Return to Work of Women after Childbirth and the Reasons of Resumption of Labor

Let me proceed to the initial analysis of the first post-natal leave now.

Summary statistics

The first post-natal leave

The table below presents the results of first simple analysis from Stata, showing the number of observations of different subgroups of women, as well as the numbers for early returns, for returns after 3 years and for those on a post-natal leave at the moment of the survey:

Table 7. Summary statistics for women who returned to work early, who returned to work after the 3^rd child's birthday and those who did not return to work at the time of the survey. The first post-natal leave of mothers of one, two and three children.

Variables	Total number of women	Early RTW (before 3 years)	RTW after 3 years	Women on PNL
PNL, total number of women in the sample	129	56	54	19
- Women who have one child - Women who have one or two more children	79 50	39 17	24 30	16 3
- Women who were employed at CB - Women who were not employed at CB	82 47	42 14	31 23	9 10
- Women who studied at CB - Women who did not study at CB	43 86	16 40	22 32	5 14
- Women who had work experience (>0 years) at CB - Women who had no work experience (=0 years) at CB	110 19	51 5	44 10	15 4
- Women whose return-to-work (RTW) decision was influenced by social norms (to RTW when a child is 3 years old) - Women whose RTW decision was not influenced by social norms	40 89	25 31	15 39	0 19
- More work committed women (who would RTW after CB, even if there would be enough money for a comfortable living) - Less work committed women (who would not RTW after CB, if there would be enough money for comfortable living)	59 70	32 24	21 33	6 13
- Women whose partner has a positive attitude to possibly long PNL - Women whose partner has a negative attitude to possibly long PNL	96 33	38 18	42 12	16 3
- Women whose child's average state of health before women's RTW is evaluated as good (relatively rare diseases) - Women whose child's average state of health before women's RTW is evaluated as bad (relatively often diseases)	85 44	42 14	30 24	13 6
- Women who stated need in their wage for a family after CB - Women who stated no need in their wage for a family after CB	112 17	55 1	42 12	15 4
- Women whose work is/was relatively far from home after CB - Women whose work is/was relatively near or at home after CB	34 95	15 41	16 38	3 16
- Women whose financial costs from home to work and back are/were significant in contrast to their salary - Women whose financial costs from home to work and back are/were insignificant in contrast to their salary	3 126	2 54	0 54	1 18
- Women who has/had access to public childcare system - Women who has/had no access to public childcare system (for whom it is/was not available)	112 17	46 10	51 3	15 4
- Women who has/had access to private childcare system - Women who has/had no access to private childcare system (for whom it is/was not available)	46 83	13 43	24 30	9 10
- Women who can/could hire a nanny for a child in order to RTW - Women who cannot/could not hire a nanny for a child	40 89	13 43	20 34	7 12
- Women whose relatives can/could help with childcare for a woman to RTW - Women whose relatives cannot/could not help with childcare for a woman to RTW	86 43	42 14	33 21	11 8
- Women who planned an early RTW (while planning to have a baby) - Women who did not plan an early RTW (while planning to have a baby)	74 55	40 16	26 28	8 11
- Women who can/could combine breastfeeding (BF) and work after RTW (it is/was possible and allowed) - Women who can/could combine breastfeeding (BF) and work after RTW (it is/was possible and allowed)	22 107	10 46	6 48	6 13
- Women who finished school at CB - Women who finished college at CB - Women who did not finish university at CB - Women who finished university at CB	11 56 21 41	4 24 8 20	7 23 10 14	0 9 3 7

To make things easier for an understanding, I will analyze the first post-natal leave in contrast to the last one, noting only the differences.

There are 129 observations in the sample for the first post-natal leave, just as in the previous case for the last post-natal leave. Of them, 56 women (43.4%) returned early, before the 3^rd birthday of their baby, 54 women (41.9%) returned after this date and 19 women (14.7%) were on post-natal leave at the moment of the study. My hypothesis that the last post-natal leave is more representative in the sense of early returns to work is confirmed, as there are 43% of women who resumed labor early after the first post-natal leave, in contrast to 47% of those who returned to work early after the last one.

The differences that are logical: 33% of the respondents have studied at the moment of the first post-natal leave versus 24% at the time of the last one; 85% of women had work experience at the time of the first post-natal leave, in contrast to 91% of those who had some work experience at the time of the last one. Also, there is a difference in time trend concerning the percentage of women who finished university at the first and at the last post-natal leave.

