Abstract
The Australian government implemented a series of private health insurance (PHI) policy reforms between 1997 and 2000. As a result, the proportion of the population with PHI coverage increased by more than 35%. However, this study found significant evidence that the policy reform disproportionately favours high-income earners. In particular, the 30% premium subsidy represents a windfall gain for households which would have purchased PHI even without the rebate. The amount of such gain is estimated to be around $900 million per year, a large proportion of which went to higher income households.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
According to the Department of Health and Ageing Factbook 2006 (Table 5), the cost to the government of the PHI rebates was A$1.5b in 1999/2000, A$2.1b in 2000/2001 and 2001/2002, A$2.3b in 2002/2003, A$2.5b in 2003/2004 and A$2.9b in 2004/2005. The Factbook can be accessed online at: http://www.healthconnect.gov.au/internet/wcms/publishing.nsf/Content/Factbook2006-1. See also Deeble [5].
This policy was initially estimated to cost the government $1.09 billion. However, with a substantial increase in the PHI take-up rate by 2001, the cost has exceeded $2 billions a year in the subsequent years.
This policy was announced in September 1999, but took effect in July 2000 to allow a 9-month “grace period”.
Specifically, an additional 2% loading is charged for every year above 30.
The variation in quality can be thought of as a reflection of waiting time, the quality of the room where the treatment is provided, or the quality of the actual treatment, e.g., a higher q means a more complete treatment (Besley et al. [1]).
Strictly speaking, in the empirical implementation of the model, we cannot call the model a demand model, since price (i.e., insurance premium) data are missing. Therefore, we will refer to it as a PHI purchase model instead. We thank John Creedy for pointing this out.
Table A1 in the appendix lists the variables used in the estimation of Eq. 1. The explanatory variables fall into seven categories, namely person descriptions, States/Territories, age, education, income and employment, health status, and health habits. Care has been taken to ensure that the variables are defined consistently across the 1995 and 2001 surveys.
We thank an anonymous referee for suggesting that we consider the use of this method.
There is the well-known index problem with this decomposition, since there is an equivalent alternative of the decomposition in the form of \( \bar Y^{01} - \bar Y^{95} = \left[ {\tfrac{1}{{N^{01} }}\sum\limits_{i = 1}^{N^{01} } {f\left( {X^{01} \hat \beta ^{01} } \right) - \tfrac{1}{{N^{95} }}\sum\limits_{i = 1}^{N^{95} } {f\left( {X^{95} \hat \beta ^{01} } \right)} } } \right] + \left[ {\tfrac{1}{{N^{95} }}\sum\limits_{i = 1}^{N^{01} } {f\left( {X^{95} \hat \beta ^{01} } \right) - \tfrac{1}{{N^{95} }}\sum\limits_{i = 1}^{N^{95} } {f\left( {X^{95} \hat \beta ^{95} } \right)} } } \right] \) which usually does not lead to the same decomposition estimates. Unfortunately, the recommended solution of using pooled estimates of \( \hat \beta \)s for estimating the first part of the decomposition (Oaxaca and Ransom [18]) is not possible for us, because the data for 2001 are available via remote access only. Furthermore, the decomposition expression in Eq. 3 is more intuitive because it asks what would have happened in 2001 in the absence of the reforms, using \( \hat \beta ^{95} \)as the weights, rather than asking what could have been in 1995 had the reforms already been implemented.
Strictly speaking, due to the lack of price (i.e., insurance premium) information, the empirical model that we estimated is not a demand model but an insurance purchase model.
The original 1995 and 2001 NHS data contain 53,828 and 26,862 records of individuals. Our sample sizes are considerably smaller, due to the elimination of records with incomplete and non-useable observations.
We distinguish between two types of households, single and family, for two reasons. First, as will be shown later, these two types of households behave differently in their decisions to purchase PHI. In addition, families have important characteristics that are not applicable to singles, e.g., family size and number of children.
The complete list of characteristics is provided in Table A2 in the appendix.
For families, the personal characteristics summarsed in Table 2 refer to those of the head of the household (husband/wife/partner/lone parent).
In the estimation, we applied the sampling weights provided in the NHS data.
Frech et al. [8] gave yet another estimate of the decline using trend analysis. See also Walker et al. [21] for further estimates based on trend analysis. These different estimates suggest that methodological differences do produce rather different values, and it is difficult to suggest whether one is superior to another.
The definition of equivalent income is provided in the Australian Bureau of Statistics [22].
We thank an anonymous referee for pointing this out.
The 95% confidence intervals for the predicted probabilities were computed on the basis of approximated standard errors via the delta method.
The lower and upper bounds were computed similarly, using the lower and upper bounds in Table 6.
