Skip to main content

Advertisement

Log in

Do healthcare tax credits help poor-health individuals on low incomes?

  • Original Paper
  • Published:
The European Journal of Health Economics Aims and scope Submit manuscript

Abstract

In several countries, personal income tax permits tax credits for out-of-pocket healthcare expenditure. Tax credits benefit taxpayers at all income levels by reducing their net tax liability and modify the price of out-of-pocket expenditure. To the extent that consumer demand is price elastic, they may influence the amount of eligible healthcare expenditure for which taxpayers may claim a credit. These effects influence, in turn, income distributions and taxpayers’ health status and therefore income-related inequality in health. Redistributive consequences of tax credits have been widely investigated. However, little is known about the ability of tax credits to alleviate health inequality. In this paper, we study the potential effects that tax credits for health expenses may have on income-related inequality in health status with reference to the Italian institutional setting. The analysis is performed using a tax-benefit microsimulation model that reproduces the personal income tax and incorporates taxpayers’ behavioral responses to changes in tax credit rate. Our results suggest that the current healthcare tax credit design tends to favor the richest part of the population.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1

Similar content being viewed by others

Notes

  1. IRPEF is currently the main tool for income redistribution policies. IRPEF accounts for around one-third of overall government tax revenues. IRPEF tax relief is provided for costs with a particular social relevance, such as those paid by taxpayers for health reasons: health costs of any type (doctor’s fees, specialist’s fees, surgery costs, pharmaceutical costs, prescription medications, and appliances such as glasses, hearing aids, etc.) qualify for a 19% tax credit.

  2. The broad list of eligible health expenses is prescribed in the legislation.

  3. Exemption from IRPEF is determined by a universal tax credit granted for specific income sources: the tax credit is applicable for either employment income or self-employment income, or pension income, with a withdrawal rate resulting in a decreasing credit as gross income increases. This tax credit contributes to the income tax progressivity design, even more so given the absence of a legal zero rate tax bracket. The no-tax hurdle is: €8000 per year for subordinate workers; €7500 for pensioners under 75 years of age; €7750 for pensioners aged 75 or older; €4800 for the self-employed. Furthermore, the no-tax hurdle increases further if there are dependent family members.

  4. Typically, the individual demand for health care services under full insurance coverage regime tends to be inelastic (for details, see [38, 50, 24]).

  5. It is worth noting that our sample (which is 39% male) also consists of individuals whose average age is quite high: 60 years old. This is not surprising, since we only considered individuals without dependent relatives.

  6. In order to obtain the expenditure in real terms, we employed a health-care specific deflator. Accordingly, we also deflated the income indicator by using the consumer price index.

  7. BETAMOD estimates, at the individual level, the conditional probability of incurring tax-relevant healthcare expenditures as a function of individual characteristics known to be predictive of health expenses, such as sex, age, health status, geographical region, marital status, income, occupation, and education. Next, BETAMOD uses the probability of healthcare spending together with fiscal data on tax relief (Ministry of Economics and Finance, 2010) to identify beneficiaries of healthcare tax relief, and to impute related amounts of expenditure (for details, see [1]).

  8. The consumer price index concerns goods and services used, such as pharmaceuticals, visits to doctors and specialists, medical services, dentistry, clinical analysis, and diagnostic tests. We assign to each individual in our sample the price index according to her region of residence, so that the price index captures regional variation only and does not account for other aspects.

  9. To obtain the annual “equivalent household income”, we divided the household disposable income by its “equivalent size”, which is calculated using the “modified OECD” equivalence scale. This scale gives a weight of 1.0 to the first adult, 0.5 to any other household member aged 14 and over, and 0.3 to each child under 14.

  10. We also tried a different specification including quadratic terms in income and health in the right-hand side of healthcare expenditure equation. For health indicator, we found no evidence of a quadratic relationship. The square of income is statistically significant but the inclusion of the square income does not significantly affect the results: the price elasticity coefficient remains very similar to the baseline model.

  11. For a Heckman model to work, it requires exclusion restrictions, i.e., regressors that enter the selection part, but not the second part of the model. We have included in the reduced form a proxy of barriers to access health care services. We have constructed an interaction term between a dummy variable that takes a value of 1 if the respondent lives in the south of Italy and a dummy variable, which indicates whether the respondent lives in a thinly populated area. In the south of Italy, in particular, the lack of financial means is one of the most relevant barriers to healthcare access [11]. For the sake of brevity, we do not include the results of the Heckman two-step estimation in the paper, but they are available upon request.

