Grandparents in the blues. The effect of childcare on grandparents’ depression

Abstract

We estimate the effect of grandchild care on the depression of grandmothers and grandfathers, using data from the Survey on Health, Ageing and Retirement in Europe and an identification strategy which exploits both the random variation in the timing of interviews across individuals and the fact that the demand for childcare declines with the age of grandchildren. We find that more childcare increases depression. The estimated effect is sizeable: ten additional hours of childcare per month increase the probability that complying grandmothers and grandfathers develop depressive symptoms by 3.2 to 3.3 percentage points and by 5.4 to 6.1 percentage points respectively.

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

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5

Notes

  1. 1.

    Pew Research Center “Family Support in Graying Societies How Americans, Germans and Italians Are Coping with an Aging Population”, May 21st 2015.

  2. 2.

    The effects of childcare provisions on female labor supply have been extensively studied, with controversial results. See for instance Posadas and Vidal-Fernandez (2012), Del Boca (2002), Rupert and Zanella (2016), Zamarro (2011), Blau and Currie (2006), and Havnes and Mogstad (2011), are recent reviews of this literature.

  3. 3.

    In these data, both regular and occasional childcare are done without the presence of parents, and we cannot distinguish between primary and supplementary care.

  4. 4.

    In Europe, there is a sharp difference in the provision of formal childcare between children younger than three and older children. See Janta (2014).

  5. 5.

    The distribution of the number of hours of childcare in our data is skewed. About 40 percent of grandparents do not provide any childcare and among those who do provide care the median number of hours is 28 for females and 20 for males.

  6. 6.

    The costs and benefits of regular childcare include lower participation to social activities (Arpino and Bordone 2017) and better cognitive functioning (Ahn and Choi 2018).

  7. 7.

    According to Sobocki et al. (2006), the total annual cost of depression in Europe is estimated at Euro 118 billion in 2004, which corresponds to a cost of 253 Euro per inhabitant. Direct costs alone comprise outpatient care (Euro 22 billion), drug costs (Euro 9 billion) and hospitalization (Euro 10 billion). Indirect costs due to morbidity and mortality are estimated at Euro 76 billion. The cost of depression corresponds to 1% of total European GDP.

  8. 8.

    This measure was proposed by the British government in 2015 and implemented in California since 2014 with the Paid Family Leave Act.

  9. 9.

    In SHARE, the information on children is reported by only one household member, the so called respondent. We impute this information to the non-respondent partner in all households with two eligible members (59 percent of the sample). About 85 percent of individuals aged 50 to 75 have one to four children. At least one child is required to qualify as potential grandparent.

  10. 10.

    Close to 94 percent of the youngest grandchildren in our sample are aged 0 to 16. We expect that childcare is not an issue for grandchildren older than 16.

  11. 11.

    Both the number of grandchildren and the number of grandchildren a grandparent cares for are endogenous variables that we exclude from the list of controls in the empirical analysis.

  12. 12.

    We have excluded from the baseline sample Israel as a non-European country, Ireland because of the very few observations and Greece because the short fieldwork period in the first wave (due to the beginning of the 2004 Olympic Games) and the use of the telephone directory as the sampling frame cast doubts on the representativeness of the Greek sample. See Mazzonna and Peracchi (2014).

  13. 13.

    Most of these symptoms refer to the last month or to the current situation.

  14. 14.

    Castro-Costa et al. (2007) and Larraga et al. (2006) use the cut-off point of 3/4 to predict both depression and short-care pervasive depression.

  15. 15.

    Our qualitative results are unaffected if we set the threshold value at three or five rather than at four. Results available from the authors upon request.

  16. 16.

    These two factors are associated with the two eigenvalues higher than one and explain about 33 percent of total variance.

  17. 17.

    The age of the oldest child is a proxy for children’s health, which correlates with grandparents’ health. Grandparents’ birth order and height are pre-determined variables that correlate with socio-economic status later in life. See Black et al. 2005, and Case and Paxton 2008.

