness in children. It is also found that even

though parent´s employment is unchanged, maternal long-term earnings are slightly

reduced in cases of young children with certain types of cancer, but no significant

reduction was found for fathers.

Only one study was found that specifically focuses on parent´s income following a

period of caring for a child with SI/D and the authors stated that no previous research on

the topic was available. This is an American study on the financial effects for families after

the death of a disabled or chronically ill child (Corden, Sloper, Sainsbury 2002). The

research is based on semi-structured interviews focusing on both immediate and long-

13

term financial impacts. It finds that all parents are affected by loss of or reduction in social

security benefits that can result in a large immediate drop in income. Also that re-entering

the labour market can be a slow process for these parents, especially those who were

unemployed during the period of caring for a child, and that the financial impact can

extend far into the period after death.

It is thus clear that previous empirical studies are limited in number and their focus on

specific diseases and findings have been somewhat mixed. This difference in findings is

not surprising as the economic effects of having children with SI/D are likely to depend

on social and institutional conditions. Thus, multiple examinations are needed that test

this effect under various settings. The final result in such a literature will not be provided

with one study, but with each additional examination adding a piece to the puzzle, as a

fuller picture emerges of cross-country patterns that may shed light on the effectiveness

of policy measures in those countries. This is one such examination.

14

3 Data

The study is based on registry data retrieved from Statistics Iceland, the holder of income-

tax files in Iceland, and The Icelandic Social Insurance Administration which identified

beneficiaries, as well as providing information on them. This data is available on personal

identification numbers assigned to each resident at birth or immigration, allowing for a

merger of the registries at the individual level. The treatment group is based on

individuals who received HCA for at least 12 consecutive months during the period from

September 1. 1997 to December 31. 2011. Only those in the treatment group, who

received no allowance in at least the last 24 months before or 12 months after the

interval, are included in the study. This makes it possible to examine whether or not

income alters at the beginning or the end of a period of receiving the allowance. Any

rupture in an interval that lasts for 3 months or less is not considered as ceased allowance.

A control group, of up to four individuals for each person in the treatment group, was

drawn from the tax register and matched on gender, age, marital status, number of

dependent children and average total annual income in one of the last two years before

the onset of the allowance.

The demographic data available are the beneficiary´s gender, marital status and

annual income. If a registered beneficiary is jointly taxed with a spouse, the spouse is

considered a beneficiary as well and included in the sample by using the average of the

couple´s earnings for the income variable. Demographic data on the children, who´s

illness or disability the HCA in based on, is the child´s age at onset and end of the

allowance (0-20 years), the year of onset and ending of each allowance-interval and

annual total amount of allowance. Also there is data on the child´s diagnosis, which are

classified as severely ill, disabled, or having disorders (Tryggingastofnun (c), 2013).

Since we only have information on the beneficiary´s annual income by calendar year,

every year containing any HCA is considered an allowance-year. The beneficiary´s annual

income during the allowance-years and the years after the allowance termination,

depending on available data, is compared to the assigned control group. This comparison

is achieved by regression analysis that control for trends over time.

15

Picture 1 Wage index for years 1995-2012

The treatment group includes 2.716 individuals who received HCA for a minimum of

one and a maximum of fifteen consecutive years. The control group includes 10.864

individuals, drawn from the Icelandic income-tax registry, who did not receive HCA. The

total income values were adjusted to the year 2012, using the wage index (Hagstofa

Íslands 2016). The wage index for years 1995 to 2012 is shown in Picture 1.

In cases where beneficiaries showed zero income on tax returns, yet showing that they

did receive HCA, their HCA was imputed as income that year. This was the case for 670

years out of a total of 12.924 allowance years. When all income figures have been

adjusted to the year 2012, in accordance with the wage index, the average yearly total

income including the allowance is ISK 4.922.735 for the treatment group and ISK

5.098.491 for the control group, or a 3,6% difference. With a simple regression we find

that the treatment group has a lower income then the control group, the coefficient is

ISK -308 thousand for the treatment group and the p-value is less than 0,001. When

Comparing the two group´s original data on income on a yearly basis, without attention

to the timing of the allowance, shows that the greatest difference can be found in the

16

year 2007, which is the height of an economic boom in Iceland and the year before the

sudden onset of an economic crisis. In that year the difference reaches 14,6% as can be

seen in Picture 2.

For the statistical research the data is adjusted. HCA is separated from beneficiary´s

other income. Since allowance years that showed zero before have been corrected,

beneficiaries do not have a negative income in any cases. If parents are jointly taxed with

a partner, the average of the couple’s earnings is calculated after the HCA has been

Picture 3 Average total income for the treatment group and the control group from the adjusted data

Picture 2 Average total income for the research group and the control group from the original data

17

retracted from the beneficiary´s total income. Outliers are then removed from the

sample. Allowance years that reach an annual income of ISK 100 million are not included,

this takes out a total of 115 observations. Also all allowance years showing an income of

zero are removed, this is a total of 12.260 observations. In total, by excluding outliers,

12.375 observations are removed and the total goes from 235.271 to 222.896. The

number of individuals in the treatment group stays the same, however the control group

is reduced by 19 individuals. Parent´s income under these new conditions is shown in

Picture 3 where the groups are now called adjusted groups. The new averages are higher,

ISK 5.003.591 for the treatment group and ISK 5.210.404 for the control group, the

difference is 4,1%. With a simple regression we find that the treatment group has a lower

income then the control group, the coefficient is ISK -201 thousand for the treatment

group and the p-value is less than 0,001.The greatest different between the groups is also

in the year 2007 in this case although only reaching 7,9%.

The average yearly amount of HCA is ISK 432.021. Picture 4 shows how HCA amounts

are divided between different groups of individuals in the treatment group. When

stratified by gender the average for males is higher. The average for those who are jointly

taxed with a partner is higher than for those without a partner. When looking into the

child’s diagnosis it can be seen that HCA is on average highest for parents of children with

disabilities, thereafter are children with illnesses and the lowest benefits are received by

those who have children have disorders.

Picture 4 Average HCA for different groups of individuals

18

The dependant variable in the statistical models is Income in all cases, which is on

average ISK 5.182.477 per year. Summary statistics can be found in Table 1. Year is one

of the independent variables that are used in all the models and the observations are

rather evenly distributed between the years. TreatmentGroup is a binary variable that

divides the observations into two groups, the treatment group and the control group. As

expected the control group is about four times larger than the research group. The

Female variable shows us that a great majority of the beneficiaries are women, so they

seem to be more likely than men to choose to stay at home when a child becomes SI/D.