  12. By “other standard-of-living information” we mean assets, housing (water, electricity, and gas bills), fuel, clothing, whether the home is owned, number of rooms per household member, overall size of dwelling (i.e., the number of square meters per person) and a battery of items on possessions in the home. These possessions include household items such as a television, satellite dish, mobile phone, computer, Internet access, hi-fi stereo, camera, washing machine, dishwasher, air conditioning, and a car [49, [39]. According to the previous literature, housing in particular is a core element of people’s material living standards. Housing conditions may strongly influence people’s health and quality of life (see [5]).

  13. PCA transforms the original set of variables into a smaller set of linear combinations that accounts for most of the variance of the original set. For a detailed discussion on how to construct asset indices, see [49].

  14. We also rescaled the index by adding a constant, which was the minimum whole number required to eliminate negative values. This rescaling does not affect the contribution of each variable to the concentration index, since the rank order remains unchanged.

  15. Note that, in contrast to C(H), the Erreygers index does not depend on the mean of health. Moreover, while the standard concentration index may give conflicting information when applied separately to good health and poor health, the Erreygers index satisfies the so-called “mirror property”, namely the inequalities in health “mirror” those in poor health [14].

  16. In the Canadian context, Smart and Stabile [44] have found price elasticities in the range of –0.27 to –0.9 across different categories of medical care spending. A review of the empirical literature on price and income elasticity of the demand for health insurance and healthcare services is given in [35].

  17. In order to check the robustness and the extensibility of our results, we have simulated the tax price elasticity for different family types: singles vs. couples. The elasticity of healthcare expenditure with respect to after-tax price is very similar (respectively −0.71915 and −0.71566) and the difference is not statically significant.

  18. IRPEF consists of five brackets, with the lowest rate (23%) applied to personal income up to €15,000 per year and the highest rate (43%) for marginal income above €75,000 per year.

References

  1. Albarea, A., Bernasconi, M., Di Novi, C., Marenzi, A., Rizzi, D., Zantomio, F.: Accounting for tax evasion profiles and tax expenditures in microsimulation modelling. The BETAMOD model for personal income taxes in Italy. Int. J. Microsimul. 8, 99–136 (2015)

    Google Scholar 

  2. Atella, V., Borgonovi, E., Collicelli, C., Kopinska, J., Lecci, F., Maietta, F.: Crisi economica, diseguaglianze nell’accesso ai servizi sanitari ed effetti sulla salute delle persone in Italia. I Quaderni della Fondazione Farmafactoring, No. 01.2015, (2015)

  3. Bago d’Uva, T.E., Van Doorslaer, Lindeboom, M., Lindeboom, O.: Does reporting heterogeneity bias the measurement of health disparities? Health Econ. 17, 351–375 (2008)

    Article  PubMed  Google Scholar 

  4. Baker, M., Stabile, M., Deri, C.: What do self-reported, objective, measures of health measure? J. Human Resour. 4, 1067–1093 (2004)

    Article  Google Scholar 

  5. Balestra, C., Sultan, J.: Home sweet home: the determinants of residential satisfaction and its relation with well-being. OECD Statistics Directorate Working Papers, No. 5, (2013)

  6. Balia, S., Jones, A.M.: Mortality, lifestyle and socio-economic status. J. Health Econ. 27, 1–26 (2008)

    Article  PubMed  Google Scholar 

  7. Benzeval, M., Judge, K., Whitehead, M.: Tackling inequalities in health: an agenda for action. King’s Fund, London (1995)

    Google Scholar 

  8. Besley, T., Hall, J., Preston, I.: The demand for private health insurance: do waiting lists matter? J. Publ. Econ. 72, 155–181 (1999)

    Article  Google Scholar 

  9. Burman, L.E.: Is the tax expenditures concept still relevant? Natl. Tax J. LVI, 613–627 (2003)

    Article  Google Scholar 

  10. Burman, L.E., Toder, E., Geissler, C.: How big are total individual income tax expenditures, and who benefits from them? Am. Econ. Rev. 98, 79–83 (2008)