I can suppose, that `work commitment' rises from the first to the last post-natal leave, as there were 46% of relatively more work committed respondents at the moment of the first post-natal leave and 52% at the moment of the last one.

There is a little difference through time in the trend concerning an availability of private childcare: 36% of women had a possibility of using a private kindergarten at the time of the first post-natal leave, in contrast to 38% of women on the last one. Although this is not consistent with the pattern concerning family income, as 49% of women who declared the need in their wage in a family income returned early after the first post-natal leave versus 53% of women after the last one.

Perhaps, through time women solve the problems of the long distance between home and work, as 26% of women stated that they lived far from work after the first post-natal leave, in contrast to 22% after the last one.

Thus, it seems that at first glance the number of children in a family, the employment status at childbirth, the length of work experience at the moment of childbirth, the degree of influence of social norms, higher degree of work commitment, a partner's attitude towards the length of the post-natal leave, average state of child's health, need in woman's wage in a family income, an availability of childcare by relatives, as long as time trend seem to matter in the return-to-work decision.

Now let me proceed to further stage of analysis.

Statistical Analysis and Results

The last post-natal leave

At the beginning, I decided to investigate the last post-natal leave, as I believed that it is more representative, as I said above. When I say `the last post-natal leave', I mean the only post-natal leave of women with one child and the second and the third post-natal leaves of mothers of two and three children, respectively. The Stata's output with the abstracts of the multinomial logistic regression and marginal effects are in the Appendix. Here I present convenient tables with clue independent variables (which appeared to be significant with P values less than 0.1), their P values and the marginal effects in the form of coefficients for three groups of women (who returned early, who returned after the 3^rd birthday of their child and those on a post-natal leave at the time of the study):

Table 8. Marginal effects for women who RTW early. Significant variables. The last post-natal leave.

Variable	Coefficient	P value
Work experience at CB	0.07	0.000
Whether social norms influence return to work (RTW), dummy	0.22	0.016
Need in woman's wage in family income, dummy	0.44	0.000
Financial costs on way to work and back, dummy	-0.34	0.000
Availability of combining breastfeeding and work, dummy	0.21	0.032
Woman's age at CB	-0.03	0.017
Trend	-0.01	0.058

The following factors are highly significant (with P values less than 0.001) in case of early return of women after childbirth:

· The length of work experience of a woman at the moment of childbirth

· Need in woman's wage in a family income

· Financial costs on way to work and back

The following factors are significant too (with P values less than 0.1):

· The degree of influence of social norms concerning the length of a post-natal leave

· Availability of combining breastfeeding and work

· Age of a woman at childbirth

Table 9. Marginal effects for a return after 3 years after childbirth. The last post-natal leave.

Variable	Coefficient	P value
Whether there are other children in a family, dummy	0.17	0.053
Child's average state of health, dummy	-0.19	0.022
Need in woman's wage in family income, dummy	-0.51	0.000
Availability of combining breastfeeding and work, dummy	-0.16	0.082
Trend	-0.01	0.04

The following factor is highly significant (with P value less than 0.001) in case of return to work after 3 years on a post-natal leave:

· Need in woman's wage in a family income

The following factors are significant too (with P values less than 0.1):

· The number of children in a family

· Average state of health of a child

· Availability of combining breastfeeding and work

· Time trend

Table 10. Marginal effects for women who have been on a post-natal leave at the time of the survey. The last post-natal leave.

Variable	Coefficient	P value
Whether there are other children in a family, dummy	-0.11	0.082
Work experience at CB	-0.04	0.004
Whether social norms influence return to work (RTW), dummy	-0.21	0.000
Work commitment, dummy	-0.17	0.000
Husband's attitude towards the length of post-natal leave (PNL), dummy	0.10	0.055
Child's average state of health, dummy	0.13	0.005
Distance between work and home, dummy	-0.16	0.000
Financial costs on way to work and back, dummy	0.57	0.000
Availability of private childcare, dummy	0.24	0.001
Availability of a nanny for a child, dummy	-0.14	0.008
Availability of relatives' help in childcare, dummy	-0.16	0.006
Trend	0.02	0.000

The following factors are highly significant (with P values less than 0.001) in case of being on a post-natal leave after childbirth:

· The degree of influence of social norms concerning the length of a post-natal leave

· The degree of work commitment of a woman

· Distance between home and work

· Financial costs on way to work and back

· Time trend

The following factors are significant too (with P values less than 0.1):

