Causal effect of education on cancer risk and survival

Cecilia Potente, University of Zurich and Christiaan Monden and Patrick Präg, University of Oxford

(Project no. 1007948)

Introduction

Cancer is becoming an increasingly common disease especially due to the aging population and the reduction in infectious diseases fatalities. Over time many more people suffer from it (Office for National Statistics 2016). Previous studies have documented the existence of an association between educational attainment and cancer risk and survival (Ø. Kravdal 2000; Woods et al. 2006). Education might influence cancer incidence and survival through greater access to resources in terms of prevention and cancer treatments. However, only few studies have tried to disentangle the causation from the selection argument for the effect of education on cancer (Buckles et al. 2013; Lager and Torssander 2012; Leuven et al. 2016; Palme and Simeonova 2015). My work has the aim of analyzing the causal effect of education on cancer incidence and survival in England and Wales. I exploit the exogenous increase in education due to the compulsory schooling law changes in 1947 and 1972 in England and Wales. This analysis takes advantage of these two reforms in order to identify the causal effect of education on cancer risk and survival.

The 1947 and 1972 Reforms

Compulsory school laws in Britain define the maximum age at which children should start education and minimum age at which they are allowed to leave education. Consequently, these laws determine how many years children spend in compulsory schooling. During the 20th century two legal changes have risen the minimum ages at which children were allowed to leave school. First, the 1944 Educational Act established an increase in leaving school age from 14 to 15. This change was implemented from the 1st of April 1947. Second, the same Education Act authorized a further increase in school leaving age from 15 to 16 which was implemented starting from 1st of September 1972. The economic and social contexts surrounding these two changes were quite different offering a more complete view of the causal effect of education on cancer risk and survival. Therefore, it results interesting to explore the causal effects of education on cancer for reforms happening in different time periods. Compulsory school law changes have been used to identify the effect of education on cancer incidence in Norway and Sweden (Lager and Torssander 2012; Leuven et al. 2016; Palme and Simeonova 2015). However, England and Wales have much more stratified societies where social inequalities in health are larger than in Nordic countries (Mackenbach et al. 2008, 2015). It remains an open question whether different institutional settings affect differently education on health.

The two aforementioned reforms in England and Wales have been used for identifying the causal effect of education on different outcomes such as earnings, fertility, assortative mating, health and mortality (Clark and Royer 2013; Geruso and Royer 2014; Oreopoulos 2006). The exogenous increase in education has been associated with higher earnings, especially for males (Devereux and Hart 2010; Grenet 2013; Oreopoulos 2006). Geruso and Royer (2014) have also found a reduction in teen pregnancies but not in completed fertility for women affected by the 1972 reform. In addition, they have found that the reform has led individuals to find more educated and younger partners. Concerning health and mortality, some studies have not found significant effects of the reform (Braakmann 2011; Clark and Royer 2013). On the contrary, Silles (2009) has identified a protective effect of education for health using the same reforms. This protective effect is in line with previous research uncovering the presence of a causal effect of education on health (Fletcher 2015; Lleras-muney 2005; Oreopoulos 2006). In summary, the causal effect of education on health does not seem to be univocally supported across different institutional contexts and identification strategies (Mazumder 2011). These mixed results leave open the debate on the causal effect of education on health and mortality. Therefore, this work has the aim of entering in the debate providing causal evidence for education on cancer risk and survival in England and Wales using the 1947 and 1972 compulsory school reforms.

Socioeconomic Inequalities in Cancer Risk and Survival

The association between cancer risk and education is highly dependent on the type of cancer. On one side, higher educational attainment reduces the risk of development of some cancers, such as colorectal, cervical and lung; on the other side, the opposite educational gradient is observable for skin, testicular, prostate and breast cancer (Braaten et al. 2005; Hemminki and Li 2003; Menvielle and Kunst 2008). Finally, leukemia and lymphoma incidences are not associated in any direction with education (Hermann et al. 2010).

Socioeconomic differences in cancer survival are more pronounced than the ones in cancer incidence (Braaten et al. 2009; H. Kravdal 2014; Sloggett et al. 2007). Sloggett and colleagues (2007) have examined the socioeconomic gradient in cancer survival in England and Wales using three individual-level measures of socioeconomic status (social class, housing tenure, and household access to a car) and one area-level indicator (carstairs index). They have found lower survival for individuals in disadvantageous socioeconomic conditions. Previous studies on cancer survival and its relationship with different socioeconomic indicators also have identified a general positive association (Ø. Kravdal 2000; Woods et al. 2006). However, so far they haven't been completely able to distinguish the causal argument from the selection one.