  18. 18.

    For each variable, we set missing values to zero and add to each regression a dummy equal to 1 if the value of the relevant variable is missing.

  19. 19.

    We cluster standard errors at the household level in the full sample (males and females), and at the individual level in the samples of males and females.

  20. 20.

    These estimates are obtained from regressions where the age of the youngest grandchild enters as a set of dummies (age 16 being the reference age) and the controls in vectors X and W are included. Omitting these controls does not qualitatively change the reported patterns.

  21. 21.

    The basic set of controls includes country by wave dummies, grandparents’ gender and age measured at the beginning of the fieldwork and the age of the youngest grandchild at the beginning of the fieldwork. The full set of controls corresponds to the sets X and W in Eq. (1).

  22. 22.

    This interpretation holds even though both Z and HC are not dummy variables.

  23. 23.

    The large first-stage effect is not an artifact due to outliers or to the skewed distribution of HC. We transformed HC in several ways to reduce skewness and we estimated Tobit models with alternative upper bounds. In all cases, the size of parameter associated to Z remained similar to those reported in Table 2.

  24. 24.

    We have also estimated (1) using as instrument a dummy equal to one if Z if equal to or higher than 6 months and to zero otherwise, with results very similar to those shown in Table 3.

  25. 25.

    For both the linear and the Probit specification we report the marginal effect of increasing childcare by one hour per month on the probability of reporting depression. Hence, the effect of an increase of ten hours per month is obtained by multiplying the estimated coefficient by 10.

  26. 26.

    To ease interpretation, we have transformed both HC and Z in the dummy variables DHC, which takes value 1 if HC > 0 and 0 otherwise, and DZ, which takes value 1 if Z is at or above its median value and 0 otherwise. We always control for X and W.

  27. 27.

    The analysis by gender yields imprecise results because of the reduction in sample size induced by the focus on single characteristics. Even in the full sample, due to the small sample size, the estimated ratio is significantly different from one only for inactivity and for residence in Italy, Spain and Poland.

  28. 28.

    Educational attainment and activity rates are lower in Italy, Spain and Poland than in the rest of the sample. The proportion of grandparents residing within 5 km from one adult child exceeds 80 percent in Italy, Spain and Poland and ranges from 39 percent in France to 69 percent in Austria and Czech Republic.

  29. 29.

    IV estimates can be interpreted as local average treatment effects if monotonicity holds. In our context, monotonicity requires that all grandparents interviewed n months after the beginning of the survey fieldwork, for whom Z = n, either hold constant or reduce their childcare compared to the level they would have provided were they assigned Z = n-1.

  30. 30.

    In the case of females, the dummies for one educational attainment and for whether parents are alive as well as the relative performance in math at age 10 turn out to be imperfectly balanced. In the case of males, unbalancing occurs for one of the dummies indicating the number of books in the house at age 10 and for the dummy indicating the presence of a co-residing partner.

  31. 31.

    In these tests, we condition on the pre-determined variables in vectors X and W.

  32. 32.

    Admittedly, the small size of samples S2 and S3 reduces the power of the tests. To evaluate whether the effect of Z varies by cohort, we have also interacted Z with the age of grandparents at the beginning of the fieldwork. Our results—available upon request—are qualitatively unchanged.

  33. 33.

    As shown by Rosembaum and Rubin (1983), conditional on the same propensity score, the joint distribution of the covariates is the same among seniors with grandchildren and seniors without. We estimate the propensity score of having grandchildren separately by gender of the senior and match each observation in the baseline sample with the observation in sample S1 which has the nearest value of the propensity score. This refinement improves comparability at the cost of reducing sample size, and is therefore not applicable to samples S2 and S3 because of their relatively small sample size.

  34. 34.

