class: center, middle, inverse, title-slide .title[ # The Impact of Chronic Diseases on Early Retirement: ] .subtitle[ ## A Difference-in-Differences Analysis ] .author[ ### Nga Chau My Ha^{1, 2}, Diana Pacheco B.^{1, 2} ] .institute[ ### ¹ University of Lucerne, ² Swiss Paraplegic Research ] .date[ ### 2024-08-29 ] --- background-image: url("images/cool.png") # Background There are two streams leading to the question of research -- .pull-left[ #### Stream 1 - The share of older workers has been increasing after many pension reforms in the Europe¹. - However, the reported actual and expected retirement ages are consistently lower than the statutory retirement age^{2, 3}. ] -- .pull-right[ #### Stream 2 - Poor health was strongly evident in its negative impact on individuals' working lives^4-7. - The increasing share of older workers is facing a rising health burden, mainly coming from chronic diseases (CDs)⁸. ] --- background-image: url(https://upload.wikimedia.org/wikipedia/commons/1/1c/Background_proj1.png) background-size: 900px background-position: center # Research question <div style="text-align: center;"> "Whether and how chronic diseases affect the decision of early retirement among workers aged 55 and over across European countries?" </div> --- background-image: url("images/cool.png") # Methodology -- .pull-left[ ### Objectives Estimate 02 outcomes: 1. The probability of early retirement (ER): P(ER = 1 | CD = 1) 1. The working years lost to ER: lost_years = NRA - ERA due to having CDs. ] .footnote[ **CD**: chronic disease; **ER**: early retirement; **ERA**: early retirement age; **NRA**: national retirement age ] -- .pull-right[ ### Data Longitudinal data from 12 European countries was collected from the Survey of Health, Ageing, and Retirement in Europe (SHARE) from 2004 to 2020. ] --- background-image: url("images/cool.png") # Methodology .pull-left[ ### Inclusion criteria - individuals aged from 55 to the national retirement age (NRA), and - initially staying in the labor market without CDs (employed or not) - excluding those with unknown retirement status. ] .footnote[ **CD**: chronic disease; **ER**: early retirement; **ERA**: early retirement age; **NRA**: national retirement age ] -- .pull-right[ ### Variables - Having CDs: **D** = 1 - Group indicator: `$$\mathcal{G}_g = \mathbf{1}\{G = g\}$$`(g: the wave at first diagnosis) - The never-treated: **C** = 1 - Having ER: **ER** = 1 - Lost years: `$$\text{Lost_years} = \text{NRA} - \text{ERA}$$` - Control variables (X): age, sex, couple, yedu, income percentile, country. ] --- background-image: url("images/cool.png") # Methodology ### Empirical approach - Applied the DiD with a staggered treatment and multiple time periods approach, as described by Callaway and Sant’ Anna (2021)⁹ to estimate the average treatment effect of having CDs for each particular sub-group at a specific time point t: **ATT(g, t)**. - **ATT(g, t)** are then selected and weighted aggregated to estimate the overall average treatment effect (ATT^O) or the event-study results (i.e., the average treatment effect among those with the same length of exposure). -- ####Technical note: - Method for ATT(g, t) estimation: inverse propability weighting (“ipw”); - Control group: never treated group (“nevertreated”); - Base period: the period right before having diagnosis (“universal”); - Bootstrap is used to compute SE (1000 interations); - Cluster by household ID. --- background-image: url("images/cool.png") # Results ## Outcome 01 #### Global estimation - Overall, people with CDs have, on average, a 16% higher risk of retiring early <img src="https://upload.wikimedia.org/wikipedia/commons/a/a2/Proj01_pic01.png" width="60%" style="display: block; margin: auto;" /> --- background-image: url("images/cool.png") # Results ## Outcome 01 #### Country-specific estimation - Having chronic diseases triggers the decision to retire early in all observed countries; however, the effect sizes and patterns vary across nations. <img