The Partner variable indicates that most of the beneficiaries are jointly taxed with a

spouse. No number of individuals is given for the partner groups because in many cases

there are changes in the beneficiary’s marital status during the allowance period. Then

there are a number of binary variables that divide the treatment group into smaller

groups. Three of them are Disability, Disorder and Illness, which are the child´s different

diagnosis. Another three are different allowance period durations named Y1-5HCA, Y6-

10HCA and Y11-15HCA. Then there are OnAge1-5, OnAge6-12 and OnAge13-20 that

divide the treatment group into three subgroups depending on the age of the child at the

onset of HCA. The same is done for the child´s age at the ending of the allowance, those

variables are called EndAge1-5, EndAge6-12 and EndAge13-20. Next is the duration

variables, which group together all observations for the first year of receiving HCA, then

the second year end so on. There are ten of them, starting with Dur1HCA and ending with

Dur10HCA. At last there are the ending variables, which group together all observations

for the first year after HCA has terminated, then the second year and so on, starting with

End1HCA and ending with End10HCA. The duration and ending variables have many

observations behind them during the first years, however they get fewer as the years go

by and the lowest value goes down to 170 observations for the Dur10HCA. This is the

reason why only ten years were chosen to be examined. For this study STATA 13 software

was used for all statistical analysis of the data.

19

Binary variables Observations Individuals Binary variables Observations Individuals

Year 222.896 222.896 HCA duration 44.969 2.716

1995 11.789 11.789 Y1-5HCA 33.315 2023

1996 11.979 11.979 Y6-10HCA 9.851 589

1997 12.118 12.118 Y11-15HCA 1.803 104

1998 12.071 12.071

1999 12.082 12.082 Age at HCA onset 44.969 2.716

2000 12.268 12.268 OnAge0-5Y 13.247 818

2001 12.365 12.365 OnAge6-12Y 20.080 1.205

2002 12.390 12.390 OnAge13-20Y 11.642 693

2003 12.611 12.611

2004 12.444 12.444 Age at HCA ending 44.969 2.716

2005 12.482 12.482 EnAge0-5Y 5.694 362

2006 12.563 12.563 EnAge6-12Y 13.788 862

2007 12.539 12.539 EnAge13-20Y 25.487 1.492

2008 12.654 12.654

2009 12.735 12.735 During allowance years

2010 12.667 12.667 Dur1HCA 2.844 2.844

2011 12.592 12.592 Dur2HCA 2.646 2.646

2012 12.547 12.547 Dur3HCA 2.000 2.000

Dur4HCA 1.391 1.391

TreatGroup 222.896 13.561 Dur5HCA 940 940

1=treatment group 44.969 2.716 Dur6HCA 678 678

0=control group 177.927 10.845 Dur7HCA 496 496

Dur8HCA 340 340

Female 222.896 13.561 Dur9HCA 237 237

1=female 182.288 11.083 Dur10HCA 170 170

0=male 40.608 2.478

After allowance years

Partner 222.896 - Aft1HCA 2798 2798

1=partner 144.778 - Aft2HCA 2365 2365

0=no partner 78.118 - Aft3HCA 2023 2023

Aft4HCA 1712 1712

Childs diagnosis 43.688 2.716 Aft5HCA 1418 1418

Disability 6.224 394 Aft6HCA 1196 1196

Disorder 23.586 1.465 Aft7HCA 966 966

Illness 13.878 857 Aft8HCA 782 782

Aft9HCA 584 584

Aft10HCA 416 416

Table 1 Summary statistics

20

4 Methods

The data is stacked time-series panel data. The foundations of linear panel model

estimation are general linear models. One such model is the Generalized least square

(GLS) estimator, which suits well for time-series as it exploits correlation structure and

adjusts for standard error with cluster consistent matrix estimators. The Random effects

(RE) model allows for individual effects and controls for unobserved heterogeneity. It

suits well when including time-invariant variables like gender and marital status

(Hamilton, 2006). According to a Hausman specification test the RE is the correct model

to be used.

Three models are estimated in this research and the dependant variable is parent´s

income change in all cases. The income variable is the beneficiaries total annual income

before tax without HCA. All models are RE, GLS models of the following form:

Yit = βXit + α + ui + εit

Where Yit represents Income, the dependant variable, for individual i at time t, which in

our case is the variable Year. Xit is a vector of independent variables, β is a vector of

coefficients, α is the unknown intercept for each entity or cluster, ui is the between-entity

error and εit is the within-entity error (Hamilton, 2006).

The first model describes the data for the whole time interval. The independent variable

of main interest is TreatmentGroup, with controls for Year, Partner and Female and

results are given for the full sample as well as four different strata; all, female, male,

partner and no partner. Following is the first model:

Incomeit = β1ResGroupit + β2Yearit + β3Partnerit + β4Femaleit +α + ui + εit (Model 1)

For the second model, ten independent variables have been added to the first model,

the first year of receiving HCA, the second year of receiving HCA and so on up to ten

years. This model examines how each year of receiving HCA is related to income.

Results are shown for eight different groups; all, female, male, partner, no partner and

the three different child’s diagnosis. Following is the second model:

Incomeit = β1ResGroupit + β2Yearit + β3Partnerit + β4Femaleit + β5Dur1HCAit + β6Dur2HCAit

21

+ β7Dur3HCAit + β8Dur4HCAit + β9Dur5HCAit + β10Dur6HCAit + β11Dur7HCAit

+ β12Dur8HCAit + β13Dur9HCAit + β14Dur10HCAit +α + ui + εit (Model 2)

For the third model, other ten independent variables have been added to the first model,

the first year after receiving HCA, the second year after receiving HCA and so on up to ten

years. This model examines how each year after HCA is terminated is related to income.