    We have also matched each observation in sample S1 with the closest one in the baseline sample in terms of the propensity score of not having grandchildren. Results are qualitatively similar.

  35. 35.

    In each falsification test, HC = 0 and Eq. (4) corresponds to the reduced form of Eq.(1).

  36. 36.

    The effect of hours of childcare on depression remains positive and statistically significant. Due to the small size of samples S2 and S3, the standard errors estimated in the falsification tests are large. Even in these case, however, the adjusted confidence intervals exclude zero in most cases.

  37. 37.

    Recall that A, the age of the youngest grandchild, is included among the regressions in Eq. (1).

  38. 38.

    We also investigate whether the estimated effects of childcare on depression vary with grandparents being maternal or paternal, but find no evidence that this is the case.

  39. 39.

    We exclude from the sample the grandparents providing childcare less than on a weekly basis. In this exercise, we look at the extensive rather than at the intensive margin. This strategy is motivated by the need of having an adequate sample size, as only about one third of the full sample provides childcare at least on a weekly basis.

  40. 40.

    Poor health is derived from a five-levels self-reported health scale. We define a dummy equal to 1 when self-reported health is poor or fair (and equal to zero when it is good, very good or excellent). Numeracy and word recall are the outcomes of cognitive tests performed by SHARE respondents, who are asked to solve simple calculations and recall up to ten words. The associated dummies are equal to 1 if individual performance exceeds median performance and to 0 otherwise. Life satisfaction is measured as a dummy equal to 1 for the top three levels of a 10-levels scale and to 0 otherwise. Control, autonomy, pleasure and self-realization are based on the corresponding four sub-scales composing the twelve-items CASP index of quality of life and wellbeing. Each trait is measured by a dummy variable equal to 1 when the underlying sub-scale is larger than or equal to 9 (the maximum being 12) and to 0 otherwise.

  41. 41.

    Further details are available from the authors upon request.

  42. 42.

    Mania is the opposite of a depressive mood, when a person feels plenty of energy and that he/she can do almost anything.

References

  1. Ahn, T., & Choi, K. (2018). Grandparent caregiving and cognitive functioning among older people: evidence from Korea. Review of the Economics of the Household, 16, 1–34.

    Article  Google Scholar 

  2. Angrist, J. & Pischke, J. (2008). Almost harmless econometrics. Princeton: Princeton University Press.

  3. Arpino, B., & Bordone, V. (2014). Does grand-parenting pay off? The effect of child care on grandparents’ cognitive functioning. Journal of Marriage and Family, 76(April 2014), 337–351.

    Article  Google Scholar 

  4. Arpino, B., & Bordone, V. (2017). Regular provision of granchild care and participation in social activities. Review of the Economics of the Household, 15, 135–174.

    Article  Google Scholar 

  5. Baker, L. A., & Silverstein, M. (2008). Preventive health behaviors among grandmothers raising grandchildren. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 63, S304–S311.

    Article  Google Scholar 

  6. Benazon, N. R., & Coyne, J. C. (2000). Living with a depressed spouse. Journal of Family Psychology, 14(1), 71–79.

    Article  Google Scholar 

  7. Black, D. A., Joo, J., LaLonde, R., Smith, J., & Taylor, E. (2015). Simple tests for selection bias: learning more from instrumental variables, IZA discussion papers 9346. Bonn, Germany: Institute for the Study of Labor (IZA).

  8. Black, S., Devereux, P., & Salvanes, K. (2005). The more the merrier? The effect of family size and birth order on children’s education. The Quarterly Journal of Economics, 120(2), 669–700.

    Google Scholar 

  9. Blau, D. M., & Currie, J. (2006). Pre-school, day care, and after-school care: who’s minding the kids? In E. A. Hanushek, & F. Welch (Eds.), Handbook of the economics of education, vol. 2 (pp. 1163–1278). New York: Elsevier.

