Understanding the growth in invalid's benefit receipt in New Zealand.

Abstract

This paper reports on research that uses the Ministry of Social Development's benefit administration data to advance our understanding of the growth in the number of people receiving the Invalid's Benefit over the decade to 2002. It investigates the growth in inflows of people to Invalid's Benefit, as this was the main cause of growth in recipient numbers. Some of the growth in inflows can be explained by population growth, population ageing, and the effects of the rise in the age of eligibility for New Zealand Superannuation. However, more than half cannot be accounted for by these factors and is explained instead by an increase in the proportion of people aged 15-59 taking up Invalid's Benefit. This increase was explained both by growth in transfers from other benefits, and by growth in entries from outside the benefit system, some of which was associated with growth in uptake by former benefit recipients returning after a period off benefit. Possible explanations for these shifts, many of which warrant further investigation, are canvassed.

INTRODUCTION

In New Zealand, income protection for working-age people unable to work due to incapacity takes three main forms:

* Invalid's Benefit, which provides for people with a long-term and severe incapacity

* Sickness Benefit, which provides for people with a short-term incapacity

* Accident Compensation Corporation (ACC) earnings-related weekly compensation, which provides for people with an accident-related incapacity. (2)

To qualify for an Invalid's Benefit, a person must be "permanently and severely" restricted in his or her capacity for work, where "permanently" means that the sickness, injury or disability is expected to continue for at least two years, or the person has a terminal illness and is not expected to live more than two years; and "severely" means that a person could not regularly work 15 hours or more per week in open employment. Payments are subject to a test of the joint income of the claimant and their partner.

The number receiving Invalid's Benefit has grown considerably over the last three decades (Figure 1). Growth was particularly rapid between 1992 and 2002, with a doubling of numbers over this period. This occurred in the context of a downward shift in unemployment. The official unemployment rate fell from its peak of 10.6% in 1992 to 6.2% in 1996 and 1997, and then rebounded to 7.5% in 1999 before falling once more to 5.3% in 2002. (3) All measures of unemployment, including long-term unemployment, were lower in 2002 than in 1992. (4)

[FIGURE 1 OMITTED]

The proportion of working-age people receiving a Sickness Benefit, an Invalid's Benefit or ACC weekly compensation rose from around 1% in the 1970s to 5% in June 2002. Most other OECD countries also experienced a rise in the proportion of the working-age population claiming incapacity benefits over this period. By the late 1990s and early 2000s, it was common for around 4-6.5% of the working-age population to receive such benefits (OECD 2003:60-63).

While the proportion of the New Zealand working-age population receiving incapacity benefits in 2002 was not high relative to other developed countries, it was acknowledged that there was a need to understand the growth in order to inform the development of policy and service responses. (5) In late 2002, the Ministry of Social Development (MSD) established a programme of research that aimed to:

* identify the key factors behind the growth in Sickness and Invalid's Benefits

* understand the populations in receipt of these benefits

* identify approaches and interventions that support the social and economic wellbeing of these populations and their participation in employment.

This paper summarises the findings of the research that was undertaken on the growth in Invalid's Benefit numbers, (6) which centred on analysis of the MSD benefit dynamics data set. This is a longitudinal research data set assembled from benefit administration records. (7) At the time the analysis was conducted it covered the period 1 January 1993 to 31 December 2002, and these dates mark the start and finish points of the study. The data set includes all people who received any main working-age benefit in that period, either as the primary recipient or as a partner. This allowed examination of the contribution to the growth of transitions from other benefit types to Invalid's Benefit.

UNDERSTANDING HOW NUMBERS GREW

The number of benefit recipients grows whenever more people come on to the benefit than leave. Figure 2 shows official data on the number of grants of Invalid's Benefits each year, and the number in receipt. The number of cessations is obtained by subtraction. (8) Over the decade to 2002, there was growth in both grants and cessations, with the number of people entering Invalid's Benefit each year exceeding the number of people leaving by a widening margin. (9)

[FIGURE 2 OMITTED]

Examination of these data suggests that the widening gap between grants and cessations between 1992 and 2002 was caused mainly by growing inflows rather than an increase in the length of time people were staying on the Invalid's Benefit. The length of stays was in fact shorter for those who entered later in the decade 1993-2002 than for those who entered earlier in the decade. (10,11)

Even though duration did not increase for those entering Invalid's Benefit over the decade, the long length of stay associated with Invalid's Benefit receipt was important in explaining how numbers grew so rapidly. Of those who entered Invalid's Benefit in 1993, for example, half remained continuously in receipt of that benefit six years later, and more than a third remained 10 years later. These long durations meant that growth in grants caused the number in receipt to rapidly cumulate.