· The number of children in a family

· The length of work experience of a woman at the moment of childbirth

· Husband's attitude towards the length of post-natal leave

· Average state of health of a child

· Availability of private childcare

· Availability of a nanny for a child

· Availability of relatives' help in childcare

When I changed the variable indicating woman's age at childbirth with the dummy variable indicating childbirth after 23 years old (as a remedy for high correlation between woman's age at childbirth and work experience at childbirth), it was insignificant in all groups, in contrast to the variable indicating woman's age at childbirth. Thus, I present here the results of such an analysis only briefly. For the group of women who return after 3 years on a post-natal leave, the variables indicating financial costs and the distance between home and work become significant, while the variable indicating an availability of combining of breastfeeding and work shows the opposite effect. For women on a post-natal leave at the time of the study, the variables indicating the number of children in a family and a partner's attitude towards the length of a post-natal leave become insignificant.

I tryied to use multinomial probit regression model too, however, in order to get the results I had to remove the variable indicating an unfinished university level of education at childbirth. I used official guides and support sites, where I found the information that in order to get read of the problem, which I faced, I need eigher to remove all the independent variables from the regression and to add one at a time and examine, whether I can run the regression with it and without the bug, and then to add only these variables into the model - this is what I have done, or I need to form a new dependent variable with less outcomes, however, this is not my case, because I wanted to study no less than three outcomes, the early return to work, return after 3 years and the group for women on a post-natal leave, I did not suppose to study only two outcomes. That is why I decided that using the multinomial logistic model is preferential, as I do not need to exclude variables for it to work properly without bugs or forming a new dependent variable with less than three outcomes.

Thus, I will not present a table with results here, however, I will briefly tell what changes with the probit model (with the excluded variable). For the group of women who return early, nothing changes. For the group of those who return after the 3^rd birthday of their child, the variables indicating distance from home to work, financial costs on way to work and back and college level of education at childbirth become significant in the multinomial pribit model, while the variables indicating the number of children in a family and an attitude of a woman's partner towards the length of a post-natal leave become insignificant.

In the multinomial probit regression model with the dummy variable indicating a birth after 23 years old instead of the continuous variable indicating woman's age at childbirth (as a remedy for the problem of high correlation between age and work experience at childbirth), the group of women who return early stays the same with the new dummy variable being insignificant. For the group of those who return after the 3^rd child's birthday, the variables indicating the distance between home and work, finanacial costs on way from home to work, an availability of a public kindergarten and college level of education at childbirth become significant, while an availability af combining of breastfeeding and work becomes insignificant. For women on a post-natal leave the variable indicating college level of education at childbirth becomes significant, in contrast to the following variables: an attitude of a woman's partner towards the length of a post-natal leave, the number of children in a family and an availability of a nanny for a child, which become insignificant.

7. The first post-natal leave

The next stage is the analysis of the first post-natal leave of mothers of one, two and three children, which is also conducted using multinomial logistic regression model. Thus, the comparisons can be made between the first and the last post-natal leaves and with the results of the previous research of the first post-natal leaves of women after childbirth in other countries. The convenient tables with the marginal effects of the same three groups of women are below (only significant variables are presented, as their P values suggest):

Table 11. Marginal effects for women who RTW early. Sign vars.

Variable	Coefficient	P value
Work experience at CB	0.06	0.005
Whether social norms influence return to work (RTW), dummy	0.18	0.048
Work commitment, dummy	0.20	0.008
Child's average state of health, dummy	0.19	0.025
Need in woman's wage in family income, dummy	0.40	0.000
Availability of private childcare, dummy	-0.28	0.001
Availability of a nanny for a child, dummy	0.18	0.093
Availability of relatives' help in childcare, dummy	0.19	0.013
Woman's age at CB	-0.04	0.064
Trend	-0.01	0.084

The following factor is highly significant (with P value less than 0.001) in case of early return of women after childbirth:

· Need in woman's wage in a family income

The following factors are significant too (with P values less than 0.1):

· The degree of influence of social norms concerning the length of a post-natal leave

· The length of work experience of a woman at the moment of childbirth

· The degree of work commitment of a woman

· Average state of health of a child

· Availability of private childcare

· Availability of a nanny for a child

· Availability of relatives' help in childcare

· Age of a woman at childbirth

· Time trend

Table 12. Marginal effects for a return after 3 years after childbirth.