  10. Börsch-Supan, A., & Krieger, U. (2013). Investigating response behavior. In F. Malter & A. Börsch-Supan (Eds.), SHARE wave 4 innovations & methodology. Munich: Munich Center for the Economics of Aging (MEA).

    Google Scholar 

  11. Case, A., Fertig, A., & Paxton, C. (2005). The lasting impact of childhood health and circumstance. Journal of Health Economics, 24(2), 365–389.

    Article  Google Scholar 

  12. Case, A., & Paxton, C. (2008). Height, health, and cognitive function at older ages. American Economic Review: Papers & Proceedings, 98(2), 463–467.

    Article  Google Scholar 

  13. Castro-Costa, E., Dewey, M., Stewart, R., Banerjee, S., Huppert, F., Mendonca-Lima, C., Bula, C., Reisches, F., Wancata, J., Ritchie, K., Tsolaki, M., Mateos, R., & Prince, M. (2007). Prevalence of depressive symptoms and syndromes in later life in ten European countries: The SHARE study. British Journal of Psychiatry, 191(5), 393–401.

    Article  Google Scholar 

  14. Castro-Costa, E., Dewey, M., Stewart, R., Banerjee, S., Huppert, F., Mendonca-Lima, C., Bula, C., Reisches, F., Wancata, J., Ritchie, K., Tsolaki, M., Mateos, R., & Prince, M. (2008). Ascertaining late-life depressive symptoms in Europe: an evaluation of the survey version of the Euro-D scale in 10 nations. The SHARE Project. International Journal of Methods in Psychiatric Research, 17(1), 12–29.

    Article  Google Scholar 

  15. Clement, S., Schauman, O., Graham, T., Maggioni, F., Evans-Lacko, S., Bezborodovs, N., Morgan, C., Rüsch, N., Brown, J. S. L., & Thornicroft, G. (2015). What is the impact of mental health-related stigma on help-seeking? A systematic review of quantitative and qualitative studies. Psychological Medicine, 45, 11–27.

    Article  Google Scholar 

  16. Coall, D. A., & Hertwig, R. (2011). Grandparental investment: a relic of the past or a resource for the future? Psychological Science, 20, 93–98.

    Google Scholar 

  17. Coyne, J. C., Kessler, R. C., Tal, M., Turnbull, J., Wortman, C. B., & Greden, J. F. (1987). Living with a depressed person. Journal of Consulting and Clinical Psychology, 55(3), 347–352.

    Article  Google Scholar 

  18. Conley, T. G., Hansen, C. B., & Rossi, P. E. (2012). Plausibly exogenous. The Review of Economics and Statistics, 94(1), 260–272.

    Article  Google Scholar 

  19. Del Boca, D. (2002). The effect of child care and part time opportunities on participation and fertility decisions in Italy. Journal of Population Economics, 15, 549–573.

    Article  Google Scholar 

  20. De Luca, G., & Lipps, O. (2005). Fieldwork and survey management in SHARE. In A. Börsch-Supan & H. Jürges (Eds.), The Survey of Health, Aging, and Retirement in Europe – Methodology. Mannheim: Mannheim Research Institute for the Economics of Aging.

  21. Dewey, M., & Prince, M. (2005). Mental health. In A. Borsch-Supan (Ed.), Health, ageing and retirement in Europe. First results from the survey of health, ageing and retirement in Europe (pp. 108–117). Mannheim: Mannheim Research Institute for the Economics of Ageing.

    Google Scholar 

  22. Di Gessa, G., Glaser, K., & Tinker, A. (2016). The health impact of intensive and non-intensive grandchild care in Europe: new evidence from SHARE. Journal of Gerontology: Social Sciences, 71(5), 867–879.

    Article  Google Scholar 

  23. Eisenberg, D., Golberstein, E., Whitlock, J. L., & Downs, M. F. (2013). Social contagion of mental health: evidence from college roommates. Health Economics, 22, 965–986.