BEGINNING TO UNDERSTAND WHY NUMBERS GREW

This section begins the process of investigating the growth in inflows of people to Invalid's Benefit, because this was the main cause of the growth in the number of recipients. (12)

Demographic Change and the Rise in the Age of Eligibility for New Zealand Superannuation

Growth in inflows could have been caused by a number of factors, including:

* growth in the size of the population aged 15-64, which increased the numbers of people potentially eligible

* population ageing, which increased the proportion of the population aged 15-64 in age groups at which the likelihood of taking up Invalid's Benefit is high

* growth in the proportion of the population in given age groups coming on to the benefits each year; i.e. an increase in age-specific "inflow rates", which, in turn, could have been caused by:

--growth in inflow rates for the population aged 60-64, as a result of the rise in eligibility for New Zealand Superannuation (13)

--growth in inflow rates for younger working-age groups.

Table 1 shows the independent contributions these various factors made to the growth in the number coming on to Invalid's Benefit, comparing 2002 and 1993. The contributions are estimated by calculating the growth that would have occurred had each factor changed in isolation while all other factors remained constant. (14) It should be noted that a different choice of years to compare would result in different contributions. (15)

Table 1 indicates that some of the growth was inevitable given the growth and ageing of the population and given the rise in the age of eligibility for New Zealand Superannuation, which led increasing numbers of people with incapacities in the 60-64 age group to take up the Invalid's Benefit.

More than half of the growth in inflows cannot be explained by demographic change and the New Zealand Superannuation age rise. It reflects, instead, an increase in inflow rates for people aged between 15 and 59. Figure 3 shows that inflow rates were higher in 2002 than in 1993 for all age groups within this band.

[FIGURE 3 OMITTED]

Inflow rates at ages 15-59 fell temporarily in the late 1990s. This may partly reflect the October 1998 change in the criteria for entry to the Invalid's Benefit from "75% incapacitated" to "permanently and severely" restricted capacity for work.

The remainder of this section concentrates on gaining a better understanding of the upward movement in inflow rates at ages 15-59, because this is the component of growth that requires the most explanation.

Growth in Transfers from Other Benefits

Benefit administration data allow us to examine whether people coming on to Invalid's Benefit came from other parts of the benefit system or from outside the benefit system. (16)

What we find is that growth in inflow rates at ages 15-59 was partly associated with growth in the number of people making direct transfers on to Invalid's Benefit from other parts of the benefit system. (17)

We estimate that growth in transfers explains 60% of the growth in Invalid's Benefit inflow rates at ages 15-59 between 1993 and 2002. (18) This translates to a 31% contribution to the overall growth in inflows to that benefit (Table 2), (19) a share of growth that is bigger than the independent contributions of either the demographic changes or the increase in inflow rates at ages 60-64 shown in Table 1.

The growth in transfers comparing 1993 and 2002 occurred in the context of:

* only a slight increase in the proportion of the population aged 15-59 receiving the Sickness Benefit, the main benefit of origin for people transferring to the Invalid's Benefit

* a large fall in the proportion of that population receiving unemployment-related benefits

* a slight fall in rates of receipt of the Widow's Benefit and Domestic Purposes Benefit, which is mainly received by sole parents (Wilson et al. 2005:28).

This means that the growth largely reflected an increase in the rate at which people on other benefits were transferring to the Invalid's Benefit, rather than an increase in the proportion of the population receiving other benefits and potentially able to transfer.

Growth in Entries by Former Benefit Recipients

Forty per cent of the growth in the proportion of the population aged 15-59 entering the Invalid's Benefit between 1993 and 2002 is explained by an increase in numbers coming on to the Invalid's Benefit from outside the benefit system ("Non-transfers" in Table 2). More detailed examination of this growth indicated that some was generated by growth in entries by people who had received a benefit for some period of time in the two years prior. (20)

Growing Inflows of People with Mental Disorders and Musculoskeletal Incapacities

We can gain some additional insights into the drivers of growth in inflows to Invalid's Benefit at ages 15-59 by examining the types of incapacity that generated most of the growth. This can only be done for the period from 1996 onwards because changes in the coding of incapacities in 1995 make comparisons over a longer period less reliable. It is also not possible to calculate population-based rates that indicate whether growth in inflow rates varied between population groups with different conditions.