Results are shown for seventeen different groups; all, female, male, partner, no partner,

the three different child’s diagnosis, the three different HCA durations, the three different

age groups for HCA onset and the three different age groups for HCA ending. Following is

the third model:

Incomeit = β1ResGroupit + β2Yearit + β3Partnerit + β4Femaleit + β5Aft1HCAit + β6Aft2HCAit

+ β7Aft3HCAit + β8Aft4HCAit + β9Aft5HCAit + β10Aft6HCAit + β11Aft7HCAit

+ β12Aft8HCAit + β13Aft9HCAit + β14Aft10HCAit +α + ui + εit (Model 3)

All model estimations are shown in Tables 2 to 6 and when using models 2 and 3, results

are presented in line charts with regards to time span in Pictures 5 to 13. The Pictures all

have an income scale covering ISK 3,5 million. Even though some of the results would do

not need such a large scale, having the same scale for all of them makes them easier to

compare. In Appendix D Tables 8 to 12 show all model estimations when using the original

data.

22

5 Results

For all statistical results, standard error is adjusted for each entity, or clusters that are

associated with each individual. All regressions are robust to control for

heteroskedasticity and the results show that residuals are not correlated with the

independant variable. Wooldridge tests were performed on each model and all

concluded that there was no first-order autocorrelation (Stata, 2013). By using Year as an

independant variable in all the models the time trend is taken into account.

Table 2 includes results for fifteen regressions, the first five using Model 1, next five

using Model 2 and the last five using Model 3. The first regression includes all individuals

and the results are all statistically significant (p-value is less than 0,1). They show that the

treatment group has a lower average total income than the control group in the time

interval of the research. The positive coefficient for the Year variable tells that income

increases with time. This rise is not due to wage inflation, which has already been

adjusted for, it is due to increasing age and work experience. The Year variable has a

positive coefficient throughout all the results of this research. The positive coefficient for

the Partner variable indicates that being jointly taxed with a spouse is positively

associated with total income. The negative coefficient for the Female variable indicates

that women have a lower average total income than men. When the sample is divided

into groups of men and women, all but one of the coefficients are statistically significant.

The results show that women in the treatment group have a lower average income than

men in the treatment group. Also that being jointly taxed with a spouse has a higher

positive coefficient for women than men. When comparing individuals with a partner and

without a partner, all but one of the results are statistically significant. The results show

that beneficieries in the treatment group who are jointly taxed with a spouse have a lower

average income then those without a spouse. Yet being female has a higher negative

coefficient when they are without a partner then when they are with a partner.

The next five regressions in Table 2 use Model 2 and the results show how receiving

HCA is associated with total income for the five different groups. Four of the independent

variables are the same as in Model 1 and will not be interpreted further. Ten new

variables have been added that represent each allowance year. Most of the coefficients

for the years are statistically significant, nine of the fifty have a p-value that is higher than

23

0,1. Without exception, all coefficients are positive. Picture 7 shows a linear trend line for

three of the groups, all, female and male, all af which are inclined upwards, showing the

most positive trend for men. Picture 9 shows a linear trend line for all, partner and no

partner, here all trend lines are similarly inclined upwards.

The last five regressions in Table 2 use Model 3 and the results show how terminating

HCA is associated with total income for the five different groups. Again the variables from

Model 1 will not be interpreted. The ten new variables that have been added this time

represent each year after HCA has terminated. Again most of the coefficients for the years

are statistically significant, thirteen of the fifty have a p-value that is higher than 0,1. All

but three coefficients are negative, those that are positive all have a p-value that is higher

than 0,1. In the results for men, only three out of ten coefficients have p-values over 0,1

and therefore the results for this group should be interpreted carefully. Picture 6 shows

the linear trend for all, female and male, the trend lines are all declined downwards and

the most negative trend is for men. Picture 8 shows the linear trend for all, partner and

no partner which also have a negative trend and the most negative trend is for the no

partner group.

In Table 3, the sample is divided into three groups, disability, disorder and illness,

depending on the child´s diagnosis. Results are given for six regressions, the first three

using Model 2 and the last three using Model 3. For the first three regressions, seventeen

out of thirty coefficients are statistically significant. Twenty nine coefficients are positive

and one is negative, the only negative one has a p-value above 0,1. Picture 9 shows linear

trend lines for income during the first 10 years of receiving HCA for childs different

diagnosis. The trend line for disability is slightly inclined downwards, for disorder the

trend line is slightly inclined upwards and for illness the trend has the most positive

incline.

The last three regressions in Table 3 use Model 3 and the results present how

terminating HCA is associated with total income for the three different groups. Here half

of the coefficients are statistically significant, or fifteen out of thirty, all but one

coefficient are negative and the positive one has a p-value higher than 0,1. Picture 10

shows linear trend lines for income during the first ten years after HCA termination for

child´s different diagnosis. This time the trend line for disability is inclined upwards

24

showing an upward trend, but both disorder and illness are inclined downwards showing

a negative trend.

Table 4 shows results for income after HCA termination for three groups with different

periods of HCA duration, using Model 3. Since the data only cover fifteen allowance years,

no data is available for the last five years for beneficiaries belonging to the group for 11-

15 years of HCA duration. For this reason five results have been omitted. Eighteen out of

twenty five coefficients are statistically significant and they are all negative. The seven

coefficients that have a p-value over 0,1 are positive in four cases and negative in three

cases. For the group with HCA duration of 1 to 5 years, only three out of ten coefficients

have p-values over 0,1 and therefore the results for this group should be interpreted

carefully. Picture 11 presents linear trend lines for each of the three groups, all trend lines

are inclined downwards and the group representing individuals with HCA duration of 6 to

10 years has the most negative trend.

Table 5 shows results for income after HCA termination for three groups with

children´s different age at HCA onset, using Model 3. Seventeen out of thirty coefficients

are statistically significant, fifteen of which are negative. For the group with children in

the age range 1 to 5 at HCA onset, only three out of ten coefficients have p-values over

0,1 and therefore the results for this group should be interpreted carefully. Picture 12

shows the linear trend lines for the groups, all af which are inclined downwards and the

group with children in the age range 6 to 12 at HCA onset has the most negative trend.

The group with children in the age range 13-20 at HCA onset is however the group with

the lowest total average income.

Table 6 shows results for income after HCA termination for three groups with chilren´s

different age at HCA ending, using Model 3. Sixteen out of thirty coefficients are

statistically significant, elleven of which are negative. For the group with children in the

age range 6 to 12 at HCA ending, none of the ten coefficients have p-values over 0,1 and

therefore the results for this group should be interpreted carefully. Picture 13 shows the

linear trend lines for the groups, all af which are inclined downwards and the group with

children in the age range 6 to 12 at HCA ending has the most negative trend. The group

with children in the age range 13-20 at HCA ending is however the group with the lowest

total average income.