    Article  Google Scholar 

  24. European Commission (2012). The role of men in gender equality: European strategies and insights, Brussels: European Commission.

  25. Ferrari, A. J., Charlson, F. J., Norman, R. E., Patten, S. B., Freedman, G., Murray, C. J. L., Vos, T., & Whiteford, H. A. (2013). Burden of depressive disorders by country, sex, age, and year: findings from the global burden of disease study 2010. PLoS Medicine, 10(11), e1001547.

    Article  Google Scholar 

  26. Glaser, K., Price, D., Di Gessa, G., Ribe, E., Stuchbury, R., & Tinker, A. (2013). Grandparenting in Europe: family policy and grandparents’ role in providing childcare. Grandparents plus report. https://www.grandparentsplus.org.uk/grandparenting-in-europe-family-policy-and-grandparents-role-inproviding-childcare.

  27. Goldman, L. S., Nancy, H., Nielsen, N. H., & Champion, H. C. (1999). Awareness, diagnosis, and treatment of depression. Journal of General Internal Medicine, 14, 569–580.

    Article  Google Scholar 

  28. Havnes, T., & Mogstad, M. (2011). Money for nothing? Universal child care and maternal employment. Journal of Public Economics, 95, 1455–1465.

    Article  Google Scholar 

  29. Hughes, M. E., Waite, L. J., LaPierre, T. A., & Luo, Y. (2007). All in the family: the impact of caring for grandchildren on grandparents’ health. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 62, S108–S119.

    Article  Google Scholar 

  30. Hwang, S., Jayadevappa, R., Zee, J., Zivin, K., Bogner, H. R., Raue, P. J., Bruce, M. L., Reynolds, III, C. F., & Gallo, J. J. (2015). Concordance between clinical diagnosis and medicare claims of depression among older primary care patients. The American Journal of Geriatric Psychiatry, 23(7), 726–734.

    Article  Google Scholar 

  31. Janta, B. (2014). Caring for children in Europe. Cambridge: RAND Europe.

  32. Kessler, R. C., & Bromet, E. J. (2013). The epidemiology of depression across cultures. Annual Review of Public Health, 34, 119–138.

    Article  Google Scholar 

  33. Ku, L., Stearns, S., Houtven, C., & Holmes, G. (2012). The health effects of caregiving by grandparents in Taiwan: an instrumental variable estimation. Review of Economics of the Household, 10(4), 521–540.

    Article  Google Scholar 

  34. Larraga, L., Saz, P., Dewey, M., Marcos, G., & Lobo, A. (2006). Validation of the Spanish version of the EURO-D scale: an instrument for detecting depression in older people. International Journal of Geriatric Psychiatry, 21, 1199–1205. 2006.

    Article  Google Scholar 

  35. Maurer, D. M. (2012). Screening for depression. American Family Physician, 85(2), 139–144.

    Google Scholar 

  36. Mausner-Dorsch, H., & Eaton, W. W. (2000). Psychosocial work environment and depression: epidemiologic assessment of the demand–control model. American Journal of Public Health, 90(11), 1765–1770.

    Article  Google Scholar 

  37. Mazzonna, F., & Peracchi, F. (2014). Unhealthy retirement?. EIEF working papers series 1409. Roma: Einaudi Institute for Economics and Finance (EIEF).

  38. Minkler, M. (1999). Intergenerational households headed by grandparents: contexts, realities, and implications for policy. Journal of Aging Studies, 13, 199–218.

    Article  Google Scholar 

  39. NIMH National Institute of Mental Health. (2015). Depression. What you need to know. Bethesda: NIMH.

    Google Scholar 

  40. Posadas, J., & Vidal-Fernandez, M. (2012). Grandparents’ childcare and female labor force participation. IZA discussion paper no. 6398. Bonn: IZA.

  41. Powdthavee, N. (2011). Life satisfaction and grandparenthood: evidence from a nationwide survey. IZA discussion paper no. 5869. Bonn: Institute for the Study of Labor (IZA).