Table 3 shows the percentage of growth in inflow rates at ages 15-59 that can be accounted for by growth in the number coming on to Invalid's Benefit in different incapacity groupings. (21) We examine the highest-ranked incapacity recorded for each individual (22) at the time their benefit was granted. (23) These incapacities are aggregated into internationally comparable standard groupings, with the exception of the large "mental disorders" grouping, for which individual subgroups are analysed.

Around two-thirds of the growth in inflow rates at ages 15-59 between 1996 and 2002 was associated with growth in inflows of people with mental disorders. Within that grouping, the largest contributions were associated with:

* depression (accounting for 19% of the growth)

* bipolar disorder (accounting for 11% of the growth)

* the residual "other psychological" category (accounting for 14% of the growth).

Growth in inflows of people with schizophrenia, stress and incapacities related to substance abuse also contributed.

Most of the remaining growth was associated with growth in musculoskeletal incapacities. With the exception of cancer and circulatory and nervous conditions, all other physical incapacity groupings also contributed, to some degree, to the growth that occurred.

Growing Inflows of Maori and Pacific People with Chronic Disease

Table 4 shows the contribution of the different incapacity groupings to growth in inflows at ages 15-59 for different ethnic groups. (24) These figures differ in scale from those presented in Table 3 because they refer to changes in inflows, rather than changes in inflow rates standardised to control for changes in the age distributions of the different population groups. (25) The five incapacity types that explain most of the overall growth are highlighted.

The rate of growth in inflows was most rapid for Maori and Pacific peoples (41% and 36% respectively, compared with 24% for the New Zealand and other European grouping). It is likely that significant changes in the recording and coding of ethnicity in benefit administration data partly account for this. (26) Variations in the rate of population growth or in the rate of population ageing between ethnic groups may also be important.

Most of the growth in the number of Maori and Pacific entrants is explained by incapacities other than the five that generated most of the overall growth.

* For Maori, this reflects larger contributions from schizophrenia (accounting for 12% of the growth in Maori inflows at ages 15-59), cancer (7%), respiratory (6%), circulatory (5%) and nervous (5%) conditions.

* For Pacific peoples, it reflects larger contributions from circulatory conditions (accounting for 23% of the growth in Pacific inflows at ages 15-59), endocrine conditions (17%) and cancer (10%).

This indicates a greater propensity for growth in inflows to be associated with chronic disease for these groups. We would expect the difference to be even more pronounced if we were able to calculate inflow rates standardised for the different age distributions of these populations, given the relatively youthful age profiles of the Maori and Pacific populations and the positive association between chronic disease and age.

QUESTIONS THAT REMAIN

Although benefit administration data give some valuable insights into how and why the growth in numbers receiving Invalid's Benefit occurred, they do not allow us to tackle some more fundamental questions about causality. In particular, did the growth in the proportion of the population coming on to Invalid's Benefit each year reflect:

* an increase in the uptake of this benefit by people with qualifying incapacities?

* a change in the way qualifying incapacities were interpreted or administered, which extended the proportion of the population assessed as qualifying?

* an increase in the prevalence of qualifying incapacities among the population as a whole, or among subgroups of the population?

Increasing Uptake by People with Qualifying Incapacities?

If the health and disability status of the population and its subgroups remained unchanged, and the interpretation and administration of qualifying incapacities remained unchanged, then the only explanation for the growth in age-specific inflow rates for people aged 15-59 was an increase in the proportion of people with qualifying incapacities who received income from Invalid's Benefit rather than from other sources.

We have shown that some of this growth was associated with shifts on to Invalid's Benefit by people who were already receiving or who had recently received other main benefits. Beatty and Fothergill (2003) set out a model that can explain this type of shift without any increase in the prevalence of incapacity or any widening of qualifying incapacities (although these changes might also occur).

Similarly, the rise in sole parenthood over the last two decades might have been associated with a shift of some people with poor health and disabilities first onto the Domestic Purposes Benefit and then onto Invalid's Benefit, perhaps in response to the extension of work testing and work-oriented case management to this benefit from the late 1990s, and the focus on "full and correct entitlement" in the administration of benefits from 2000.

There may have also been a "benefit shift" from Sickness Benefit to Invalid's Benefit by people with mental and other health problems of a severity that could potentially qualify them for the Invalid's Benefit. This may have been partly associated with the reductions in Sickness Benefit payment rates, which occurred in 1991 and 1998, and the later focus on "full and correct entitlement".