25

GLS RE model

Dep. Variable:

Income Coef P>/z/ Coef P>|z| Coef P>|z| Coef P>|z| Coef P>|z|

TreatGroup -154 0,001*** -179 0,000*** -34 0,823 -296 0,000*** 74 0,168

Year 121 0,000*** 122 0,000*** 119 0,000*** 124 0,000*** 107 0,000***

Partner 1146 0,000*** 1317 0,000*** 279 0,001*** - - - -

Female -469 0,000*** - - - - -219 0,001*** -1179 0,000***

During HCA, years 1 to 10

Dur1HCA 95 0,036** 89 0,038** 176 0,274 -23 0,699 285 0,000***

Dur2HCA 234 0,000*** 206 0,000*** 404 0,032* 84 0,218 440 0,000***

Dur3HCA 301 0,000*** 268 0,000*** 475 0,023** 148 0,048** 490 0,000***

Dur4HCA 479 0,000*** 468 0,000*** 560 0,026** 297 0,003*** 668 0,000***

Dur5HCA 441 0,000*** 481 0,000*** 263 0,322 403 0,009*** 461 0,000***

Dur6HCA 280 0,001*** 299 0,001*** 197 0,351 136 0,190 407 0,000***

Dur7HCA 558 0,000*** 456 0,000*** 1235 0,089* 491 0,025** 534 0,000***

Dur8HCA 490 0,001*** 473 0,003*** 614 0,029** 372 0,072* 588 0,000***

Dur9HCA 482 0,001*** 393 0,009*** 1008 0,038** 275 0,070* 825 0,004***

Dur10HCA 276 0,085* 161 0,329 1020 0,047** 238 0,262 334 0,152

TreatGroup -229 0,000*** -250 0,000*** -133 0,349 -333 0,000*** -28 0,595

Year 120 0,000*** 121 0,000*** 118 0,000*** 124 0,000*** 107 0,000***

Partner 1146 0,000*** 1317 0,000*** 275 0,001*** - - - -

Female -470 0,000*** - - - - -219 0,001*** -1183 0,000***

After HCA, years 1 to 10

Aft1HCA -97 0,079* -71 0,245 -262 0,035** -98 0,208 -116 0,060*

Aft2HCA -164 0,002*** -184 0,001*** -141 0,352 -148 0,035** -199 0,006***

Aft3HCA -150 0,019** -172 0,012** -129 0,454 -120 0,152 -215 0,011**

Aft4HCA -59 0,582 -175 0,079* 406 0,304 -38 0,765 -123 0,512

Aft5HCA -281 0,002*** -417 0,000*** 254 0,463 -244 0,061* -432 0,000***

Aft6HCA -374 0,000*** -471 0,000*** 11 0,971 -43 0,000*** -370 0,070*

Aft7HCA -315 0,008*** -368 0,005*** -146 0,604 -302 0,076* -484 0,000***

Aft8HCA -341 0,050** -353 0,077* -342 0,307 -260 0,307 -674 0,000***

Aft9HCA -649 0,000*** -591 0,001*** -922 0,000*** -570 0,013** -930 0,000***

Aft10HCA -806 0,001*** -708 0,019** -1175 0,000*** -773 0,036** -1015 0,000***

TreatGroup -84 0,065* -98 0,029** 5 0,970 -234 0,000*** 169 0,002***

Year 124 0,000*** 125 0,000*** 121 0,000*** 128 0,000*** 112 0,000***

Partner 1141 0,000*** 1313 0,000*** 274 0,001*** - - - -

Female -470 0,000*** - - - - -218 0,001*** -1181 0,000***

Note: All coefficients are in thousands of ISK. P-value: * = P<0,1, ** = P<0,05, *** = P<0,01.

No partnerPartner

Obs. 222.896 Obs. 182.288 Obs. 40.608

All Female Male

Obs. 144.778 Obs. 78.118

Table 2 Results for income for all, female, male, partner and no partner (models 1, 2 and 3)

26

Picture 5 Coefficients for the first 10 years of receiving HCA for all, female and male (model 2)

Picture 6 Coefficients for the first 10 years after HCA termination for all, female and male (models 3)

27

Picture 7 Coefficients for the first 10 years of receiving HCA for all, partner and no partner (model 2)

Picture 8 Coefficients for the first 10 years after HCA termination for all, partner and no partner (model 3)

28

Table 3 Results for income for child´s different diagnosis (models 2 and 3)

GLS RE model

Dep. Variable:

Income Coef P>|z| Coef P>|z| Coef P>|z|


Dur1HCA 201 0,233 172 0,001*** -108 0,116

Dur2HCA 275 0,082* 292 0,000*** 87 0,345

Dur3HCA 393 0,045** 315 0,000*** 157 0,117

Dur4HCA 240 0,086* 476 0,000*** 594 0,009***

Dur5HCA 116 0,530 468 0,000*** 670 0,051*

Dur6HCA 248 0,281 324 0,001*** 149 0,297

Dur7HCA 730 0,053* 397 0,002*** 713 0,101

Dur8HCA 363 0,140 506 0,030** 564 0,008***

Dur9HCA 69 0,758 418 0,034** 1118 0,003***

Dur10HCA 144 0,681 279 0,182 394 0,180

TreatGroup -456 0,000*** -247 0,000*** -55 0,490

Year 122 0,000*** 121 0,000*** 123 0,000***

Partner 121 0,000*** 1150 0,000*** 1220 0,000***

Female -435 0,000*** -450 0,000*** -462 0,000***


Aft1HCA -352 0,002*** -23 0,736 -50 0,645

Aft2HCA -310 0,037** -149 0,015** -89 0,370

Aft3HCA -397 0,011** -123 0,122 -109 0,311

Aft4HCA 255 0,514 -205 0,045** 176 0,424

Aft5HCA -12 0,972 -289 0,009*** -159 0,428

Aft6HCA -248 0,458 -454 0,000*** -48 0,842

Aft7HCA -91 0,814 -301 0,067* -233 0,168

Aft8HCA -63 0,886 -503 0,004*** 235 0,544

Aft9HCA -600 0,041** -692 0,000*** -442 0,037**

Aft10HCA -752 0,003*** -786 0,027** -704 0,000***

TreatGroup -305 0,004*** -800 0,143 13 0,860

Year 123 0,000*** 124 0,000*** 123 0,000***

Partner 1210 0,000*** 1144 0,000*** 1219 0,000***

Female -434 0,000*** -451 0,000*** -462 0,000***


Disorder

Obs. 201.513

Illness

Obs. 191.805

Child´s diagnosis

Obs. 184.151

Disability

29

Picture 9 Coefficients for the first 10 years of receiving HCA for child´s different diagnosis (model 2)