src="https://upload.wikimedia.org/wikipedia/commons/d/de/Proj01_pic02.png" width="90%" style="display: block; margin: auto;" /> --- background-image: url("images/cool.png") # Results ## Outcome 2 - Overall, people with CDs lose, on average, 3.7 working months to early retirement <div class="datatables html-widget html-fill-item" id="htmlwidget-4f5788f5b1171bf9a67f" style="width:100%;height:auto;"></div> <script type="application/json" data-for="htmlwidget-4f5788f5b1171bf9a67f">{"x":{"filter":"none","vertical":false,"data":[["1","2","3","4","5","6","7","8","9","10","11","12"],["Slovenia","Belgium","Germany","Italy","Austria","Switzerland","Denmark","Spain","Sweden","Czech Republic","Estonia","France"],["0.99 = 11.9 months","0.60 = 7.2 months","0.50 = 6.0 months","0.45 = 5.4 months","0.35 = 4.2 months","0.23 = 2.8 months","0.17 = 2.0 months","0.17 = 2.0 months","0.14 = 1.7 months","0.10 = 1.2 months","0.03 = 0.4 months","0.04 = 0.5 months"],["0.61 - 1.38","0.36 - 0.83","0.29 - 0.71","0.23 - 0.67","0.13 - 0.58","0.11 - 0.35","0.09 - 0.27","0.03 - 0.32","0.04 - 0.25","0.02 - 0.20","0.01 - 0.06","-0.03 - 0.11"],["*","*","*","*","*","*","*","*","*","*","*","-"]],"container":"<table class=\"display\">\n <thead>\n <tr>\n <th> <\/th>\n <th>Country<\/th>\n <th>Average lost years<\/th>\n <th>95% CI<\/th>\n <th>p-val<\/th>\n <\/tr>\n <\/thead>\n<\/table>","options":{"pageLength":6,"columnDefs":[{"orderable":false,"targets":0},{"name":" ","targets":0},{"name":"Country","targets":1},{"name":"Average.lost.years","targets":2},{"name":"X95..CI","targets":3},{"name":"p.val","targets":4}],"order":[],"autoWidth":false,"orderClasses":false,"lengthMenu":[6,10,25,50,100],"rowCallback":"function(row, data, displayNum, displayIndex, dataIndex) {\nvar value=data[1]; $(this.api().cell(row, 1).node()).css({'background':'lightblue','angle':'-90','background-size':'98% 88%','background-repeat':'no-repeat','background-position':'center'});\nvar value=data[2]; $(this.api().cell(row, 2).node()).css({'background':'lightblue','angle':'-90','background-size':'98% 88%','background-repeat':'no-repeat','background-position':'center'});\nvar value=data[3]; $(this.api().cell(row, 3).node()).css({'background':'lightblue','angle':'-90','background-size':'98% 88%','background-repeat':'no-repeat','background-position':'center'});\nvar value=data[4]; $(this.api().cell(row, 4).node()).css({'background':'lightblue','angle':'-90','background-size':'98% 88%','background-repeat':'no-repeat','background-position':'center'});\n}"}},"evals":["options.rowCallback"],"jsHooks":[]}</script> --- background-image: url("images/cool.png") # Discussion - Persons with chronic conditions, on average, have a 16% higher risk of early retirement and lose about 3.7 working months compared to their healthy counterparts. -- - These findings highlight the need for prevention and rehabilitation interventions as well as policies promoting the integration or reintegration of people with CDs in the workplace. -- - Further research should focus on studying the impact of specific conditions and the mechanism of deciding to stay/leave the labour market after having CDs for better-targeted policy implementation. --- background-image: url("images/cool.png") # References .footnote[ ¹ European Commission. Directorate General for Economic and Financial Affairs. (2020). The 2021 ageing report: Underlying assumptions and projection methodologies. Publications Office. https://data.europa.eu/doi/10.2765/733565 ² Hess, M., Naegele, L., Becker, L., Mäcken, J., & De Tavernier, W. (2021). Planned Retirement Timing in Europe: Are Europeans Adapting to the Policy of Extending Working Lives. Frontiers in Sociology, 6, 691066. https://doi.org/10.3389/fsoc.2021.691066 ³ OECD. (2023). Pensions at a Glance 2023: OECD and G20 Indicators. OECD. https://doi.org/10.1787/678055dd-en ⁴ De Oliveira, C., Saka, M., Bone, L., & Jacobs, R. (2023). The Role of Mental Health on Workplace Productivity: A Critical Review of the Literature. Applied