Variable	Coefficient	P value
Child's average state of health, dummy	-0.23	0.007
Need in woman's wage in family income, dummy	-0.48	0.000
Availability of public childcare, dummy	0.21	0.075
Availability of combining breastfeeding and work, dummy	-0.17	0.084

The following factor is highly significant (with P value less than 0.001) in case of return to work after 3 years on a post-natal leave:

· Need in woman's wage in a family income

The following factors are significant too (with P values less than 0.1):

· Average state of health of a child

· Availability of public childcare

· Availability of combining breastfeeding and work

Table 13. Marginal effects for women who have been on a post-natal leave at the time of the survey.

Variable	Coefficient	P value
Work experience at CB	-0.03	0.004
Whether social norms influence return to work (RTW), dummy	-0.20	0.000
Work commitment, dummy	-0.12	0.029
Husband's attitude towards the length of post-natal leave (PNL), dummy	0.15	0.000
Availability of public childcare, dummy	-0.21	0.023
Availability of private childcare, dummy	0.25	0.000
Availability of a nanny for a child, dummy	-0.10	0.076
Availability of relatives' help in childcare, dummy	-0.12	0.062
Woman's age at CB	0.04	0.008
Trend	0.01	0.004

The following factors are highly significant (with P values less than 0.001) in case of being on a post-natal leave after childbirth:

· The degree of influence of social norms concerning the length of a post-natal leave

· Husband's attitude towards the length of post-natal leave

· Availability of private childcare

The following factors are significant too (with P values less than 0.1):

· The length of work experience of a woman at the moment of childbirth

· The degree of work commitment of a woman

· Availability of public childcare

· Availability of a nanny for a child

· Availability of relatives' help in childcare

· Age of a woman at childbirth

· Time trend

I used multinomial probit model also. However, I had to remove four variables from the regression for the model to work properly Again, the guides and support says that I need either to exclude some variables (this is exactly what I have done), or to form the dependent variable in such a way to make less outcomes. However, the latter did not seem like a good idea to me. (as there were bugs which did not allow to get the results). Thus, I decided to choose the multinomial logistic model instead of multinomial probit model, as it gives the results with all the variables.

When I changed the variable indicating woman's age at childbirth with the dummy variable indicating childbirth after 23 years old (as a remedy for high correlation between woman's age at childbirth and work experience at childbirth), it appeared to be significant only for the third group of women, who were on post-natal leave at the time of the survey. With this variable in the analysis, the variables indicating an availability of a nanny becomes insignificant for early returns, an availability of public childcare becomes insignificant for returns to work after 3 years on post-natal leave; an availability of a nanny, an availability of public childcare, and work commitment become insignificant for women on post-natal leave at the time of the survey.

Out of curiosity and as an additional analysis, I examined the second post-natal leave (for the second child) for women with two and three children. However, there were lack of observations - only 50 - which did not allowed me to get adequate results, as even when I ran the multinomial logistic regression on even five independent variables, four out of five variables were insignificant with p values very close to 1 for women who have been on post-natal leave at the time of the survey Output from Stata 17. Appendix. Multinomial logistic regression with five independent variables.. And almost all the variables for these group of women were insignificant when I ran a regression on the same number of variables, as in the analysis of the first post-natal leave.

8. Discussion

The last post-natal leave

Table 14. (Rewritten Tables 8-10) The last post-natal leave. Marginal effects for women who RTW early, return after 3 years on PNL or are at home on a post-natal leave. Significant variables.

Variable	The group	Coefficient	P value
Work experience at CB	RTW early Non-RTW	0.07 -0.04	0.000 0.004
Whether there are other children in a family, dummy	RTW after 3 Non-RTW	0.17 -0.11	0.053 0.082
Whether social norms influence return to work (RTW), dummy	RTW early Non-RTW	0.22 -0.20	0.016 0.000
Work commitment, dummy	Non-RTW	-0.17	0.000
Child's average state of health, dummy	RTW after 3 Non-RTW	-0.19 0.13	0.022 0.005
Husband's attitude towards the length of post-natal leave (PNL), dummy	Non-RTW	0.10	0.055
Need in woman's wage in family income, dummy	RTW early RTW after 3	0.44 -0.51	0.000 0.000
Financial costs on way to work and back, dummy	RTW early Non-RTW	-0.34 0.57	0.000 0.000
Availability of private childcare, dummy	Non-RTW	0.24	0.001
Availability of a nanny for a child, dummy	Non-RTW	-0.14	0.008
Availability of relatives' help in childcare, dummy	Non-RTW	-0.16	0.006
Availability of combining breastfeeding and work, dummy	RTW early RTW after 3	0.21 -0.16	0.032 0.082
Woman's age at CB	RTW early	-0.03	0.017
Trend	RTW early RTW after 3 Non-RTW	-0.01 -0.01 0.02	0.058 0.04 0.000

For the last-post natal leave, every year of work experience at the moment of childbirth increases the probability of returning to work early by 7% and decreases the probability of not returning to work (by staying on a post-natal leave) by 4%. The effect of this factor on the group of women who return after 3 years on a post-natal leave is not significant.