Picture 10 Coefficients for the first 10 years after HCA termination for child´s different diagnosis (model 3)

30

Table 4 Results for income after HCA termination for different HCA duration (model 3)

Picture 11 Coefficients for the first 10 years after HCA termination for different HCA duration (model 3)

GLS RE model

Dep. Variable:



Aft1HCA 9 0,893 -344 0,000*** -632 0,000***

Aft2HCA -111 0,060* -227 0,081* -819 0,000***

Aft3HCA -32 0,672 -468 0,000*** -936 0,002***

Aft4HCA 98 0,451 -523 0,001*** -1477 0,000***

Aft5HCA 133 0,205 -814 0,000*** -681 0,000***

Aft6HCA 176 0,147 -1213 0,000*** (omit) -

Aft7HCA -137 0,302 -1254 0,000*** (omit) -

Aft8HCA -193 0,315 -1190 0,000*** (omit) -

Aft9HCA -562 0,000*** -1062 0,001*** (omit) -

Aft10HCA -695 0,005*** -2660 0,000*** (omit) -

TreatGroup -140 0,004*** 192 0,838 279 0,304

Year 123 0,000*** 124 0,000*** 124 0,000***

Partner 1153 0,000*** 1202 0,000*** 1216 0,000***

Female -456 0,000*** -437 0,000*** -430 0,000***


(omit) is omitted because of collinearity.

Obs. 211.242 Obs. 187.778 Obs. 179.730

1-5 years 6-10 years 11-15 years

HCA duration

31

Table 5 Results for income after HCA termination for child´s different age at HCA onset (model 3)

GLS RE model

Dep. Variable:



Aft1HCA 150 0,196 -106 0,161 -360 0,000***

Aft2HCA 201 0,074* -140 0,049** -617 0,000***

Aft3HCA 151 0,143 -120 0,181 -530 0,000***

Aft4HCA 345 0,073* -158 0,252 -311 0,229

Aft5HCA 105 0,516 -299 0,039** -640 0,000***

Aft6HCA -85 0,510 -451 0,001*** -537 0,056*

Aft7HCA 245 0,164 -329 0,124 -884 0,000***

Aft8HCA 591 0,178 -458 0,040** -1178 0,000***

Aft9HCA -294 0,090* -549 0,041** -1220 0,000***

Aft10HCA -218 0,203 -1133 0,000*** -867 0,413

TreatGroup -603 0,000*** 11 0,856 360 0,000***

Year 123 0,000*** 124 0,000*** 123 0,000***

Partner 1209 0,000*** 1161 0,000*** 1191 0,000***

Female -456 0,000*** -453 0,000*** -413 0,000***


Child´s age at HCA onset

Age 13-20

Obs 189.569

Age 0-5 Age 6-12

Obs 191.174 Obs 198.007

Picture 12 Coefficients for the first 10 years after HCA termination for child´s different age at HCA onset (model 3)

32

Table 6 Results for income after HCA termination for child´s different age at HCA ending (model 3)

Picture 13 Coefficients for the first 10 years after HCA termination for child´s different age at HCA ending (model 3)

GLS RE model

Dep. Variable:



Aft1HCA 319 0,203 75 0,445 -271 0,000***

Aft2HCA 422 0,028** 45 0,587 -398 0,000***

Aft3HCA 317 0,065* 148 0,159 -412 0,000***

Aft4HCA 713 0,051* 111 0,462 -331 0,027**

Aft5HCA 314 0,285 7 0,953 -587 0,000***

Aft6HCA 80 0,666 -74 0,561 -657 0,000***

Aft7HCA 504 0,063* 145 0,542 -854 0,000***

Aft8HCA 1314 0,098* -250 0,143 -889 0,000***

Aft9HCA -144 0,572 -376 0,154 -976 0,000***

Aft10HCA 52 0,820 -772 0,000*** -1110 0,054*

TreatGroup -623 0,000*** -466 0,000*** 245 0,000***

Year 124 0,000*** 124 0,000*** 124 0,000***

Partner 1225 0,000*** 1179 0,000*** 1157 0,000***

Female -444 0,000*** -441 0,000*** -441 0,000***


Obs 183.621 Obs 191.715 Obs 203.414

Age 0-5 Age 6-12 Age 13-20

Child´s age at HCA ending

33

6 Conclusion and discussion

The current research has examined the effect of having SI/D children on parents‘ income.

The main results indicate that the treatment group has a lower average total income than

the control group in the time interval of the research. Having a partner is positively

associated with income and being female is negatively associated with income. These

results are all in accordance with what was expected.

When examining how the onset of a child´s S/ID affects parent´s income, the results

show that each year of receiving HCA is positively associated with total income. This

suggests that parent´s finances are positively affected while caring for an S/ID child,

indicating that labour supply is increased. This is unexpected and not in accordance with

the background literature for this research. One study finds that mothers typically

terminate or reduce work (Miedemaq, Easley, Fortin, Hamilton, Mathews, 2008), another

concludes that parents give up or reduce labour-market employment in order to care for

their child (Eiser, Upton, 2006). Another research finds that parent´s employment is not

adversely affected, and that the reason may be the extensive welfare options (Syse,

Larsen, Tretli, 2011). The expected conclusion for this study would have been that parents

either kept an unchanged level of income or experienced an income decrease. One

possible explanation is that since HCA is the only amount that is subtracted from parents

total income in this research, other types of financial support might be effecting the total

income and therefore also the results.

When examining how income progresses after a period of caring for an SI/D, the

results show that each year following a period of receiving HCA is negatively associated

with income. This indicates that parents do not easily adapt and reach their former

strength in the labour market. The negative effect is not confined to the first few years

after HCA termination, it rather seems to extend far into the future and get worse with

time. This is an indication that social services do not meet the needs of individuals who

face these circumstances. This is the expected conclusion and in accordance with the

background literature. One study concludes that the economic burden of caring for an ill

child can have long-term effects on the financial security of the family (Miedemaq, Easley,

Fortin, Hamilton, Mathews, 2008). Another one finds that all parents are affected by loss

of social security benefits that can result in a large immediate drop in income and that re-

34

entering the labour market can be a slow process and the financial impact can extend far

into the future (Corden, Sloper, Sainsbury 2002).

