Are the gender gaps in informal caregiving intensity and burden closing due to the COVID-19 pandemic? Evidence from the Netherlands.

This research note presents the findings of changes in the gender gap in informal care provision and caregiver burden during the Spring 2020 COVID-19 lockdown in the Netherlands. Government measures in response to the pandemic strongly restricted informal caregivers in providing help and care to persons with health-related needs. At the same time, formal care was scaled back and informal caregivers' urge to help their loved ones was likely higher than before the pandemic. Generally, women pick up a larger share of informal care and experience more caregiver burden. We assessed whether the COVID-19 pandemic affected the gender gaps in informal caregiving by analyzing unique data on Dutch informal caregivers (N = 1672 caregiving situations in 1196 respondents). Results showed that women, compared to men, were more likely to have stopped and reduced caregiving, but also to have cared more during the pandemic. Moreover, based on longitudinally comparing levels during and before the lockdown, we found that the gender gap in caregiver burden narrowed down, especially because burden decreased for women. This means that the caregiver burden was more equally distributed among women and men during the Spring 2020 COVID-19 lockdown than before.

INTRODUCTION

Informal care provision, meaning giving help or care to older or disabled persons in the personal network due to health‐related needs, has been shown to negatively affect subjective well‐being and employment outcomes, especially when care is provided for many hours a week (Gomez‐Leon et al., 2019; Kelle, 2020; Verbakel, 2014). In pre‐corona times, we know that women tended to provide informal care more often, and for more hours, than men (Schmid et al., 2012; Verbakel et al., 2017) and experienced a higher caregiver burden than men (Pinquart & Sorensen, 2006; Swinkels et al., 2017). These patterns were also confirmed by our data. This research note will present the findings of how the gender gap in informal care provision and caregiver burden changed during the Spring 2020 COVID‐19 lockdown in the Netherlands. The intriguing part of the COVID‐19 pandemic in this respect is that it may give us a preview of a realistic future situation—caused by aging and increasing chronic illnesses rather than a virus—in which societies face a high demand for care combined with a shortage of formal and informal care. We are interested in the potential consequences of gender inequality that come with it.

Although the Dutch government did not enforce a complete lockdown during the first wave of the COVID‐19 pandemic, it began to take strong measures from mid‐March, which were labeled as an “intelligent lockdown.” Several of the measures the Dutch government took strongly restricted informal caregivers in their core tasks. From March 15, people were strongly advised to keep 1.5‐m distance, to reduce social contacts to a minimum, and to avoid contact completely in case of (mild) signs of illness (Government of the Netherlands, 2020a). Visits to care recipients in nursing homes were not allowed from March 20 onward (Government of the Netherlands, 2020b). While being restricted by the rules, informal caregivers, at the same time, were likely to feel a stronger urge to care because, in times of a pandemic, they are more concerned about the (frail) health of their loved ones and feel an even stronger responsibility to care for them (de Boer et al., 2020). Additionally, formal care was reduced as many home care organizations scaled back the help provided to care recipients and short‐term care and daycare were closed (Ministry of Health, Welfare & Sports, 2020). The Spring 2020 lockdown period ended in the Netherlands on June 1, 2020.

In this research note, we will answer the following research questions: To what extent did the gender gap (a) in the intensity of informal care provision and (b) in caregiver burden decrease or increase during the Spring 2020 lockdown in the Netherlands?

METHODS

Data

We used Dutch caregiver data collected within the Longitudinal Internet Studies for the Social sciences panel administered by CentERdata (Tilburg University, The Netherlands). This online panel is based on a representative sample of the Dutch population (see www.lissdata.nl and Scherpenzeel & Das, 2010 for more information). We collected data twice. We fielded a questionnaire in March 2020 with extensive information on informal care situations and experiences now and in the past. The data were organized in current or past caregiving spells with a maximum of up to seven spells per respondent. Among respondents with at least one current caregiving spell reported in March 2020, we fielded a follow‐up questionnaire in July 2020 on the impact of COVID‐19 on their caregiving situations and experiences during April/May, that is, during the Spring 2020 lockdown. The caregiving spells reported in March were reloaded so that all questions were answered for each spell separately. This follow‐up study formed the basis of the analyses in this research note. We used the longitudinal character for constructing some of our outcome variables. Our sample only included caregiving spells that continued during the Spring 2020 lockdown as well as those that ended because the COVID‐19 crisis prevented the caregiver from providing care (N = 1671). We excluded caregiving spells that ended before April/May because the care recipient died, the care recipient was institutionalized, the care recipient did not need help anymore, the care provision was taken over by someone else than the respondent, or the caregiver could not handle the caregiving anymore. The 1671 caregiving spells were distributed over 1196 caregivers, 677 women and 519 men; most of the caregivers cared for one person to seven persons at maximum. The unit of analysis were the 1671 caregiving spells. For the analyses on burden, we excluded the spells that ended before April/May (due to the COVID‐19 crisis preventing the caregiver from providing care; N = 261), as no comparison in burden could be made in these cases. For the models, we did listwise deletion resulting in different observation numbers for different models.