Other overseas research could indicate why, in the context of a shift onto incapacity-related benefits from within the benefit system, the growth was dominated by depression and other mental disorders. In Australia, 34% of unemployed income support recipients and 45% of unpartnered women with children receiving income support experience a diagnosable mental health disorder in a 12-month period, compared with 19% of people not in receipt of income support (Butterworth 2003). (27) If recipients of non-incapacity-related benefits in New Zealand experienced health and particularly mental health problems at a similar rate, then many would have been able to qualify for Invalid's Benefit.

Changes in the Way that Qualifying Incapacities were Interpreted or Administered?

Determining whether a person is sufficiently incapacitated to qualify for Invalid's Benefit is not straightforward. The eligibility criteria are subject to interpretation, and a range of factors other than medical eligibility (including the demand for labour in the local labour market) may influence the decisions of assessing doctors (Lennan 2000, White 2000). This may have been compounded by the growth in the number of applicants with mental disorders and musculoskeletal conditions. It is difficult not only to diagnose some of these conditions, but also to assess their implications for work capacity and to predict how they will evolve in the future (OECD 2003:85).

Changes in doctors' interpretation of the medical criteria for entry may have extended the proportion of the population potentially eligible. This could have occurred as a result of an increase in the recognition and diagnosis of conditions such as stress, depression and over-use injuries, perhaps in response to increased advocacy by mental health consumers and disability groups. It could have occurred as a result of changes in the demand for labour or a change in the structure of the available employment; or it could have been in response to policy and administrative changes, including:

* the reduction in payment rates for the Sickness Benefit

* an increase in referrals for assessment with the extension of work testing and work-oriented case management to new groups of benefit recipients, the increased focus on case managing the long-term unemployed as unemployment fell, and the move to ensure "full and correct entitlement" (see Lennan 2000:13)

* a change in assessment forms supplied to doctors in 1995, which began to list possible incapacities as tick boxes which included "stress" and "depression". (28)

Increasing Prevalence of Qualifying Incapacity?

A final set of possible explanations relates to changes in the prevalence of incapacity.

Comparisons of the prevalence of mental illnesses over time are difficult to make, but rates may have been rising. Possible contributors include the effects on mental health of the earlier period of high rates of long-term unemployment. Unemployment is associated with poorer mental health (Wilson 2000) a relationship that longitudinal studies indicate is in part causal (see Wilson 2000, Kasl and Jones 2000 for reviews).

There are also suggestions that changes in the nature of work may be creating new health risks. Internationally, the growth of "precarious" employment, and an increase in the intensity of work, for example, may be linked to the growth in inflows to incapacity benefits by people with stress and depression (see Kemp 2006:19 for a review).

Could the conditions associated with growth in Invalid's Benefit inflows for Maori and Pacific peoples indicate an increase in the prevalence of chronic disease among these populations (or subgroups of these populations)? The contribution of cancer to the growth in inflows found here for Maori and Pacific peoples is consistent with the increasing mortality rate from cancer over the 1980-1999 period found for these groups by Ajwani et al. (2003). That study found widening ethnic disparities in overall chronic disease mortality, a finding that the authors link at least in part to the more adverse effects of economic restructuring on the Maori and Pacific populations and ethnic differences in access to health care.

We observe patterns of growth for Pacific people that are consistent with the likely effects of the current rise in the prevalence of diabetes. It is possible that the associated growth in obesity could be contributing to the growth in the number of people coming on to Sickness and Invalid's Benefits with musculoskeletal conditions.

CONCLUSION

This study contributes new information to establishing how the number of people receiving the Invalid's Benefit grew in the last decade. It shows that the growth appears to have been generated mainly by increased inflows rather than longer stays. It also helps us begin to understand why the growth in inflows occurred. It shows that some of that growth was inevitable given population growth, population ageing, and the effects of the rise in the age of eligibility for New Zealand Superannuation from 60 to 65 years. But half the growth in inflows is explained by an increase in the proportion of people aged 15-59 who came on to Invalid's Benefit each year.

Benefit dynamics data allow us to begin to unravel the reasons for growth in inflow rates at these ages. We find that growth in transfers from other benefits played an important role, and this growth may have been associated with changes in policy and administration affecting other benefits. Growth in entries from outside the benefit system was also important, and we find that some of that growth was associated with growth in uptake by former benefit recipients returning after a period off benefit. While this will partly reflect an increase in the proportion of the population who were former benefit recipients as unemployment fell, it might also indicate that some who left benefits for employment were in a precarious position with respect to their employment and/or their health. It is also possible that people returning to the benefit system were more likely to be case managed on to Invalid's Benefit in recent years with the extension of work expectations to a wider range of benefits and the move to ensure full and correct entitlement.