    Article  Google Scholar 

  11. Cavalieri, M.: Geographical variation of unmet medical needs in Italy: a multivariate logistic regression analysis. Int. J. Health Geograph. 12, 27 (2013)

    Article  Google Scholar 

  12. Contoyannis, P., Jones, A.M.: Socio-economic status, health and lifestyle. J. Health Econ. 23, 965–995 (2004)

    Article  PubMed  Google Scholar 

  13. Contoyannis, P., Jones, A.M., Rice, N.: The dynamics of health in the British Household Panel Survey. J. Appl. Econ. 19, 473–503 (2004)

    Article  Google Scholar 

  14. Costa-Font, J., Hernandez-Quevedo, C., Jimenez-Rubio, D.: Income inequalities in unhealthy life styles in England and Spain. Econ. Human Biol. 13, 66–75 (2014)

    Article  Google Scholar 

  15. Crivelli, L., Salari, P.: The inequity of the Swiss health care system financing from a federal state perspective. Int. J. Equity Health 13, 1–13 (2014)

    Article  Google Scholar 

  16. de Belvis, A.G., Ferrè, F., Specchia, M.L., Valerio, L., Fattore, G., Ricciardi, W.: The financial crisis in Italy: implications for the healthcare sector. Health Policy 10, 10–16 (2012)

    Article  Google Scholar 

  17. Deaton, A.: Health, inequality, and economic development. J. Econ. Lit. 41, 113–158 (2003)

    Article  Google Scholar 

  18. Di Novi, C.: The Influence of traffic-related pollution on individuals’ life-style: results from the BRFSS. Health Econ. 19, 1318–1344 (2010)

    Article  PubMed  Google Scholar 

  19. Dirindin, N.: Chi paga per la salute degli italiani?. Il Mulino, Bologna (1996)

    Google Scholar 

  20. Dixon, H., Siciliani, L.: Waiting-time targets in the healthcare sector: how long are we waiting? J. Health Econ. 28, 1081–1098 (2009)

    Article  PubMed  Google Scholar 

  21. Erreygers, G.: Correcting the concentration index. J. Health Econ. 28, 504–515 (2009)

    Article  PubMed  Google Scholar 

  22. Fattore, G., Mariotti, G., Rebba, V.: Review of waiting times policies: country case studies, Italy. In: Siciliani, L., Borowitz, M., Moran, V. (eds.) Waiting time policies in the health sector—what works?, Chapter 9, OECD Health Policy Studies (2013)

  23. France, G., Taroni, F., Donatini, A.: The Italian health-care system. Health Econ. 14, 187–202 (2005)

    Article  Google Scholar 

  24. Getzen, T.: Health care is an individual necessity and a national luxury: applying multilevel decision models to the analysis of health care expenditures. J. Health Econ. 19, 259–270 (2000)

    Article  CAS  PubMed  Google Scholar 

  25. Grossman, M.: On the concept of health capital and the demand for health. J. Polit. Econ. 80, 223–255 (1972)

    Article  Google Scholar 

  26. Heckman, J.J.: Sample selection bias as a specification error. Econometrica 47, 153–161 (1979)

    Article  Google Scholar 

  27. Idler, E.L., Benyamini, Y.: Self-rated health and mortality: a review of twenty-seven community studies. J. Health Soc. Behav. 38, 21–37 (1997)

    Article  CAS  PubMed  Google Scholar 

  28. ISTAT: Le differenze nel livello dei prezzi al consumo tra i capoluoghi delle regioni italiane. (http://www.istat.it/it/archivio/6279) (2010)

  29. ISTAT: Health for all. (http://www.istat.it/it/archivio/14562) (2015)

  30. Jakobsson, U.: On the measurement of the degree of progression. J. Public Econ. 5, 161–168 (1976)

    Article  Google Scholar 

  31. Kaplow, L.: The income tax as insurance: the casualty loss and medical expense deductions and the exclusion of the medical insurance premiums. NBER Working Paper, No. 3723, (1991)

  32. Kenkel, D.S.: The demand for preventive medical care. Appl. Econ. 26, 313–325 (1994)

    Article  Google Scholar 

  33. Kennedy, B.P., Kawachi, I., Glass, R., Prothrow-Stith, D.: Income distribution, socio-economic status, and self-rated health in the United States: multilevel analysis. Br. Med. J. 317, 917–921 (1998)