This research is limited by shortcoming of data. With more individuals it would have

been possible to get more significant results. Also it would have been possible to examine

more years both during and after receiving HCA. One of the main strengths of this study

is the concentration of the data base because it covers an exhaustive list of all

beneficieries for the whole of Iceland.

Even though this study has some limitations and concerns it sheds light on the

relationship between caring for SI/D children and parents income. In Iceland this

relationship has not been studied much. Some of the results are supported by other

studies, other are not. It certainly adds to the literature on the efficiency of selected social

services in Iceland. Yet many questions are still unanswered. For future research it would

be interesting to use data from other countries for comparison.

35

7 References

Alþingi. (1997). Lög um breytingu á lögum nr. 118/1993, um félagslega aðstoð, með síðari breytingum. Downloaded on April 9. 2015 from the website http://www.althingi.is/altext/stjt/1997.092.html

Alþingi. (2015). Svar félags- og húsnæðismálaráðherra við fyrirspurn frá Brynhildi S. Björnsdóttur um rétt foreldra til stuðnings vegan missis barns. Downloaded on April 1. 2016 from the website http://www.althingi.is/altext/145/s/0374.html

Berntsson, L. Köhler, L. (1999). Long-term illness and psychosomatic complaints in children aged 2-17 years in the five Nordic countries: Comparison between 1984 and 1996. Downloaded on April 6. 2015 from the website http://eurpub.oxfordjournals.org/content/eurpub/11/1/35.full.pdf.

Corden, A. Sloper, P. Sainsbury, R. (2002). Financial effects for families after the death of a disabled or chronically ill child: a neglected dimension of bereavement. Downloaded on September 24. 2013 from the website http://onlinelibrary.wiley.com/doi/10.1046/j.1365-2214.2002.00267.x/full

Eiser, C. Upton, P. (2006). Costs of caring for a child with cancer: a questionnaire survey. Downloaded on September 24. 2013 from the website http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2214.2006.00710.x/full

Grossman, M. (1972). On the concept of health capital and the demand for health. Journal of Political Economy, 80, 223–255.

Hagstofa. (2016). Breytingar mánaðarlegrar launavísitölu 1989-2016. Downloaded on March 15. 2016 from the website

http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2214.2006.00710.x/full

36

http://px.hagstofa.is/pxis/pxweb/is/Samfelag/Samfelag__launogtekjur__1_launavisitala__1_launavisitala/VIS04000.px/?rxid=4699c7f7-d504-4d38-93a2-b5942f260c5b

Hamilton, L. (2006). Statistics with Stata. Belmont, CA : Thomson Brooks/Cole.

Jacobson, L. (2000). The family as producer of health, an extended Grossman model. Journal of Health Economics, 19, 611–637.

NPA miðstöðin. (2016). Hvað er NPA? Downloaded on April 22. 2016 from the website http://www.npa.is/hvad-er-npa

Miedema, B. Easley, J. Fortin, P. Hamilton, R. Mathews, M. (2008). The economic impact on families when a child is diagnosed with cancer. Downloaded on September 24. 2013 from the website http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2528308/.

Missen, A. Hollingworth, W. Eaton, N. Crawley, E. (2011). The financial and psychological impacts on mothers of children with chronic fatigue syndrome (CFS/ME). Downloaded on September 24. 2013 from the website http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2214.2011.01298.x/full

Montes, G. Halterman, J. (2007). Association of childhood autism spectrum disorders and loss of family income. Downloaded on September 24. 2013 from the website http://pediatrics.aappublications.org/content/121/4/e821.full.pdf+html

Stata. (2013). Manuals 13. Fixed-, between-, and random-effect and population-averaged linear models. Downloaded on april 5. 2016 from the website http://www.stata.com/manuals13/xtxtreg.pdf

Syse, A. Larsen, I. Tretli, S. (2011). Does cancer in a child affect parent’s employment and earnings? A population-based study. Downloaded on September 24. 2013 from the website http://www.sciencedirect.com/science/article/pii/S1877782110001529.

Tryggingastofnun. (2015a). Downloaded on April 17. 2015 from the website http://www.tr.is/tungumal/enska/caregiver-payments/

http://www.npa.is/hvad-er-npa

http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2214.2011.01298.x/full

37

Tryggingastofnun. (2015b). Downloaded on April 17. 2015 from the website http://www.tr.is/tungumal/enska/caregiver-payments/who-receive-assistance/.

Tryggingastofnun. (2015c). Downloaded on April 17. 2015 from the website http://www.tr.is/tryggingastofnun/tryggingastofnun_i_tolum/stadtolur/toflur-fyrir-arid-2013/.

Tryggingastofnun. (2015d). Downloaded on April 17. 2015 from the website http://www.tr.is/barnafjolskyldur/umonnunargreidslur/hve-mikil-er-adstodin/

Tryggingastofnun. (2015e). Downloaded on April 7. 2015 from the website http://www.tr.is/media/skjol-a-pdf/reglugerd-504_1997-med-breytingum.pdf

Venter, K. (2011). Fathers care too: The impact of family relations on the experience of work for parents of disabled children. Downloaded on April 6. From the website http://www.socresonline.org.uk/16/3/1.html

Zuleyha, C. Marcus, S, Mandell, D. (2012). Implications of childhood autism for parental employment and earnings. Downloaded on September 24. 2013 from the website http://pediatrics.aappublications.org/content/129/4/617.full.pdf+html

http://www.tr.is/tungumal/enska/caregiver-payments/who-receive-assistance/

http://pediatrics.aappublications.org/content/129/4/617.full.pdf+html

38

Appendix A

On the website for the Icelandic Social Insurance Administration, information on number

of beneficiaries and expenditure on home-care allowances is listed in table 1.4 for each

year from 2003 to 2013. No information was found on the years before 2003 or after

2013. The information has been collected and depicted in the two charts below, the

numbers are at the price level for year 2013. The charts show that women are a great

majority in the number of beneficiaries, and therefore they also receive most of the

amounts. The charts also show that both number of beneficiaries and total amount of

expenditure was very stable in the time interval. The average for total number of

beneficiaries is 2.116,46 and the standard deviation is 0,64. The average for annual

expenditure is ISK 1.636,59 million and the standard deviation is 61,43 million

(Tryggingastofnun (c), 2015)

Picture 14 Total number of HCA beneficiaries in December each year in years 2003 to 2013

39

Picture 15 Annual expenditure on HCA in ISK millions in years 2003 to 2013 (price index 2013)

40

Appendix B

On the website for the Icelandic Social Insurance Administration this information is listed

in a table showing variable possible amounts of monthly home-care allowance. As the

table shows the minimum amount is ISK 36.338 and the maximum is ISK 145.351. These

amounts are valid since January 2015 according to the website (Tryggingastofnun (d),

2015).