When we examine the types of incapacities that account for the growth, we find that growth in inflows of people with mental disorders accounts for around two-thirds of the recent growth in Invalid's Benefit inflow rates at ages 15-59, with growth in inflows of people with musculoskeletal incapacities accounting for most of the remainder. For Maori and Pacific peoples, growth in inflows has been associated with a wider range of incapacities. For Maori, the largest additional contributions come from schizophrenia, cancer, and respiratory, circulatory and nervous conditions; for Pacific peoples they come from circulatory conditions, endocrine conditions and cancer. These findings indicate that a greater share of the growth was associated with chronic disease for these ethnic groups.

Although we are unable to examine the underlying drivers of these shifts, it is possible that there were three broad contributors: uptake rates by people with qualifying incapacities may have increased; changes in assessment may have widened the range of incapacities that were assessed as qualifying; and the prevalence of qualifying incapacities may have increased for some population subgroups.

Our hope is that further research into these and other possibilities will bring new insights into the growth in receipt of Invalid's Benefit and greater understanding of the circumstances of people in receipt of these benefits. The growth has perhaps made more visible the size of the population for whom poor health or disability, perhaps in conjunction with other disadvantages, causes difficulty. Good evidence on the characteristics, experiences and aspirations of this population will strengthen the new policy focus on this group.

Beatty and Fothergill's model:

* In the period of full employment, many people with poor health and disabilities were "carried" within the employed workforce. These people had incapacities sufficient to qualify them for incapacity-related benefits but, given the labour market context at that time, these did not stop them from working.

* With economic recession, people with poor health and disabilities were more vulnerable than others to job loss. Some went directly on to incapacity-related benefits, but others who viewed their status primarily as unemployed took up unemployment benefits. Many in this group remained at the back of the queue for new jobs because of their poor health or disability, particularly when this was compounded by low qualifications, older age or other disadvantages. They remained on unemployment benefits for long periods as a result.

* Through disillusion with their prospects as job seekers, or as a result of being moved by the employment services or by benefit administrators, they progressively moved on to incapacity-related benefits.

* This "benefit shift" could take some years to work through to its full effect. Over time, less joblessness would be measured as unemployment and more measured as incapacity-related non-participation.

REFERENCES

Ajwani, S., T. Blakely, B. Robson and M. Tobias (2003) Decades of Disparity: Ethnic Mortality Trends in New Zealand 1980-1999, Ministry of Health and University of Otago, Wellington.

Beatty, C. and S. Fothergill (2003) "Incapacity benefit and unemployment" in P. Alcock, C. Beatty, S. Fothergill, R. Macmillan and S. Yeandle (eds.) Work to Welfare: How Men Become Detached from the Labour Market, Cambridge University Press, Cambridge.

Butterworth, P. (2003) Estimating the Prevalence of Mental Disorders Among Income Support Recipients: Approach, Validity and Findings, Commonwealth Department of Family and Community Services, Policy Research Paper 21, Canberra.

Danziger, S., A. Kalil and N. Anderson (2000) "Human capital, physical health and mental health of welfare recipients: Co-occurrence and correlates" Journal of Social Issues, 56(4):635-654.

Jayakody, T. and D. Stauffer (2000) "Mental health problems among single mothers: Implications for work and welfare reform" Journal of Social Issues, 56(4):617-634.

Jensen, J., S. Sathiyandra, M. Rochford, D.L. Jones, V. Krishnan and K. McLeod (2005) "Work participation amongst people with disabilities: Does the type of disability influence the outcome?" Social Policy Journal of New Zealand, 24:134-159.

Jude Miller Consulting (2006) International Literature Review: Approaches and Interventions for Sickness Benefit and Invalid's Benefit Clients, Ministry of Social Development, Wellington.

Kasl, S. and B. Jones (2000) "The impact of job loss and retirement on health" in L. Berkman and I. Kawachi (eds.) Social Epidemiology, Oxford University Press, New York.

Kemp, P.A. (2006) "Comparing trends in disability benefit receipt" in P.A. Kemp, A. Sunden and B. Bakker Tauritz (eds.) Sick Societies? Trends in Disability Benefits in Post-Industrial Welfare States, International Social Security Association, Geneva.