According to the results, the presence of more than one child in a family increases the probability of resuming labor after 3 years on a post-natal leave by 17% and decreases the probability of not returning to work by 11%, However, here may be more to correlation than to causal relation, as there may be other factors that influence both the decision about the number of children in a family and the decision about the length of a post-natal leave, for example such a characteristic as more family-orientation of a woman.

The fact of influence of social norms on a woman's return-to-work decision increases the probability of resuming work early by 22%, decreasing the probability of not returning by 20%. This is high enough, thus I can tell that my hypothesis about the importance and influence of social norms on women's decision concerning the length of a post-natal leave can be confirmed. More work commitment of a woman decreases the probability of not returning to work by 17%, which is rather logical.

Rare diseases or a good state of health of a child on average decreases the probability of returning to work after the 3^rd child's birthday, at the same time increasing the probability of being on a post-natal leave by 13%. This is rather interesting result, as I hypothesized that child's good state of health on average allows women to resume labor earlier, and here I see that it increases the probability of not returning to work. Perhaps, mothers of children with not so good state of health on average are forced to return earlier as they want to earn money for something that helps to support the immunity and health of their child, for example, swimming pool of a trip to the sea (this is just my assumption).

Positive attitude of a partner towards the possibility of long post-natal leave increases the probability of staying on a post-natal leave (or not returning to work) by 10%. Thus, the hypothesis that preferences of a woman's family are among the influential factors, which matter in a woman's return-to-work decision.

The factor of need of a woman's salary in a family budget is very important for early returns and for returns after the 3^rd birthday of a child, which is consistent with `Omnibus VCIOM' results that financial need is one of the most important reasons of resuming labor by women after childbirth. If there is a need in woman's wage, the probability of resuming labor after 3 years on a post-natal leave decreases by 51% and the probability of early return to work increases by 44%.

The presence of significant costs on way to work and back in contrast to a woman's wage is as important, as the presence of need: the probability of early returns decreases by 34% and the probability of staying on a post-natal leave increases by 57%, if costs are significant. So, the other my hypothesis about the importance of financial costs on a way to work and back home is confirmed.

An interesting pattern an availability of a private kindergarten shows - the probability of not returning to work increases by 24%, if private childcare is available to a family, which contradicts with one of my hypotheses. I hypothesized that if alternative private childcare services are available to a family, then woman returns to work earlier. However, here the variable of an availability of private childcare may behave like a proxy for the level of husband's or family's income. Thus, if income is high enough for a family to afford private childcare, than a woman may stay at home longer, I suppose.

An opposite effect shows the factor indicating an availability of a nanny for a child, if a nanny is available to a family, the probability of staying on a post-natal leave decreases by 14%. The similar effect shows the factor of an availability of help of relatives - if such help with childcare is available to a family, the probability of not returning to work decreases by 16%. An availability of combining of breastfeeding and work increases the probability of resuming labor early by 21%. This means that an availability of various options of childcare and the possibility to combine family and work responsibilities are among the important factors, which influence a woman's decision whether to return to work, and if yes, than when to return.

Every year of woman's age at the moment of childbirth decreases the probability of early return by 3%.

The influence of the time trend is not great, but the factor is highly significant. The results suggest that every point of time trend decreases the probabilities of return to work before and after the 3^rd child's birthday by 1% for each outcome, but increases the probability of staying on a post-natal leave by 2%. The measure of the time trend is not just the year of birth of a child, but the difference between the year of childbirth and the minimum year of childbirth among the respondents plus one, created in order to get `the time scale' to see how time influences women's decisions.

Thus, among the factors, which are very important for the resuming labor before the 3^rd birthday of a child are:

· The length of work experience of a woman at childbirth (every year increases the probability of early return by 7%)

· The influence of social norms

· Need of a woman's salary in a family budget

· An availability of combining of breastfeeding and work

Among the factors, which are important for the resuming work after the 3^rd birthday of a child and for the returning to work after a post-natal leave are:

· The length of work experience of a woman at the moment of childbirth

· Influence of social norms

· More work commitment of a woman

· Need of a woman's salary in a family budget

· Negative attitude of a woman's partner to the possibility of a long post-natal leave

· An availability of a nanny for a child

· An availability of help of relatives

Thus, such factors as the length of work experience of a woman at childbirth, the presence of the social norms limiting the length of a post-natal leave, work commitment of women, the need in their salary in a family income, the size of financial costs of a way to work and back home and an availability of various forms of childcare and the possibility to combine work and family responsibilities are of great importance for the decision about resuming labor by new mothers after the last post-natal leave.