Measures

We were interested in two types of outcomes. First, we studied changes in the intensity of informal care provision. This was measured by self‐reported changes in caregiving on the question “Did the corona crisis affect the amount of time you helped <> in April/May? Indicate which statement best describes your situation.” There were five possible, ordered outcomes: compared to the period before the corona crisis, the caregiver did (1) care much less, (2) care less, (3) continue to care the same amount, (4) care more, or (5) care much more. The longitudinal design of our data allowed for an extra category (0) that comprised caregivers who stopped providing care to the care recipient because the COVID‐19 crisis prevented the caregiver from providing care.

Second, we studied changes in caregiver burden, and we used two measurements for that. The first measurement reflected a subjective evaluation of the respondent based on the question: “Indicate for each statement to what extent it applied to you in April/May in comparison to the period before the corona crisis: I found providing help to <> hard.” Answer categories were (1) much less in April/May than before the corona crisis, (2) less in April/May than before the corona crisis, (3) as much in April/May than before the corona crisis, (4) more in April/May than before the corona crisis, or (5) much more in April/May than before the corona crisis. In the second measurement, an objective measurement on caregiver burden, we compared responses to the item “I find/found providing help to <> hard” (answer categories ranging from 1 “completely disagree” to 5 “completely agree”) as provided by the respondent in the March data with that to the July data. Note that the assessment of the caregiver referred to the same care recipient in both time points. The comparison resulted in three options: (1) less burden, (2) no change, or (3) more burden in the lockdown period than before.

Our main predictor sex is measured binary (men 0, women 1). Further, we controlled for age and for how many people the person was caring for at the moment of the interview.

Analytical strategy

We first compared the bivariate descriptives of changes in intensity and burden of women and men. In a second step, we ran multinomial logistic regressions including the control variables, with clustered standard errors in STATA 16 to take into account that caregivers can have multiple caregiving relationships. We compared categories that implied changed intensity or burden to the reference category “the same.” If the outcome variable is truly ordinal rather than categorical (which may be the case for the two measurements of caregiver burden), multinomial logistic regression models may yield conservative results because in that scenario the number of parameters is overestimated. Therefore, as a robustness check, we ran ordinal logistic regression models for subjective and objective burden to check whether we missed statistically significant results.

RESULTS

Changes in the gender gap in the intensity of informal care provision

Looking at the descriptive results for the change in the intensity of caregiving due to COVID‐19 in Spring 2020, we see that most of the women and men did not change their intensity (see Figure 1). Stability in caregiving intensity was higher among men (48%) than among women (39%). A non‐negligible group of caregivers increased the intensity of care provision, and this was slightly more common among women than among men: in total, 18% of the women cared (much) more and 14% of the men cared (much) more. However, women were also the ones who stopped the most due to the restrictions the corona crisis put on care provision (17% compared to 13% of the men) and the ones who changed most often to much less caregiving (10% compared to 8% men). This two‐sided pattern was confirmed by the multinomial logistic regression. We found that relative to no change, women were significantly more likely than men to have stopped caregiving (b = 0.56, p = 0.001 in Table A1 in the Appendix) and to have cared much less (b = 0.40, p = 0.04), but also to have cared more during the lockdown than before (b = 0.47, p = 0.006 in Table A1 in the Appendix). We did not find significant differences between women and men relative to no change for less or much more caregiving. In sum, our findings showed a two‐sided picture regarding the gender gap in the intensity of informal care during the COVID‐19 lockdown: Women were the ones who more often stopped and provided much less care, but at the same time also the ones who more often cared more.