Lennan, M. (2000) Review of the Designated Doctor Scheme, paper prepared for Department of Work and Income, New Zealand.

Lunt, N. (2006) "Sickness and Invalid's Benefits: New developments and continuing challenges" Social Policy Journal of New Zealand, 27:77-99.

OECD (2003) Transforming Disability into Ability: Policies to Promote Work and Income Security for Disabled People, Organisation for Economic Co-operation and Development, Paris.

Peace, R., A. Wolf, S. Crack, I. Hutchinson and M. Roorda (2004) Wellbeing, Employment, Independence: The views of Sickness and Invalids' Benefit Clients, Ministry of Social Development Working Paper 07/04, Wellington.

White, P. (March 2000) The Role of Designated Doctors, paper prepared for Department of Work and Income, New Zealand.

Wilson, M. (1999) "The duration of benefit receipt: New findings from the benefit dynamics data set" Social Policy Journal of New Zealand, 13:59-82.

Wilson, M. (2001) Longitudinal Analysis of Administrative Data: The Ministry of Social Policy's Benefit Dynamics Project, paper presented at the Ministry of Social Policy Longitudinal Research and Social Policy Seminar, 5-6 April, Wellington.

Wilson, M., K. McLeod and S. Sathiyandra (2005) Growth in Numbers of Sickness and Invalid's Benefit Recipients 1993-2002: Cohort Analysis Technical Report, Ministry of Social Development, Wellington.

Wilson, N. (2000) "Labour force status and health" in P. Howden-Chapman and M. Tobias (eds.) Social Inequalities in Health: New Zealand 1999, Ministry of Health, Wellington.

Worth, H. and K. McMillan (2004) "Ill prepared for the labour market: Health status in a sample of single mothers on welfare" Social Policy Journal of New Zealand, 21:83-97.

Moira Wilson

Keith McLeod (1)

Centre for Social Research and Evaluation

Ministry of Social Development

(1) The authors would like to thank the external reviewers who provided helpful comments on an earlier draft: Dr Lixin Cai and Dr Roger Wilkins of the Melbourne Institute of Applied Economic and Social Research, University of Melbourne; Dr Michael O'Brien of the School of Social and Cultural Studies, Massey University; and David Preston, an independent researcher. Thanks are also due to the many people within the Ministry of Social Development and other agencies who provided assistance by reviewing drafts, checking material or participating in discussions, and to an anonymous referee for helpful comments. The views expressed in this paper are those of the authors and do not necessarily reflect the views of the Ministry of Social Development.

(2) The Sickness Benefit and Invalid's Benefit are income tested, non-contributory social assistance benefits. Neither payment is time limited, but both are payable only as long as the incapacity for work persists. ACC weekly compensation is employment-related social insurance, where payments are related to past earnings but not tied to individual contributions. This payment is payable only as long as injury-related incapacity for work persists.

(3) Averages for years ended March (Labour Market Statistics 2002, Table 1.01).

(4) Labour Market Statistics 2002, Tables 3.05 and 3.13.

(5) See Lunt (2006) for a review and analysis of policy and service developments since 2002.

(6) For a full, technical report on the analysis, including a parallel analysis of the growth in Sickness Benefit numbers, see Wilson et al. (2005). Other reports flowing from the research programme include Peace et al. (2004), Jensen et al. (2005) and Jude Miller Consulting (2006).

(7) For more information on the benefit dynamics data set and its strengths and limitations, see Wilson (1999) and Wilson (2001).

(8) Actual counts of cessations vary from those shown. This is because details are extracted as at the time the cessation action was entered into the SWIFTT benefit payments system and not as at the date the action may take effect. The change in number in receipt over a year may not exactly equal grants less recorded cessations as a result. By presenting cessations obtained by subtraction in the figures shown, we avoid this problem.

(9) It is likely that all of the 1988 and part of the 1991 peak inflows were the result of administratively driven cessation and re-grant activity because both resulted in less marked increases in the number in receipt. This highlights the unreliability of pre-SWIFTT start dates for benefit spells and is one of the reasons why the analysis does not cover the period prior to the introduction of SWIFTT in 1992/1993.

(10) Duration fell and then partially recovered, but was shorter for those entering at the end of the decade than for those entering at the beginning. The shifts that occurred coincided with changes in the scheduling of medical re-assessments of entitlement. See Wilson et al. (2005:15-19) for these findings.