    Article  CAS  Google Scholar 

  34. Kohn, J.L.: What is health? A multiple correspondence health index. East. Econ. J. 38, 223–250 (2012)

    Article  Google Scholar 

  35. Liu, S., Chollet, D.: Price and income elasticity of the demand for health insurance and health care services: a critical review of the literature. http://www.mathematica-mpr.com/publications/pdfs/priceincome.pdf (2006)

  36. Mackenbach, J.: The persistence of health inequalities in modern welfare states: the explanation of a paradox. Soc. Sci. Med. 75, 761–769 (2012)

    Article  PubMed  Google Scholar 

  37. Ministry of Economy and Finance: (http://www.dt.tesoro.it/en) (2010 and 2015)

  38. Newhouse, J.P., Phelps, C.E.: Price and income elasticities of the demand for medical care services. In: Perlman, M. (ed.) The economics of health and medical care, pp. 139–161. Macmillan, London (1974)

    Chapter  Google Scholar 

  39. O’Donnell, O., van Doorslaer, E., Wagstaff, A., Lindelow, M.: Analyzing health equity using household survey data: a guide to techniques and their implementation. World Bank Publications, No. 434, Washington (2008)

  40. OECD, Health Statistics: FOCUS on health spending. Available at http://www.oecd.org/els/health-systems/health-data.htm (2015)

  41. Prasad, N.: Policies for redistribution: the use of taxes and social transfers. International Institute for Labour Studies Discussion Paper, No. 194 (2008)

  42. Poterba, J.M.: Economic analysis of tax expenditures. Natl. Tax J. 64, 451–458 (2011)

    Article  Google Scholar 

  43. Ramalho, E.A., Ramalho, J.J.S., Enriques, P.D.: Fractional regression models for second stage DEA efficiency analyses. J. Prod. Anal. 34, 239–255 (2010)

    Article  Google Scholar 

  44. Smart, M., Stabile, M.: Tax credits, insurance, and the use of medical care. Can. J. Econ. 38, 345–365 (2005)

    Article  Google Scholar 

  45. Tyson, J.: Reforming tax expenditures in Italy: what, why, and how?. IMF Working Paper, January 2014 (2014)

  46. Toder, E., Harris, B.H., Lim, K.: Distributional effects of tax expenditures in the United States. In: Philipps, L., Brooks, N., Li, J. (eds.) Tax expenditures: state of the art. Canadian Tax Foundation, Toronto (2011)

    Google Scholar 

  47. Toder, E., Baneman, D.: Distributional effects of individual income tax expenditures: an update. Urban-Brookings Tax Policy Center, February 2 (2012)

  48. Undén, A.L., Elofsson, S.: Do different factors explain self-rated health in men and women? Gend. Med. 3, 295–308 (2006)

    Article  PubMed  Google Scholar 

  49. Vyas, S., Kumaranayake, L.: Constructing socioeconomic status indices: how to use principal components analysis. Adv. Access Publ. 9, 459–468 (2006)

    Google Scholar 

  50. Wagstaff, A.: The demand for health theory and applications. J. Epidemiol. Community Health 40, l–11 (1986)

    Article  Google Scholar 

  51. Wagstaff, A., van Doorslear, E., Paci, P.: On the measurement of horizontal inequity in the delivery of health care. J. Health Econ. 10, 169–205 (1991)

    Article  CAS  PubMed  Google Scholar 

  52. Wagstaff, A., van Doorslaer, E.: Measuring and testing for inequity in the delivery of health care. J. Human Resour. 35, 716–733 (2000)

    Article  Google Scholar 

  53. Wagstaff, A., van Doorslaer, E.: What makes the personal income tax progressive? A comparative analysis for fifteen OECD countries. Int. Tax Publ. Finance 8, 299–315 (2001)

    Article  Google Scholar 

  54. Wilkinson, R.G., Pickett, K.E.: Income inequality and population health: a review and explanation of the evidence. Soc. Sci. Med. 62, 1768–1784 (2006)

    Article  PubMed  Google Scholar 

Download references

Acknowledgements

An earlier draft of this paper was presented at the annual meeting of the Italian Public Economic Associations, Lecce, Italy, and at the annual meeting of the Italian Health Economics Association, Bologna, Italy. The authors wish to thank the conference participants for their detailed and helpful comments. The authors in particular wish to thank Andrea Albarea for his assistance in gathering and processing the data. The paper benefitted from comments from Michele Bernasconi, Enrica Croda, Gianmaria Martini and Francesca Zantomio. The usual disclaimer applies.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Cinzia Di Novi.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (DOC 179 kb)

Appendices

Appendix A

See Fig. 1.