Data

The data source is the ONS Longitudinal Study (ONS LS) for England and Wales. ONS LS contains census linked information concerning vital events for 1% of the population of England and Wales from 1971 to 2011. Events registered are births, deaths, immigrations, emigrations, and cancer. In particular, I make use of the linkage with National Cancer Registry which records several information concerning the age at diagnosis, stage and type of cancer. Data are also linked to death records allowing for examining cancer survival. The sample members followed longitudinally across different censuses (1971-2011) are around 500 thousand. Among those, 135 thousand experience a cancer event and they have information concerning their diagnosis and survival to the disease.

The value of this project is highly public relevance because it might inform policy-makers on the effect of increasing educational attainment for important health outcomes as cancer. Evidence on this area of research are still scarce, and it would an important step forward to learn more about the relationship between cancer and education.

Empirical Strategy

The aim of this work is the identification of the causal effect of education on cancer risk and survival using changes in compulsory school laws in England and Wales. The identification is achieved through a regression discontinuity approach (Imbens and Lemieux 2008; Lee and Lemieux 2010). In the first stage it is important to show the effect of the changes in these laws on educational outcomes. The first stage is estimated by Clark and Royer (2013). The reforms have a clear effect on educational attainment of different individuals. The second stage examines the impact of compulsory schooling changes on cancer incidence and survival using a reduced form estimation. First, in order to study how education is causally related to cancer risk and survival, I estimate multistate hazard models: the hazard of developing cancer and subsequently the hazard of death are estimated using age and month-year birth cohort fixed effects. Second, once these month-year birth cohort fixed effects are calculated, I use them as outcome variables in local linear regression models. From these local linear regressions the discontinuous changes in the hazard of developing and surviving cancer are estimated. These are going to be our parameter of interest.

Discussion

This analysis contributes the existing literature in three ways. First, it provides additional evidence on the causal effect of education on health focusing on cancer incidence and survival. Mixing evidence have been found concerning the effect of an increase in compulsory schooling on health (for review see Mazumder 2011). The current work offers a more detailed picture on the topics by examining the effect on cancer risk and survival. Second, the literature so far has examined this effect only in Nordic countries where the compulsory school reforms were accompanied by other changes in the educational systems (Lager and Torssander 2012; Leuven et al. 2016; Palme and Simeonova 2015). The study could provide causal evidence of the effect of education on cancer in a different institutional setting, England and Wales. Finally, the focus on both incidence and survival helps to shed light on the mechanisms connecting education and cancer. Understanding better causes of educational inequalities in cancer is extremely important given the increasing number of people affected.

References

Braakmann, N. (2011). The causal relationship between education, health and health related behaviour: Evidence from a natural experiment in England. Journal of Health Economics, 30(4), 753-763. http://doi.org/10.1016/j.jhealeco.2011.05.015

Braaten, T., Weiderpass, E., Kumle, M., & Lund, E. (2005). Explaining the socioeconomic variation in cancer risk in the Norwegian Women and Cancer Study. Cancer Epidemiology Biomarkers and Prevention, 14(11 I), 2591-2597. http://doi.org/10.1158/1055-9965.EPI-05-0345

Braaten, T., Weiderpass, E., & Lund, E. (2009). Socioeconomic differences in cancer survival: the Norwegian Women and Cancer Study. BMC Public Health, 9, 178. http://doi.org/10.1186/1471-2458-9-178

Buckles, K., Hagemann, A., Malamud, O., Morrill, M. S., & Wozniak, A. K. (2013). The effect of college education on health. NBER Working Paper Series, (19222), 54.

Clark, D., & Royer, H. (2013). The Effect of Education on Adult Mortality and Health: Evidence from Britain. American Economic Review, 103(6), 2087-2120. http://doi.org/10.1257/aer.103.6.2087

Devereux, P., & Hart, R. A. (2010). Forced to Be Rich? Returns to Compulsory Schooling in Britain. The Economic Journal, 120(549), 1345-1364. http://doi.org/10.1111

Fletcher, J. M. (2015). New evidence of the effects of education on health in the US: Compulsory schooling laws revisited. Social Science and Medicine, 127, 101-107. http://doi.org/10.1016/j.socscimed.2014.09.052

Geruso, M., & Royer, H. (2014). The Impact of Education on Family Formation: Quasi-Experimental Evidence from the UK. University of California, mimeo. Retrieved from http://www.ucdenver.edu/academics/colleges/CLAS/Departments/economics/Se... 2014/geruso_royer.pdf

Grenet, J. (2013). Is Extending Compulsory Schooling Alone Enough to Raise Earnings? Evidence from French and British Compulsory Schooling Laws. Scandinavian Journal of Economics, 115(1), 176-210. http://doi.org/10.1111/j.1467-9442.2012.01739.x

Hemminki, K., & Li, X. (2003). Level of Education and the Risk of Cancer in Sweden. Cancer Epidemiology , Biomarkers and Prevention, 12(August), 796-802.