9. The first post-natal leave

The next step was to analyze the first post-natal leave of different groups of women, among them were women who returned to work early, i. e. before 3 years of their child, women who returned after the 3^rd birthday of their child and women who were on post-natal leave at the time of the survey. In addition, there was aim to examine the characteristics of these groups, to understand what they differ in. For these aims I ran multinomial logistic regression. Let me now discuss the results of this stage of analysis, using the rewritten Tables 9-11 with the marginal effects for three groups of women:



Table 15. (Rewritten Tables 11-13) The first post-natal leave. Marginal effects for women who RTW early, return after 3 years on PNL or are at home on a post-natal leave. Significant variables.

Variable	The group	Coefficient	P value
Work experience at CB	RTW early Non-RTW	0.06 -0.03	0.005 0.004
Whether social norms influence return to work (RTW), dummy	RTW early Non-RTW	0.18 -0.20	0.048 0.000
Work commitment, dummy	RTW early Non-RTW	0.20 -0.12	0.008 0.029
Child's average state of health, dummy	RTW early RTW after 3	0.19 -0.23	0.025 0.007
Husband's attitude towards the length of post-natal leave (PNL), dummy	Non-RTW	0.15	0.000
Need in woman's wage in family income, dummy	RTW early RTW after 3	0.40 -0.48	0.000 0.000
Availability of public childcare, dummy	RTW after 3 Non-RTW	0.21 -0.21	0.075 0.023
Availability of private childcare, dummy	RTW early Non-RTW	-0.28 0.25	0.001 0.000
Availability of a nanny for a child, dummy	RTW early Non-RTW	0.18 -0.10	0.093 0.076
Availability of relatives' help in childcare, dummy	RTW early Non-RTW	0.19 -0.12	0.013 0.062
Availability of combining breastfeeding and work, dummy	RTW after 3	-0.17	0.084
Woman's age at CB	RTW early Non-RTW	-0.04 0.04	0.064 0.008
Trend	RTW early Non-RTW	-0.01 0.01	0.084 0.004

Every year of work experience at the moment of childbirth increases the probability of resuming labor before the 3^rd birthday of a child by 6% and decreases the probability of non-returning by 3%. The effect of this factor on the first post-natal leave is quite similar to its effect on the last one.

The influence of social norms increases the probability of returning early by 18% and decreases the probability of non-returning by 20%. The presence of work commitment increases the probability of resuming labor early by 20% and decreases the probability of non-returning by 12%, which is again quite similar to the last post-natal leave.

What differs is that child's good state of health, or relatively rare diseases, in other words, increases the probability of returning early by 19% (and decreases the probability of returning after 3 years on post-natal leave by 23%). Thus, for the first post-natal leave the hypothesis, that good child's health allows a woman to return to work earlier, can be confirmed, in contrast to the last post-natal leave.

Need in woman's salary in a family income increases the probability of returning early by 40%, and this factor is highly significant for this group. It is also highly significant for the group of women who return after 3 years on post-natal leave and the need in woman's wage decreases the probability of returning after the 3^rd child's birthday by the whole 48%. This factor, thus, is the most important for the first post-natal leave, as well as for the last one.

Positive husband's attitude towards the possibility of a long post-natal leave increases the probability of non-returning by 15% (10% for the last post-natal leave). So, for the first post-natal leave the hypothesis about the importance of preferences of a woman's family can also be confirmed.

An availability of public kindergartens decreases the probability of non-returning by 21%, increasing the probability of returning after the 3^rd birthday of a child by the same 21%, which is not surprisingly and is rather logical, and my hypothesis about the importance of an availability of public childcare system is confirmed.

A very interesting result concerns an availability of private childcare - it decreases the probability of early return by 28%, increasing the probability of non-returning by 25%. Such a result here may also be explained by the assumption that an availability of private childcare behaves like a proxy for a high level of husband's income, making it possible for a woman not to return to work or to do it later, just as for the last post-natal leave.