Health Economics and Health Policy, 21(2), 167–193. https://doi.org/10.1007/s40258-022-00761-w ⁵ Fouad, A. M., Waheed, A., Gamal, A., Amer, S. A., Abdellah, R. F., & Shebl, F. M. (2017). Effect of Chronic Diseases on Work Productivity: A Propensity Score Analysis. Journal of Occupational and Environmental Medicine, 59(5), 480–485. https://doi.org/10.1097/JOM.0000000000000981 ] --- # References .footnote[ ⁶ Kessler, R. C., Greenberg, P. E., Mickelson, K. D., Meneades, L. M., & Wang, P. S. (2001). The effects of chronic medical conditions on work loss and work cutback. Journal of Occupational and Environmental Medicine, 43(3), 218–225. https://doi.org/10.1097/00043764-200103000-00009 ⁷ Roskes, K., Donders, N. C. G. M., & van der Gulden, J. W. J. (2005). Health-related and work-related aspects associated with sick leave: A comparison of chronically ill and non-chronically ill workers. International Archives of Occupational and Environmental Health, 78(4), 270–278. https://doi.org/10.1007/s00420-004-0596-0 ⁸ Vos, T., Lim, S. S., Abbafati, C., Abbas, K. M., Abbasi, M., Abbasifard, M., Abbasi-Kangevari, M., Abbastabar, H., Abd-Allah, F., Abdelalim, A., Abdollahi, M., Abdollahpour, I., Abolhassani, H., Aboyans, V., Abrams, E. M., Abreu, L. G., Abrigo, M. R. M., Abu-Raddad, L. J., Abushouk, A. I., … Murray, C. J. L. (2020). Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: A systematic analysis for the Global Burden of Disease Study 2019. The Lancet, 396(10258), 1204–1222. https://doi.org/10.1016/S0140-6736(20)30925-9 ⁹ Callaway, B., & Sant’Anna, P. H. C. (2021). Difference-in-Differences with multiple time periods. Journal of Econometrics, 225(2), 200–230. https://doi.org/10.1016/j.jeconom.2020.12.001 ] --- class: center, middle **Thank you!** PhD Student: Nga Chau My Ha [chaumynga.ha@paraplegie.ch](chaumynga.ha@paraplegie.ch) [chau.ha@stud.unilu.ch](chau.ha@stud.unilu.ch) <div style="margin-top: 20px;"> <img src="https://www.paraplegie.ch/sites/default/files/2018-05/schweizer_praplegiker-forschung_logo_rgb_en.jpg" width="80px" /><img src="https://upload.wikimedia.org/wikipedia/de/a/a1/Logo_Universit%C3%A4t_Luzern.svg" width="80px" /> --- background-image: url("images/cool.png") # Appendix - Descriptive statistics | Variables | Total sample <br> (n = 11,903) | With CDs <br> (n = 2,224) | Without CDs <br> (n = 9,679) | |----------------------------------|-------------------------------:|--------------------------:|-----------------------------:| | **Men**, n (%) | 6,273 (52.7) | 1,180 (53.1) | 5,093 (52.6) | | **Age**, years, mean (SD) | 58.7 ± 2.6 | 59.6 ± 1.8 | 58.5 ± 2.7 | | **Number of children**, mean (SD)| 2.2 ± 1.3 | 2.2 ± 1.3 | 2.1 ± 1.3 | | **In partnership**, n (%) | | | | | - Yes | 9,389 (79.2) | 1,730 (77.9) | 7,659 (79.5) | | **Education years**, mean (SD) | 12.6 ± 4.1 | 12.6 ± 4.2 | 13 ± 4.1 | | **Income percentile**, mean (SD) | 6.5 ± 2.9 | 6.5 ± 2.9 | 6.5 ± 2.9 | | **Early retired**, n (%) | | | | | - Yes | 1,482 (12.5) | 604 (27.2) | 878 (9.1) | | - No | 10,421 (87.5) | 1,620 (72.8) | 8,801 (90.9) | --- background-image: url("images/cool.png") # Appendix - Outcome 1 <img src="https://upload.wikimedia.org/wikipedia/commons/d/de/Proj01_pic02.png" width="100%" style="display: block; margin: auto;" /> --- background-image: url("images/cool.png") # Appendix - Outcome 2 Country | Average years lost to ER | 95% CI | p-val ------------|:------------------------:|:-----------:|:-------: Slovenia | 0.99 = 11.9 months | 0.61 - 1.38 | * Belgium | 0.60 = 7.2 months | 0.36 - 0.83 | * Germany | 0.50 = 6.0 months | 0.29 - 0.71 | * Italy | 0.45 = 5.4 months | 0.23 - 0.67 | * Austria | 0.35 = 4.2 months | 0.13 - 0.58 | * Switzerland | 0.23 = 2.8 months | 0.11 - 0.35 | * Denmark | 0.17 = 2.0 months | 0.09 - 0.27 | * Spain | 0.17 = 2.0 months | 0.03 - 0.32 | * Sweden | 0.14 = 1.7 months | 0.04 - 0.25 | * Czech | 0.10 = 1.2 months | 0.02 - 0.20 | * Estonia | 0.03 = 0.4 months | 0.01 - 0.06 | * France | 0.04 = 0.5 months |-0.03 - 0.11 | -