  42. Prince, M. J., Beekman, A. T., Deeg, D. J., Fuhrer, R., Kivela, S. L., & Lawlor, B. A., et al. (1999). Depression symptoms in late life assessed using the EURO-D scale: effect of age, gender and marital status in 14 European centres. British Journal of Psychiatry, 174, 339–345.

    Article  Google Scholar 

  43. Rosenbaum, P. R., & Rubin, D. (1983). The central role of the propensity score in observational studies for causal effects. Biometrika, 70(1), 41–55.

    Article  Google Scholar 

  44. Rupert, P., & Zanella, G. (2016). Grandchildren and their grandparents’ labour supply. New York: Mimeo.

  45. Reinkowski, J. (2013). Should we care that they care? Grandchild care and its impacts on grandparent health. ifo working paper 165. Munich: ifo Institute.

  46. Silverstein, M., Giarrusso, R., & Bengtson, V. L. (2003). Grandparents and grandchildren in family systems: a social-developmental perspective. In V. L. Bengtson & A. Lowenstein (Eds.), Global aging and challenges to families (pp. 75–103). New York: Aldine de Gruyter.

    Google Scholar 

  47. Sobocki, P., Jönsson, B., Angst, J., & Rehnberg, C. (2006). Costs of depression in Europe. Journal of Mental Health Policy Economics, 9(2), 87–98.

    Google Scholar 

  48. Zamarro, G. (2011). Family labor participation and child care decisions. The role of grannies. RAND working paper WR-833. Santa Monica: Rand.

Download references

Acknowledgements

We are grateful to Marco Bertoni, Hope Corman, Maria De Paola, Jeffrey DeSimone, José Escarce, Andrea Ichino, Ariela Lowenstein, Giovanni Mastrobuoni, Raffaele Miniaci, Michele Pellizzari, Enrico Rettore, Lucia Rizzica, Silvia Salcuni, Anna Sanz de Galdeano, Marcello Sartarelli, Elena Stancanelli and the audiences at seminars and conferences in Alicante, Brescia, Bolzano, Ispra (JRC), Trento (IRVAPP), Jerusalem (Taub Center), San Diego and UCLA for comments and suggestions. This paper uses data from SHARELIFE release 1 and SHARE release 2.6.0. The SHARE data collection has been primarily funded by the European Commission through the 5th frame work programme (project QLK6-CT-2001-00360 in the thematic programme Quality of Life), through the 6th framework programme (projects SHARE-I3, RII-CT-2006 062193, COMPARE, CIT5-CT-2005-028857, and SHARELIFE, CIT4-CT-2006-028812) and through the 7th frame work programme (SHARE-PREP, 211909 and SHARE-LEAP, 227822). Financial support by Fondazione Cariparo Starting Grant “Education, Retirement and Household Behavior” is gratefully acknowledged. The usual disclaimer applies.

Funding

This study was funded by the CARIPARO Foundation Starting Grant on “Education, Retirement and Household Behavior”.

Author information

Affiliations

Authors

Corresponding authors

Correspondence to Giorgio Brunello or Lorenzo Rocco.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Appendix

Measuring depression

Depression is the second leading cause of disability worldwide and a major contributor to the burden of suicide and ischemic heart disease (Ferrari et al. 2013). It is a syndrome characterized by the presence of specific symptoms. However, not all depressed people display the same set of symptoms and their severity, frequency and duration vary from individual to individual. Hence, “…no two people are affected the same way by depression…” (NIMH 2015). The symptoms typically associated with depression are:

  1. 1.

    persistent sad, anxious, or empty mood

  2. 2.

    feelings of hopelessness, pessimism

  3. 3.

    feelings of guilt, worthlessness, helplessness

  4. 4.

    loss of interest or pleasure in hobbies and activities

  5. 5.