However, there is an opposite effect for the factor indicating an availability of a nanny for a child - though there was a high correlation between this variable and those for an availability of private childcare - it increases the probability of early returns by 18% and decreases the probability of non-returning to work after a post-natal leave. Availability of relatives' help in childcare increases the probability of early resumption of labor by 19%, decreasing the probability of non-returning by 12%. An availability of combining of breastfeeding and work decreases the probability of the return after 3 years on a post-natal leave by 17%.

Every year of woman's age at childbirth in my sample decreases the probability of early return to work by 4% (and by 3% for the last post-natal leave), increasing the probability of non-returning by 4% at the same time. In addition, every point of the variable indicating trend decreases the probability of early resumption of labor by 1%, increasing the probability of non-returning by 1%.

Thus, among the factors, which are very important for the resuming labor before the 3^rd birthday of a child are:

· The length of work experience of a woman at childbirth (every year increases the probability of early return by 6%)

· The influence of social norms

· Work commitment of a woman

· Child's good state of health, or relatively rare diseases

· Need of a woman's salary in a family budget

· An availability of a nanny for a child

· An availability of help of relatives

Among the factors, which are important for the resuming work after the 3^rd birthday of a child and for the returning to work after a post-natal leave are:

· The length of work experience of a woman at the moment of childbirth

· Influence of social norms

· More work commitment of a woman

· An availability of public kindergartens

· Negative attitude of a woman's partner to the possibility of a long post-natal leave

· An availability of a nanny for a child

· An availability of help of relatives

Thus, the following factors appeared to be very important in the decision about resuming labor by new mothers after the first post-natal leave: the length of work experience of a woman at childbirth, the presence of the social norms limiting the length of a post-natal leave, work commitment of women, the need in their salary in a family income and an availability of various forms of childcare (relatives' care, public childcare system and a nanny). The difference in the results between the first and the last post-natal leave is that for women after the first post-natal leave a child's state of health and an availability of a public kindergarten are significant, while for the last one, an availability of combining of breastfeeding and work and the size of financial costs on a way to work also make a difference. The other contributing factors are quite similar for both post-natal leaves.

Positive husband's attitude towards the possibly long post-natal leave and an availability of private childcare increase the probability of staying on post-natal leave in both cases.

At this point, the comparisons can be made between my results of the analysis of the first post-natal leave and the previous research of scholars.

Shapiro and Mott and Saurel-Cubizolles at all. in their papers suggested that social norms and policy, which guarantee a workplace and wages of women before and after childbirth might help new mothers to return to work after childbirth and to show more work attachment, and my results about the importance of social norms and women's work commitment in the decision of resuming labor by women after childbirth is consistent with the surveys of Saurel-Cubizolles at all and Shapiro and Mott.

Susan McRae suggests that the creation of new forms of childcare by the government and employers may give new possibilities for working mothers in Britain. Klerman and Leibowitz also emphasizes the role of relatives's in childcare for a woman to resume labor sooner. My study confirms these results, as an availability of different forms of child daycare and the possibility of combining work and family responsibilities allow new mothers to return to work sooner. In addition, my survey confirm Susan McRae's results that financial need and the number of children in a family are very important factors in the return-to-work decision, with the former to be the most influential.

Eva Bernhardt wrote about the importance of work experience and that it influences earlier return to work, which is true for our reality too, as the length of work experience is of great importance both for resuming labor both before and after the 3^rd child's birthday.

According to Desai and L. Waite, `women's preference of being employed over being at home', which I call work commitment in my survey, is among the clue factors, and my work confirms this result.

However, I presented some results, which differ from the previous research greatly. For example, the length of education, which was emphasized by many scholars as one of the main factors, is not a significant factor neither for the first, nor for the last post-natal leave in our country nowadays. It become significant after some manipulations with the variables, and only the college status becomes significant, but not in the basic versions of the model.

An interesting result was for the private childcare in the form of private kindergartens. However, this may be because of scarcity outside the capital region and because of the price, which is only available for families with a high level of income. I suppose that if subsidies would be paid for the use of this alternative form of childcare when a child is under 3 years old, then more women could return to work earlier.

In addition, an availability of combining of breastfeeding and work and the size of financial costs on a way to work are among clue factors contributing to the return-to-work decision of women after childbirth. However, these factors are important for the last, not the first post-natal leave.