Table 7 Monthly Home-care allowance in ISK in January 2015

Flokkur 1 (kr/mán.) 100%

145.351 kr.

50%

72.676 kr.

25%

36.338 kr.


123.548 kr.

43%

62.501 kr.

25%

36.338 kr.


101.746 kr.

35%

50.873 kr.

25%

36.338 kr.

Flokkur 4 (kr/mán.) 0 0 25%

36.338 kr.

Flokkur 5 (kr/mán.) 0 0 0

41

Appendix C

Economical background

Grossman´s theory on the demand for health argues that good health is a commodity

which is produced and consumed by the individual and the optimal level of investment in

health occurs when marginal cost and marginal benefit of health are equal (Grossman,

1972). Though the Grossman model has been extremely influential in the field of health

economics it does not take into account that most people lead their lives within a family

that influences their behaviour. Jacobson´s theory extends the Grossman model and

presents the family as a producer of health where family members have common

preferences and will allocate their joint resources in the production of health (Jacobson,

2000). Accordingly, parents will allocate their resources to produce their child´s health,

but a change in the child´s health conditions will change the price and utility of their

investment.

42

Appendix D

Table 8 Results for income for all, female, male, partner and no partner for the original data, see table 2 for the adjusted data (models 1, 2 and 3)

GLS RE model

Dep. Variable:

Income Coef P>/z/ Coef P>|z| Coef P>|z| Coef P>|z| Coef P>|z|

TreatGroup -254 0,000*** -311 0,000*** 8 0,968 -383 0,000*** -64 0,442

Year 133 0,000*** 134 0,000*** 129 0,000*** 134 0,000*** 117 0,000***

Partner 1298 0,000*** 1426 0,000*** 593 0,000*** - - - -

Female -396 0,000*** - - - - -193 0,063* -898 0,000***


Dur1HCA 151 0,039** 73 0,111 559 0,117 28 0,795 276 0,000***

Dur2HCA 266 0,004*** 205 0,002*** 474 0,154 134 0,339 415 0,000***

Dur3HCA 472 0,041** 223 0,000*** 1620 0,191 444 0,234 471 0,000***

Dur4HCA 395 0,000*** 400 0,000*** 393 0,084* 190 0,086* 639 0,000***

Dur5HCA 317 0,003*** 334 0,004*** 258 0,369 205 0,212 397 0,000***

Dur6HCA 264 0,004*** 255 0,009*** 327 0,188 146 0,287 362 0,003***

Dur7HCA 582 0,000*** 480 0,000*** 1172 0,103 539 0,026** 598 0,000***

Dur8HCA 477 0,003*** 439 0,012** 721 0,033** 355 0,137 554 0,004***

Dur9HCA 516 0,001*** 422 0,009*** 1046 0,052* 323 0,093* 825 0,004***

Dur10HCA 284 0,007*** 177 0,309 891 0,157 236 0,329 332 0,205

TreatGroup -337 0,000*** -373 0,000*** -172 0,287 -436 0,000*** -164 0,040**

Year 132 0,000*** 134 0,000*** 127 0,000*** 133 0,000*** 116 0,000***

Partner 130 0,000*** 1426 0,000*** 584 0,000*** - - - -

Female -397 0,000*** - - - - -193 0,062* -903 0,000***


Aft1HCA -234 0,001*** -181 0,007*** -516 0,020** -227 0,024** -198 0,006***

Aft2HCA -286 0,000*** -257 0,000*** -469 0,048** -256 0,009*** -268 0,001***

Aft3HCA -287 0,000*** -260 0,001*** -465 0,074* -219 0,052* -299 0,002***

Aft4HCA -191 0,099* -260 0,014*** 64 0,881 -114 0,447 -198 0,281

Aft5HCA -373 0,000*** -481 0,000*** 66 0,847 -231 0,118 -501 0,000***

Aft6HCA -376 0,004*** -424 0,003*** -192 0,504 -298 0,093* -362 0,072*

Aft7HCA -444 0,000*** -440 0,001*** -500 0,085* -299 0,122 -563 0,000***

Aft8HCA -397 0,024** -349 0,085* -637 0,052* -175 0,521 -632 0,000***

Aft9HCA -723 0,000*** -601 0,001*** -1261 0,000*** -567 0,022** -860 0,000***

Aft10HCA -812 0,001*** -673 0,025*** -1338 0,000*** -685 0,066* -916 0,000***

TreatGroup -151 0,016** -210 0,000*** 137 0,549 -304 0,000*** 467 0,545

Year 137 0,000*** 139 0,000*** 134 0,000*** 137 0,000*** 122 0,000***

Partner 1293 0,000*** 1422 0,000*** 579 0,000*** - - - -

Female -397 0,000*** - - - - -192 0,064* -902 0,000***


No partnerPartner

Obs. 235.271 Obs. 192.235 Obs. 43.036

All Female Male

Obs. 151.459 Obs. 83.812

43

Table 9 Results for income for child´s different diagnosis for the original data, see table 3 for the adjusted data (models 2 and 3)

GLS RE model

Dep. Variable:



Dur1HCA 338 0,163 219 0,004*** 100 0,585

Dur2HCA 378 0,167 293 0,002*** 359 0,145

Dur3HCA 296 0,122 281 0,000*** 881 0,248

Dur4HCA 219 0,122 411 0,001*** 432 0,044**

Dur5HCA 31 0,867 303 0,008*** 632 0,059*

Dur6HCA 293 0,219 235 0,038** 280 0,100*

Dur7HCA 752 0,048* 366 0,012** 915 0,039**

Dur8HCA 392 0,135 439 0,081* 642 0,010***

Dur9HCA 87 0,714 434 0,049** 1217 0,003***

Dur10HCA 107 0,763 222 0,346 624 0,065*

TreatGroup -589 0,000*** -349 0,000*** -170 0,053*

Year 136 0,000*** 134 0,000*** 136 0,000***

Partner 1362 0,000*** 1301 0,000*** 1376 0,000***

Female -330 0,000*** -353 0,000*** -385 0,000***


Aft1HCA -494 0,000*** -170 0,029** -211 0,121

Aft2HCA -459 0,003*** -262 0,000*** -274 0,043**

Aft3HCA -490 0,003*** -255 0,004*** -301 0,039**

Aft4HCA 60 0,871 -314 0,003*** -37 0,875

Aft5HCA -76 0,814 -404 0,000*** -323 0,104

Aft6HCA -314 0,312 -410 0,009*** -230 0,338

Aft7HCA -179 0,620 -416 0,013** -503 0,005***

Aft8HCA -55 0,895 -579 0,001*** 67 0,860

Aft9HCA -853 0,021** -797 0,000*** -674 0,006***

Aft10HCA -741 0,006*** -812 0,021** -787 0,000***

TreatGroup -398 0,001*** -154 0,023** 15 0,894

Year 136 0,000*** 137 0,000*** 137 0,000***

Partner 1362 0,000*** 1297 0,000*** 1376 0,000***

Female -329 0,000*** -354 0,000*** -385 0,000***


Disorder

Obs. 214.038

Illness

Obs. 203.709

Childs diagnosis

Obs. 195.738

Disability

44

Table 10 Results for income after HCA termination for different HCA duration for the original data, see table 4 for the adjusted data (model 3)

Table 11 Results for income after HCA termination for child´s different age at HCA onset for the original data, see table 5 for the adjusted data (model 3)

GLS RE model

Dep. Variable:



Aft1HCA -241 0,004*** -437 0,000*** -776 0,000***

Aft2HCA -337 0,000*** -280 0,040** -933 0,000***

Aft3HCA -290 0,002*** -538 0,000*** -1185 0,001***

Aft4HCA -160 0,255 -529 0,002*** -1336 0,000***

Aft5HCA -368 0,001*** -797 0,000*** -614 0,008***

Aft6HCA -314 0,036** -1148 0,000*** (omit) -

Aft7HCA -433 0,002*** -1134 0,000*** (omit) -

Aft8HCA -392 0,043*** -1108 0,000*** (omit) -

Aft9HCA -756 0,000*** -1175 0,002*** (omit) -

Aft10HCA -823 0,001*** -2871 0,000*** (omit) -

TreatGroup -30 0,666 -133 0,184 233 0,402

Year 138 0,000*** 137 0,000*** 138 0,000***

Partner 1304 0,000*** 1355 0,000*** 1367 0,000***

Female -384 0,000*** -335 0,000*** -325 0,001***


(omit) is omitted because of collinearity.

1-5 years 6-10 years 11-15 years

HCA duration

Obs. 222.967 Obs. 198.205 Obs. 189.621

GLS RE model

Dep. Variable:



Aft1HCA -108 0,455 -170 0,040** -498 0,000***

Aft2HCA -574 0,705 -217 0,007*** -682 0,000***

Aft3HCA -115 0,449 -206 0,034** -631 0,000***

Aft4HCA 29 0,895 -222 0,108 -386 0,132

Aft5HCA -94 0,594 -369 0,010*** -686 0,000***

Aft6HCA -244 0,101 -345 0,087* -579 0,038**

Aft7HCA -18 0,926 -438 0,040** -933 0,000***

Aft8HCA 395 0,362 -480 0,030** -1169 0,000***

Aft9HCA -392 0,040 -659 0,007*** -1230 0,000***

Aft10HCA -265 0,171 -1096 0,022** -937 0,360

TreatGroup -569 0,000*** -105 0,136 263 0,010***

Year 137 0,000*** 137 0,000*** 137 0,000***

Partner 1365 0,000*** 1316 0,000*** 1334 0,000***

Female -375 0,000*** -350 0,000*** -315 0,001***


Age 13-20

Obs 200.051

Age 0-5 Age 6-12

Obs 201.760 Obs 208.982

Childs age at HCA onset

45

Table 12 Results for income after HCA termination for child´s different age at HCA onset for the original data, see table 6 for the adjusted data (model 3)

GLS RE model

Dep. Variable:



Aft1HCA -22 0,944 -65 0,526 -367 0,000***

Aft2HCA 149 1.000 -117 0,201 -438 0,000***

Aft3HCA -84 0,772 -45 0,694 -462 0,000***

Aft4HCA 295 0,494 -77 0,613 -363 0,001***

Aft5HCA 103 0,744 -119 0,366 -631 0,000***

Aft6HCA -176 0,442 53 0,826 -676 0,000***

Aft7HCA 31 0,920 55 0,820 -912 0,000***

Aft8HCA 1080 0,166 -317 0,080* -884 0,000***

Aft9HCA -247 0,389 -398 0,138 -1089 0,000***

Aft10HCA 23 0,933 -781 0,000*** -1081 0,033**

TreatGroup -474 0,047** -544 0,000*** 136 0,000***

Year 138 0,000*** 137 0,000*** 137 0,000***

Partner 1382 0,000*** 1343 0,000*** 1301 0,000***

Female -361 0,000*** -336 0,000*** -350 0,000***


Childs age at HCA ending

Obs 193.795 Obs 202.383 Obs 214.615

Age 0-5 Age 6-12 Age 13-20

46

Appendix E

Declaration of interest

I had a daughter named Fanney Edda who was born in 2007. She had a neurodegenerative

disease called BVVL that brought her to death in 2010. She became severely ill and

disabled in the year 2008 and after that she needed very specialized care in her daily life.

That is when my family started receiving home-care allowance along with many other

social services. At this time I was an undergraduate student studying economics and

became very interested in the structure of the services we received and the economic

decisions made around it. Since then I have closely followed discussions about services

for disabled and severely ill children. I want to use my experience and insight in this area,

as well as my education in economics and disability studies, to contribute to the

development of these services in Iceland. Even though my life experience affected my

choice of materials and is noted here, I have no personal interest in the outcome of the

project and am no longer a recipient of any transfers related to the projects topic.

MS ritgerð Heilsuhagfræði The effect of children´s disability and … · Leiðbeinandi minn er Tinna Laufey Ásgeirsdóttir, doktor í hagfræði og prófessor við hagfræðideild

Documents