    decreased energy, fatigue, being slowed

  6. 6.

    difficulty concentrating, remembering or making decisions

  7. 7.

    insomnia, early-morning awakening or oversleeping

  8. 8.

    appetite and/or weight loss, or overeating and weight gain

  9. 9.

    thoughts of death or suicide; suicide attempts

  10. 10.

    restless and irritability

  11. 11.

    persistent physical symptoms that do not respond to treatment, such as headaches, digestive disorders and chronic pain

There are many types of depressive disorders. The most common are the Major Depression Disorder (MDD) and Dysthymia. MDD is a severe condition of depressed mood that hampers all dimensions of a person’s life, including work, home and social life. In MDD symptoms last for at least two weeks. Although some people suffer for just one episode of MDD, in most cases the problem is recurrent. The probability of experiencing another episode after the first has been treated and concluded is high and exceeds 75 percent. Even after a successful treatment some residual symptoms remain in 10 to 30 percent of the cases (Maurer 2012). Dysthymia is less severe than MDD but the symptoms are continuously present over a much longer period of time, for at least 2 years. This is considered a chronic form of depression. Other types of depression include Adjustment Disorder with Depressed Mood, that typically follows deep changes in a person life, the Seasonal Affective Disorder generally emerging in winter when the length of daylight shortens, the Postpartum Disorder, related to the abrupt fall in the quantity of hormones soon after the childbirth, and the Bipolar Disorder, characterized by sudden swings from depressed mood to maniasFootnote

Mania is the opposite of a depressive mood, when a person feels plenty of energy and that he/she can do almost anything.

(NIMH 2015).

Depression is twice as common among women as men. Generally, the first episode of depression occurs quite early, the median age being in the mid-twenties (Kessler and Bromet 2013). Many factors may play a role in depression, including genetics, brain biology and chemistry, and life events such as trauma, loss of a loved one, a difficult relationship, an early childhood experience, or any stressful situation (NIMH 2015). Limited evidence suggests also the possibility that depression may be “contagious” and that partners living with a depressed person could develop symptoms of depression. Two early studies analyzing small samples of partners of depressed persons found higher prevalence of symptoms compared to the general population (Coyne et al. 1987; Benison and Coyne 2000). However, selection problems and contextual effects are likely to be responsible for the result, as pointed out by Eisenberg et al. (2013), who exploit the exogenous assignment of roommates in a college dormitory and do not find evidence of contagion.

There are no lab tests that can diagnose depression, although lab tests can be used to exclude the presence of other medical conditions responsible for some of the typical symptoms of depression. Rather, the diagnosis is performed by means of a careful clinical interview conducted by a psychiatrist. A patient is diagnosed with depression if he or she meets the criteria stated by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM- IV). For instance, for MDD these criteria include most of the symptoms used to compile the EURO-D index that we use in this paper. They also verify the presence of conditions that can mimic depression (such as substance abuse, medical illness and other psychiatric disorders). Typically, before visiting a psychiatrist, patients refer to their family doctors, who use screening tools such as the General Health Questionnaire, the Beck Depression Inventory, the Symptom Checklist, the Inventory of Depressive Symptoms, and the Zing Depression Scale. Scores above a predetermined cut-off call for a deeper evaluation by a psychiatrist. These screens have sensitivities of 70 percent to 85 percent and specificities of about 80 percent. The Center for Epidemiologic Studies Depression scale and the shortened Geriatric Depression Scale, the base for EURO-D, are particularly valuable tools to detect depression among the elderly (Goldman et al. 1999).