Appendix B

The health index

IT-SILC provides three different measures of health: self-assessed health (SAH), the presence of chronic diseases, and the presence of conditions limiting daily activities.

SAH is measured with a five-point categorical variable following the conventional levels recommended by the World Health Organization: “very poor”, “poor”, “fair”, “good”, and “very good”. Hence, the self-assessed health indicator has been included as a five-category ordered variable (ranging from very poor to very good health). SAH has been widely used in previous studies that examined the relationship between health and socioeconomic status (e.g., [32, 12, 18]). SAH is supported by a body of literature that shows a strong predictive relationship between people’s self-ratings of their own health and mortality or morbidity [27, 33]. Moreover, SAH correlates strongly with more complex health indices such as functional ability or indicators derived from health service usage [48]. SAH is a subjective measure of health that may involve biases in the measurement of inequalities. Indeed, SAH may be systematically correlated with characteristics such as sex, age, income level, or education. SAH may also be subject to measurement errors caused by the poor design of questionnaires, misunderstood concepts, inadequately trained interviewers, different conceptions of health in general, different expectations for own health or financial incentives to report ill health (see [13, 3, 34] for a discussion of biases associated with self-assessed health).

In order to support the reliability of our measure of health inequality, we also employ two objective (albeit self-reported) functional measures of health: limitations to activities of daily living because of health problems (ADL) and an indicator reporting a chronic (long-standing) illness or condition. The question about global chronic diseases asked respondents whether they had suffered from or had any chronic (long-standing) illness or conditions in the form of health problems. Answers to this question were kept in its original binary coding (“no” or “yes”). Hence, the chronic condition indicator is a dummy with a value of 1 if a person mentions a chronic illness, and 0 otherwise. Finally, respondents were asked if they had been limited in activities, which people normally do, in the last 6 months due to health problems. The answer categories ranged from “severely limited”, “limited but not severely” to “not limited at all”. Hence, we use an ordinal scale of 3 points, from 3 (not limited) to 1 (severely limited). Table 9 shows how each indicator is distributed in our sample.

Table 9 Health index components (%)

Our choice is explained by the observation that “the specificity of the questions constrains the likelihood that respondents rationalize their own behavior through their answer” [4].

Finally, in order to have a single number that reflects overall health, we constructed a health index through multiple correspondence analysis (MCA), which reduces the multiple discrete indicators described above into a continuous variable. MCA is more appropriate than other techniques such as principal component analysis (PCA) when constructing an index based on ordered categorical variables. Empirical evidence suggests, in fact, that answers to the SAH question, for instance, cannot be scored on a simple scale from 1 to 5 because the true scale will not be equidistant between categories. If PCA was used on an ordered categorical variable such as SAH, or for other discrete or binary indicators of health problems (such as ADL or the presence of chronic illness), the underlying assumption would be that individuals consider the distance between the categories to be equivalent (for details, see [34]). Therefore, the health index has been computed from the weights for each measure of health using row scores based on the indicator matrix of MCA. We also standardize the index to lie on a continuous scale between 0 (poorest health) to 1 (best health) to aid in interpretation.

Appendix C

Healthcare expenditure concentration index

Table 10 shows the healthcare expenditure concentration index in the five policy scenarios proposed in our paper.

Table 10 Healthcare expenditure concentration index

The results are in line with those presented in the paper, which concerned health inequality. Again, tax relief decreasing in income is conducive to reducing health inequality and also presents a better redistributive effect with respect to the baseline.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Di Novi, C., Marenzi, A. & Rizzi, D. Do healthcare tax credits help poor-health individuals on low incomes?. Eur J Health Econ 19, 293–307 (2018). https://doi.org/10.1007/s10198-017-0884-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10198-017-0884-8

Keywords

JEL Classification

Navigation