(11) It is possible that the lengthening stays of people who came on to the benefit prior to 1993 contributed to the growth in numbers in receipt in the decade 1993-2002 (through the increased life expectancy of people with some conditions or the longer waiting time for New Zealand Superannuation for example), but this is not something that can be established from the available data. See Wilson et al. (2005:19).

(12) From this point on, we use the term "inflows" rather than "grants" because this better reflects the measure we analyse. Where grants count all benefit commencements, we use inflows to refer only to those commencements that occur more than two weeks after a cessation of the same benefit and that are associated with a spell on benefit of at least one day. The purpose of this is to exclude administratively generated grant and cessation activity.

(13) Between April 1992 and June 2001, the age of eligibility for New Zealand Superannuation gradually increased from 60 to 65 years of age. From 1 October 1994, the age of eligibility for the Transitional Retirement Benefit, designed to assist those affected by the transition, also began to rise; it was completely phased out by 1 April 2004. Many people with incapacities who would formerly have qualified for New Zealand Superannuation (or the Transitional Retirement Benefit in the transition period) now remained on or newly took up incapacity-related benefits. The New Zealand Superannuation age rise is likely to be the main explanation for the growth in inflow rates at ages 60-64.

(14) These factors are not independent of each other. For example, increased inflow rates will be exacerbated when combined with population growth. These additional "interaction" effects of multiple factors cannot easily be apportioned to individual factors, and are therefore represented as "interactions between factors" in Table 2. However, since the net effects are relatively small (12%), they are not particularly important in the context of the overall growth. Nevertheless, it should be noted that each of the individual factors listed has contributed to these interaction effects to some degree.

(15) It is also likely that the contribution of population growth is overstated (and the contribution of increased inflow rates, in particular, understated) because population growth over the period was partly due to net migration gains, and eligibility to Invalid's Benefit requires at least 10 years of residence.

(16) In Wilson et al. (2005) we also examined whether there has been a growth in inflows of former ACC recipients, which could be related to the introduction of work capacity testing within that system. Comparing 1999 (the first year for which data were available) and 2002, we estimated that growth in the number of former ACC recipients coming on to the Invalid's Benefit from outside the benefit system explained 4% of the growth in inflow rates at ages 15-59. Growth in the number of former ACC recipients coming on to the Invalid's Benefit via other benefits made up a further 4%, giving a total contribution of 8%. As growth in inflow rates at ages 15-59 accounted for half of the overall growth in Invalid's Benefit inflows between 1999 and 2002, this means that growth in entries by former ACC recipients can explain 4% of the overall growth in inflows over that period. We were unable to assess whether some of the growth in inflows was the result of a fall in the proportion of ACC claims being granted following the introduction of work-capacity testing or whether it was the result of an increase in the number taking up Invalid's Benefit while waiting for an ACC claim to be assessed.

(17) Defined as occurring where a grant of Invalid's Benefit is made within two weeks of another benefit ceasing.

(18) This estimate is age-standardised to control for the effects of population ageing on inflows from different benefits.

(19) The proportion 31% is obtained by multiplying the 60% share of growth in inflow rates at ages 15-59 by the 52% contribution that growth in inflow rates at ages 15-59 makes to the overall growth in inflows shown in Table 1.

(20) See Wilson et al. (2005:26)

(21) The overall percentages are age-standardised to control for the effects of population ageing (the proportion of entrants with a musculoskeletal incapacity generally increases with age, for example, so that some of the raw growth in the inflow rates associated with that incapacity grouping reflects population ageing). The 1996 population is used as the standard population for the age-standardisation.

(22) Each individual can have up to four incapacities recorded. These are ranked by medical practitioners in order of severity. In many cases, it appears that the ranking given is not well recorded as it often follows the order in which incapacities appear on the form. In most cases, only a single reason code is present so this does not impact on accuracy. For around four in ten people granted IB more than one reason code is entered, however, and in these cases the highest-ranked code may not always be the most severe. The extent of the error is uncertain.

(23) Changes in highest-ranked incapacity that might occur as the spell on benefit progresses are not, therefore, captured. An appendix to Wilson et al. (2005) gives the full names of the groupings analysed for each benefit and the SWIFTT codes and descriptors for conditions that are included in each.

(24) Note that the ethnicity code used for each individual is that most recently recorded in SWIFTT at the extraction date for the data analysed. The effect of this is to minimise the proportion of individuals for whom ethnicity is unknown. Given this approach, the figures shown will differ from time series data drawn directly from SWIFTT. These are based on ethnicity as it was recorded for each individual at each date in the time series.