FIGURE 1

Descriptive differences between men and women in the change in informal care intensity during the Spring 2020 lockdown in the Netherlands. N = 1671

TABLE A1

Multinomial logistic regression with clustered standard errors

Change in intensity (ref. no change)	Stopped care	Much less care	Less care	More care	Much more care
Women (ref. men)	0.555** (0.175)	0.396* (0.195)	0.139 (0.157)	0.472** (0.171)	0.451 (0.359)
Age	−0.000 (0.007)	0.001 (0.006)	−0.010 (0.005)	−0.020*** (0.005)	0.003 (0.010)
Number of people cared for	0.286** (0.089)	0.082 (0.090)	0.058 (0.089)	0.095 (0.107)	0.309 (0.168)
Constant	−1.889*** (0.460)	−1.998*** (0.410)	−0.594 (0.351)	−0.476 (0.345)	−3.915*** (0.748)
Observations	1671	1671	1671	1671	1671

Note: Coefficients reported as log odds are relative to the reference category (no change). Standard errors in parentheses.

Significance levels: *p < 0.05, **p < 0.01, ***p < 0.001.

Changes in the gender gap in caregiver burden

For the change in caregiver burden, we looked at two indicators. First, we discuss how women and men perceived their burden change due to the corona crisis with a subjective measure. Second, we discuss the results from the comparison between burden indicated in March and burden indicated in April/May.

Subjective change in caregiver burden

Again, and even more strongly than with intensity, we found that most of the caregivers did not perceive a change in their caregiver burden (see Figure 2): 73% of the men and 68% of the women found caregiving as hard as before the lockdown. This also means that again men were more likely to be stable in their caregiving compared to women. Descriptively, women seemed to be more likely than men to feel more subjective burden during the Spring 2020 lockdown than before, with 13% of the women feeling more burden and 6% of the women feeling much more burden compared to, respectively, 11% and 4% of the men. For the changes toward less burden, women and men had similar frequencies. Women experienced 6% less and 7% much less burden and men experienced 7% less and 6% much less burden. Judging from the results of the multinomial (and also ordered) logistic regressions, men and women did not differ significantly in the change in subjective burden (see Tables A1 and A2 in the Appendix). Consequently, when measured in terms of subjective change, our findings did not provide evidence for a widening or shrinking gender gap in caregiver burden.

FIGURE 2

Descriptive differences between women and men in the subjective change in caregiver burden during the Spring 2020 lockdown in the Netherlands. N = 1209

TABLE A2

Ordered logistic regressions with clustered standard errors

Predictors	Change in subjective burden	Change in objective burden
Women (ref. men)	0.141 (0.138)	−0.269* (0.117)
Age	−0.011* (0.005)	−0.005 (0.004)
Number of people cared for	0.057 (0.090)	0.230** (0.074)
Threshold parameter 1	−3.099*** (0.344)	−1.062*** (0.281)
Threshold parameter 2	−2.338*** (0.335)	0.958*** (0.281)
Threshold parameter 3	1.192*** (0.321)
Threshold parameter 4	2.603*** (0.328)
Observations	1209	1145

Note: Coefficients reported as log odds and standard errors in parentheses.

Significance levels: *p < 0.05, **p < 0.01, ***p < 0.001.

Objective change in caregiver burden

Our second measure compared for each caregiving spell how much burden the caregiver reported in April/May with the burden reported in March.

Figure 3 summarizes the changes toward less, stable, or more burden in April/May than in March by gender. Overall, most caregivers (46%) did not experience any difference in burden, whereas equally sized groups had less burden (27%) and more burden (27%). However, these patterns differed by gender. Women were more likely than men to change toward less burden (30% of the women compared to 23% of the men), whereas men were more likely than women to change toward more burden (29% of the men compared to 25% of the women).

FIGURE 3

Descriptive differences between men and women in the change in caregiver burden between March and April/May 2020 in the Netherlands. N = 1145

The multinomial regression analysis confirmed that women were statistically more likely than men to have less burden in April/May than in March compared to no change in burden (b = 0.35, p = 0.02, see Table A1 in the Appendix). The tendency that caregiver burden increased more often for men than for women appeared nonsignificant (b = −0.03, p = 0.83, see Table A1 in the Appendix). The ordered logistic regression confirmed that, overall, men had a significantly higher increase in burden than women during the Spring 2020 lockdown (b = −0.27, p = 0.02, see Table A2 in the Appendix).

To get a more detailed picture of the changes in burden between March and April/May for men and women, Table 1 presents the cross‐tabulation of burden by gender. Table 1 demonstrates that a ceiling effect was not explaining the closing gender gap. After all, in March, only a small minority of men and women reported a heavy burden (score 5). A bottom effect could be in place, as in March more men (190/524 = 36.3%) than women (174/621 = 28.0%) reported to experience no burden at all (score 1). This implies that more men than women were, by definition, unable to experience a decrease in burden. The most striking difference in the patterns of men and women was the substantially higher likelihood of women to change from the heaviest level of burden to lower levels. In all, we can conclude that the gender gap in burden became smaller during the 2020 Spring lockdown in the Netherlands and that this was especially the result of women being more likely to experience a decrease in burden.