References
Besley, T., Hall, J., Preston, I.: The demand for private health insurance: do waiting lists matter. J. Public Econ. 72, 155–181 (1999)
Propper, C.: Constrained choice sets in the UK demand for private medical insurance. J. Public Econ. 51(3), 287–307 (1993)
Besley, T., Coate, S.: Public provision of private goods and the redistribution of income. Am. Econ. Rev. 81, 979–984 (1991)
Wright, D.J.: Insurance and monopoly power in a mixed private/public hospital system. Econ. Rec. 82(259), 460–468 (2006)
Deeble, J.: The private health insurance rebate: report to state and territory health ministers. National Center for Epidemiology and Population Health, The Australian National University (2003)
Cormack, M.: Private health insurance: the problem child faces adulthood. Aust. Health Rev. 25(2), 38–51 (2002)
Butler, J.R.G.: Policy change and private health insurance: did the cheapest policy do the trick? Aust. Health Rev. 25(6), 33–41 (2002)
Frech III, H.E., Hopkins, S., MacDonald, G.: The Australian private health insurance boom: was it subsidies or liberalised regulation? Econ. Pap. 22(1), 58–64 (2003)
Palangkaraya, A., Yong, J.: Effects of recent carrot-and-stick policy initiatives on private health insurance coverage in Australia. Econ. Rec. 81(254), 262 (2005)
Barrett, G.F., Conlon, R.: Adverse selection and the decline in private health insurance coverage in Australia: 1989–1995. Econ. Rec. 79(246), 279–296 (2003)
Palangkaraya, A., Yong, J.: How effective is “lifetime health cover” in raising private health insurance coverage in Australia? An assessment using regression discontinuity. Appl. Econ. 39(11), 1361–1374 (2007)
Fairlie, R.W.: An extension of the Blinder–Oaxaca decomposition technique to logit and probit models. J. Econ. Soc. Meas. 30(4), 305–316 (2005)
Fairlie, R.W.: The absence of the African–American owned business: an analysis of the dynamics of self-employment. J. Labor Econ. 17(1), 80–108 (1999)
Blinder, A.S.: Wage discrimination: reduced form and structural variables. J. Hum. Resour. 8, 436–455 (1973)
Oaxaca, R.: Male–female wage differentials in urban labor markets. Int. Econ. Rev. 14(October), 693–709 (1973)
Jones, F.L.: On decomposing the wage gap: a critical comment on Blinder’s method. J. Hum. Resour. 18(1), 126–130 (1983)
Cain, G.G.: The economic analysis of labor market discrimination: a survey. In: Ashenfelter, O., Laynard, R. (eds.) Handbook of labor economics, vol. 1. Elsevier Science, Amsterdam (1986)
Oaxaca, R., Ransom, M.R.: On discrimination and decomposition of wage differentials. J. Econom. 61, 5–21 (1994)
Australian Bureau of Statistics (ABS): Census of population and housing, selected social and housing characteristics, 2015.0. ABS, Canberra (2001)
Oaxaca, R., Ransom, M.R.: Calculation of approximate variances for wage decomposition differentials. J. Econ. Soc. Meas. 24, 55–61 (1998)
Walker, A., Percival, R., Thurecht, L., Pearse, J.: Distributional impact of recent changes in private health insurance policies. Aust. Health Rev. 29(2), 167–177 (2005)
Australian Bureau of Statistics (ABS): 1996 census of population and housing: socio-economic indexes for areas. Information paper, ABS #2039.0. ABS, Canberra (1998)
Private Health Insurance Administration Council (PHIAC): Operations of the registered health benefits organisations annual report 2001–2002. PHIAC, Canberra (2002)
Harper, I.: Preserving choice: a defence of public support for private health care funding in Australia. Report prepared for Medibank Private Limited (2003)
Emmerson, C., Frayne, C., Goodman, A.: Should private medical insurance be subsidised? pp. 49–65. Health Care UK, Spring (2001)
Acknowledgements
This paper arose out of a consultancy project funded by the Victorian Department of Premier and Cabinet. The views expressed in this paper are those of the authors and do not represent the views of the Victorian Government. Peter Dawkins worked at the Melbourne Institute during the consultancy period. He now works at the Victorian Department of Treasury and Finance.
Author information
Authors and Affiliations
Corresponding author
Appendix
Appendix
Rights and permissions
About this article
Cite this article
Palangkaraya, A., Yong, J., Webster, E. et al. The income distributive implications of recent private health insurance policy reforms in Australia. Eur J Health Econ 10, 135–148 (2009). https://doi.org/10.1007/s10198-008-0111-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10198-008-0111-8