(25) This is because it is not possible to reliably combine SWIFTT-based ethnicity data with Census-based population estimates. See Wilson et al. (2005:40-41).

(26) See Wilson et al. (2005:40-41).

(27) Recent research on welfare reform in the United States shows a significantly higher prevalence of mental disorders among single mother welfare recipients compared with non-recipients (Danziger et al. 2000, Jayakody and Stauffer 2000). These overseas findings are echoed by a small New Zealand survey of newly work-tested Domestic Purposes Benefit recipients in the late 1990s, which found that self-reported physical and mental health was poorer for this population than for the overall population of New Zealand women of comparable age on all measures (Worth and McMillan 2004).

(28) Doctors had previously hand-written the condition. It is possible that this change widened doctors' notions of the range of conditions that could qualify a person for the benefits.

Table 1 Factors Contributing to Growth in Numbers Coming on to Invalid's
Benefit, 2002 Compared with 1993

Factors contributing to growth in            % share of growth
numbers coming on to Invalid's Benefit       in numbers coming
                                              on between 1993
                                             and 2002 resulting
                                             from each factor
                                                in isolation

Growth in the population aged 15-64                  11
Ageing of the population aged 15-64                  6
Increased inflow rates at ages 60-64                 20
Increased inflow rates at ages 15-59                 52

Net effect of interactions between factors           12

Total                                               100

Table 2 Contribution of Transfers to Increased Invalid's Benefit
Inflow Rates at Ages 15-59 and to Overall Growth in Inflows, 2002
Compared with 1993

                                       % share of
                                       increased    % share of
                                         inflow     increased
                                        rates at    inflows at
                                       ages 15-59   ages 15-64

Transfer from Sickness Benefit             38           20

Transfer from unemployment-related         7            3
benefits

Transfer from DPB or WB                    11           6

Transfer from benefit as partner           5            2

All transfers                              60           31

Non-transfers                              40           20

Increased inflow rates at ages 15-59      100           52

Notes:
Percentages may not sum due to rounding.

Unemployment-related benefits include Unemployment Benefit, Emergency
Unemployment Benefit, 55+ Benefit, Independent Youth Benefit, Young
Job Seekers Allowance, Training Benefit and their Community Wage
equivalents.

DPB is Domestic Purposes Benefit. WB is Widow's Benefit.

Calculations of percentage shares of increased inflow rates at ages
15-59 are standardised to control for the effect of changes in age
structure on inflows from different benefits of origin.

Table 3 Percentage Share of Growth in Invalid's Benefit Inflow Rates
at Ages 15-59 Associated with Different Incapacity Groups,
Age-Standardised to Control for Population Ageing, 2002 Compared
with 1996

                             % share of increased
                             inflow rates at ages
                                    15-59

Mental disorders:
  Bipolar disorder                    11
  Depression                          19
  Intellectual disability              0
  Schizophrenia                        6
  Stress                               9
  Substance abuse                      5
  Other psychological                 14

All mental disorders                  65
Circulatory                           -6
Congenital                             4
Endocrine                              1
Injury                                 7
Musculoskeletal                       22
Cancer                                -3
Nervous                               -4
Respiratory                            3
Other                                  7
Unspecified                            4

All                                  100

Note: Percentages may not sum due to rounding.

Table 4 Percentage Share of Growth in Invalid's Benefit Inflows at Ages
15-59 Associated with Different Incapacity Groups, by Ethnic Group,
2002 Compared with 1996

                                   NZ and Other  Maori  Pacific Peoples
                                     European

Mental disorders:
  Bipolar disorder                      10         4           5
  Depression                            16         8           2
  Intellectual disability                0         1           4
  Schizophrenia                          0        12           5
  Stress                                 6         6           3
  Substance abuse                        4         4           3
  Other psychological                   11         7           0

All mental disorders                    48        44          21
Circulatory                             -2         5          23
Congenital                               3         3           0
Endocrine                                1         3          17
Injury                                   6         5           6
Musculoskeletal                         27        12          10
Cancer                                   3         7          10
Nervous                                  0         5          -4
Respiratory                              4         6           5
Other                                    7         5           5
Unspecified                              3         5           6

All                                    100       100         100

Percentage growth in inflows at         24        41          36
ages 15-59, 1996-2002

Note: Percentages may not sum due to rounding.