TABLE 1

Cross‐tabulation of caregiver burden in March and caregiver burden in April/May split by sex

Women	Burden in March
Burden in Apr/May	1—No burden	2	3	4	5—Heavy burden	Total
1—No burden	97	42	17	12	9	177
	55.75	23.33	11.97	12.50	31.03	28.50
2	51	96	40	17	2	206
	29.31	53.33	28.17	17.71	6.90	33.17
3	22	23	55	31	7	138
	12.64	12.78	38.73	32.29	24.14	22.22
4	4	18	25	27	9	83
	2.30	10.00	17.61	28.13	31.03	13.37
5—Heavy burden	0	1	5	9	2	17
	0.00	0.56	3.52	9.38	6.90	2.74
Total	174	180	142	96	29	621
	100.00	100.00	100.00	100.00	100.00	100.00

Note: First row has frequencies and second row has column percentages.

CONCLUSION

In this research note, we looked at changes in the gender gap in informal caregiving intensity and caregiver burden due to COVID‐19. Dutch caregivers were not affected in one and the same direction by the COVID‐19 pandemic. The measurements by the government forced some of the caregivers to increase their caregiving intensity, whereas others decreased it or even stopped providing care due to the COVID‐19 circumstances. Our results showed that women, more often than men, reduced caregiving, including stopping altogether, as well as increased caregiving during the pandemic. This implies that no straightforward answer could be given to the question of changes in the gender gap in caregiving intensity during the Spring 2020 lockdown. Instead, we observed a two‐sided pattern. One part of the story is that women stopped or substantially reduced providing care more often than men and the other part of the story is that women more likely increased their intensity.

The varying consequences of the COVID‐19 pandemic for caregivers also showed from our results regarding caregiver burden. Providing care became more burdensome for some, whereas for others it resulted in relieving levels of burden. These patterns were partly gender‐specific: the gender gap in caregiver burden, measured by comparing caregivers' levels in April/May with those in March, was narrowing down during Spring 2020, especially because women, more likely than men, experienced lower burden than before the pandemic. This means that the COVID‐19 pandemic redistributed the burden of caregiving more equal to women and men.

When measured in terms of subjective change in caregiver burden, we did not find significant differences between women and men. The fact that we found different results for the subjective and objective change in burden is insightful for other studies that only include one type of measurement, for instance, because longitudinal data are unavailable. The subjective measure has the advantage that it directly asked whether the change in burden was due to the COVID‐19 pandemic. This is relevant because burden generally increases with caregiving duration (see Swinkels et al., 2019); hence, the subjective measure may be better in ruling out other factors that could have increased burden. However, we have no reason to believe that the general increase in burden due to, for instance, deteriorated health of the care recipient or cumulated levels of stress for the caregiver, over this short time span, should be different for women and men. The downside of the subjective measure is that the way of answering such a question might be gendered. Previous research found that women tend to report more burden in the same caregiving situations than men because women often put more emphasis on caregiving than men and, therefore, feel greater concerns (Swinkels et al., 2017). Considering this, it could be that women are more hesitant to admit directly that they felt the same or even less burden during a global pandemic. This issue is solved with the difference measure, using longitudinal data, where the already higher reported burden is accounted for and we focused on the objective change in burden.

Our research note had the aim to describe how the gender gap in informal care provision and caregiver burden changed during the Spring 2020 COVID‐19 lockdown in the Netherlands. A next step would be to address the question of why the gender gap in burden narrowed down and why the intensity of caregiving changed differently for women and men. Some explanatory factors might be that women and men deal differently with the advice by the government, felt more or less urge to help more, or had more or less time for caregiving depending on their employment or other duties. Another explanatory factor could be that women and men are in different caregiving situations and are, therefore, differently affected by the pandemic. Future research will have to disentangle these interconnected factors in detail because of the implication informal caregiving has on the well‐being of caregivers and their labor market participation. For now, we saw that a societal change like the COVID‐19 pandemic impacted the gender gap in caregiver burden and made it more equal. Aging societies might lead to a similar shortage of formal and informal help like during the COVID‐19 pandemic and our results give a foundation to believe that in this case the burden coming with this shortage might be more equally distributed among women and men.

CONFLICT OF INTEREST

No conflict of interest has been declared by the authors.

Supplementary Material

Click here for additional data file.