In Europe and the US, 52 to 74 percent of people with mental disorders do not receive treatment (Clement et al 2015) and, even after initiation of mental health care, non-adherence to treatment programs and early withdrawal from services are common. Many causes explain the lack of treatment and the poor adherence. On the patients’ side, disclosure concerns, perception of social stigma against mental illness, fear of the psychotherapy and reluctance to take antidepressants play an important role. Goldman et al. (1999), state: “…Many patients are reluctant to acknowledge to themselves or their physicians that they are experiencing emotional distress. Patients may deny or minimize symptoms, rationalize them as expectable because of life stresses or as due to other general medical problems, believe them to be failures of will or moral shortcomings, or not see them as within the physician’s purview or capabilities. […]. Similarly, patients may be reluctant to disclose information they fear could be included in insurance or employment records… (p. 574).” Physicians too might be responsible for the lack of treatment as they often fail to recognize depression, especially because of the absence of confirmatory laboratory or radiologic tests.

Particularly, in older adults depression may remain undetected because tiredness, sleeping problems, irritability and grumpiness might be considered a normal event in old age. Older adults may also have additional medical conditions such as heart disease, stroke, or cancer, which may cause depressive symptoms, or they may be taking medications with side effects that contribute to depression. Hwang et al. (2015), find that, in a sample of older primary care patients screened for depression according to the DSM-IV criteria, only 42 percent of those turning out to be depressed had Medicare claims for depression (i.e. were treated for depression).

The SHARE fieldwork

Fieldwork was expected to start in May 2004 for wave 1 and in October 2006 for wave 2. In practice, mainly because of funding delays, interviewers did not start working simultaneously in all countries. Depending on when each country started its operations, the intensity of data collection changed. Some countries started their operations earlier than expected. Specifically, the first interviews of wave 1 were carried out in Germany already in February 2004 (3 months earlier than expected), followed by a few other interviews in Denmark and Sweden in March and 90 interviews distributed across several countries in April. Because of this, we select February 2004 as the starting point of wave 1 operations. Similarly, for wave 2, the first interviews were in September 2006 (in Germany, 1 month in advance of the expected start). Therefore, we set September 2006 as the starting point of the wave 2 campaign.

Interviewers were instructed to attempt contacting sample units at least 5 times in wave 1 and 8 times in later waves. Contacts could be by telephone or in person, although a certain number of contacts in person were compulsory, and had to occur at different times of the day and in different days of the week. In general, not all sample units in SHARE were immediately contacted. The reason is that a relatively small number of interviewers were active at the beginning of the fieldwork. In all countries but Sweden, interviewers, in planning their work, progressively contacted new households as long as they completed previous interviews. Therefore, the date of first contact is approximately uniformly distributed, as documented by De Luca and Lipps (2005).

On average, each interviewer was assigned 43 households. Interviewers had the task of deciding how to organize their work over the established fieldwork period. According to Börsch-Supan and Krieger (2013), “…driving distances between households is … [interviewers’] … foremost consideration when planning their work. While interviewers get some compensation for travel cost, their main income results from finalized interviews. Thus, they try to optimize driving distances between addresses… (p.53)”.

To increase households’ propensity to collaborate, SHARE emailed or mailed an advance letter to each household informing about upcoming calls or visits by an interviewer, communicating the nature and motivation of the survey and explaining the importance of participating. After the interview, a thank-you letter was mailed out to each respondent in order to increase the propensity to participate in future waves of the survey. In most SHARE countries, incentives to respondents were distributed in order to gain their cooperation. In several countries, individuals received a small gift before completing the interview (e.g. a lottery ticket in Sweden, a box with a set of ball-pens in Germany, a sweet in Austria, a voucher for department stores in Spain). In other countries, incentives were given at the end of the interview (15 Euro per completed household in the Netherlands). Denmark was the only country where incentives were considered inappropriate.

Electronic supplementary material

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Brunello, G., Rocco, L. Grandparents in the blues. The effect of childcare on grandparents’ depression. Rev Econ Household 17, 587–613 (2019). https://doi.org/10.1007/s11150-018-9432-2

Download citation

JEL Codes

  • J13
  • I12

Keywords

  • Childcare
  • Grandparents
  • Depression
  • Europe