Hermann, S., Rohrmann, S., Linseisen, J., Nieters, A., Khan, A., Gallo, V., … Riboli, E. (2010). Level of education and the risk of lymphoma in the European prospective investigation into cancer and nutrition. Journal of Cancer Research and Clinical Oncology, 136(1), 71-77. http://doi.org/10.1007/s00432-009-0638-9

Imbens, G. W., & Lemieux, T. (2008). Regression discontinuity designs: A guide to practice. Journal of Econometrics, 142(2), 615-635. http://doi.org/10.1016/j.jeconom.2007.05.001

Kravdal, H. (2014). Widening educational differences in cancer survival in Norway. European Journal of Public Health, 24(2), 270-275. http://doi.org/10.1093/eurpub/ckt082

Kravdal, Ø. (2000). Social inequalities in cancer survival. Population Studies: A Journal of Demography, 54(1), 1-18. http://doi.org/10.1080/713779066

Lager, A. C. J., & Torssander, J. (2012). Causal effect of education on mortality in a quasi-experiment on 1.2 million Swedes. Proceedings of the National Academy of Sciences, 109(22), 8461-8466. http://doi.org/10.1073/pnas.1105839109

Lee, D. S., & Lemieux, T. (2010). Regression Discontinuity Designs in Economics. Journal of Economic Literature, 48(2), 281-355. http://doi.org/10.1257/jel.48.2.281

Leuven, E., Plug, E., & Rønning, M. (2016). Education and cancer risk. Labour Economics, 43, 106-121. http://doi.org/10.1016/j.labeco.2016.06.006

Lleras-muney, A. (2005). The Relationship Between Education and Adult Mortality in the United States. Review of Economic Studies, 72, 189-221.

Mackenbach, J. P., Kulhánová, I., Bopp, M., Deboosere, P., Eikemo, T. A., Hoffmann, R., … White, C. (2015). Variations in the relation between education and cause-specific mortality in 19 European populations: A test of the "fundamental causes" theory of social inequalities in health. Social Science and Medicine, 127, 51-62. http://doi.org/10.1016/j.socscimed.2014.05.021

Mackenbach, J. P., Stirbu, I., Roskam, A. R., Schaap, M. M., Menvielle, G., Leinsalu, M., & Kunst, A. (2008). Socioeconomic inequalities in health in 22 European countries. The New England Journal of Medicine, 358(23), 2468-81. http://doi.org/10.1056/NEJMc081414

Mazumder, B. (2011). The Effects of Education on Health and Mortality. Nordic Economic Policy Review.

Menvielle, G., & Kunst, A. (2008). Social inequalities in cancer incidence and cancer survival: Lessons from Danish studies. European Journal of Cancer, 44(14), 1933-1937. http://doi.org/10.1016/j.ejca.2008.06.034

Office for National Statistics. (2016). Cancer Registration Statistics, England: 2014. Statistical Bulletin.

Oreopoulos, P. (2006). Estimating Average and Local Average Treatment Effects of Education When Compulsory Schooling Laws Really Matter. American Economic Review, 96(1), 152-175.

Palme, M., & Simeonova, E. (2015). Does women's education affect breast cancer risk and survival? Evidence from a population based social experiment in education. Journal of Health Economics, 42, 115-124. http://doi.org/10.1016/j.jhealeco.2014.11.001

Silles, M. a. (2009). The causal effect of education on health: Evidence from the United Kingdom. Economics of Education Review, 28(1), 122-128. http://doi.org/10.1016/j.econedurev.2008.02.003

Sloggett, A., Young, H., & Grundy, E. (2007). The association of cancer survival with four socioeconomic indicators: a longitudinal study of the older population of England and Wales 1981-2000. BMC Cancer, 7, 20. http://doi.org/10.1186/1471-2407-7-20

Woods, L. M., Rachet, B., & Coleman, M. P. (2006). Origins of socio-economic inequalities in cancer survival: A review. Annals of Oncology, 17(1), 5-19. http://doi.org/10.1093/annonc/mdj007