An interesting question is what can be done in order to help women to return to work earlier after a post-natal leave? According to my results, I can suggest the following:

· To save the norms, which regulate the length of a post-natal leave after childbirth and ensure the saving of a woman's workplace while she is on a post-natal leave

· Methods, which may help in increasing work commitment of women. Of course, this may be more to a woman's character, however, work commitment may be increased to some degree by higher wages, better job perspectives, friendly atmosphere and relations at a workplace and other material and non-material bonuses, which may be especially important for new mothers

· Creating alternative public and private forms of child daycare facilities

· Subsidies for using private forms of childcare

· Subsidies on way to work and back home for new mothers

· Better child's health care

· Facilities allowing for combining work and breastfeeding in order to help women to save breastfeeding, which becomes especially important for health and relationships of a mother and a baby, when a new mother returns to work



Conclusion

In my work I attempted to study the issue of resuming labor by women after childbirth. Specifically, I wanted to answer the questions - when and why women return to work after childbirth, which factors influence their decision the most. I attempted to ask these questions by conducting my own retrospective survey in the Stary Oskol region.

In order to be able to give answers to the questions above, I created a questionnaire, conducted a survey and used several approaches, including examining the first, the second and the last post-natal leaves.

In my sample of 129 respondents, 43% of women returned before the 3^rd birthday of a child after the first post-natal leave, 42% of women returned to work after this date and 15% of women were on a post-natal leave at the time of the survey. For the last post-natal leave, figures are very similar.

The following factors are among the most important in the decision about resuming labor by new mothers after the first post-natal leave: the length of work experience of a woman at childbirth, the presence of the social norms limiting the length of a post-natal leave, work commitment of women, the need in their salary in a family income, an availability of various forms of childcare (relatives' care, public childcare system and a nanny) and a child's state of health. For the last post-natal leave an availability of combining of breastfeeding and work and the size of financial costs on a way to work are also among clue factors.

Thus, my hypotheses about the importance of professional and financial reasons, as long as social policy concerning the length of a post-natal leave can be confirmed, which is consistent with the previous research. Also, more personal factors like preferences of a woman and her family, the factor of child's health, an availability of various forms of child daycare and an the possibility of combining of work and family responsibilities are of great importance to new mothers in the decision, whether to return to work, and if yes, then when to return.

The hypothesis of the importance of educational status for resuming labor early can be rejected, according to the results of my work, which is rather new to the previous research. Also, the hypothesis of the importance of an availability of private kindergartens can be rejected.

After the analysis, I may suggest some methods aiming to help women to return to work earlier: saving and developing the social policy and norms, which limit the length of a post-natal leave and guarantee returning to the same workplace after childbirth, using different methods attempting to increase work commitment of women, creating alternative public and private forms of child daycare facilities, paying out subsidies for using private forms of childcare and subsidies on way to work and back home for new mothers, better child's health care, facilities allowing for combining work and breastfeeding. All this methods may be developed and implemented by the government and employer, and they may help women in resuming work sooner after childbirth.

Besides these assumptions above, there are also several contributions of my work. At first, this is a new place, a new country. Secondly, the survey is relevant and new. My survey answers the questions about the resuming labor by women after childbirth and the factors contributing to this decision nowadays in our country. Moreover, the idea of the studying of the last post-natal leave is new, in contrast to the previous research. The other contribution is that my analysis of the first post-natal leave allows for the comparisons with the previous studies to be made. Lastly, as the information from the survey is unique and new for our reality, it may be useful for the understanding of the limitations of women, who wish to resume labor earlier, but cannot do so for some reasons, and to help such women to fulfill their career plans, making them and the society as a whole a little bit happier.



References

1. `Family Economics', Browning, Chiappori and Weiss, 2011.

2. `Long-term Employment and Earnings of Women in Relation to Employment Behavior Surrounding the First Birth', D. Shapiro and F. Mott, `The Journal of Human Resources', 1994.

3. `Returning to Work after Childbirth in France, Italy, and Spain', Saurel-Cubizolles, Romito, Escriba-Aguir, Lelong, Pons and Ancel, `European Sociological Review', 1999.

4. `Returning to Work after Childbirth: opportunities and inequalities', S. McRae, `European Sociological Review', 1993.

5. `Child Care and Women's Return to Work after Childbirth', Klerman and Leibowitz.

6. `Women's Home Attachment at First Birth: the Case of Sweden', Eva Bernhardt, `European Journal of Population', 1986.

7. `Women's Employment during Pregnancy and after the First Birth: occupational characteristics and work commitment', S. Desai and L. Waite, `American Sociological Review', 1991.

8. Labor Code of Russian Federation.

http://www.trudkod.ru/chast-4/razdel-12/glava-41/st-256-tk-rf.

9. `Omnibus VCIOM', 1994.

http://sophist.hse.ru/db/oprview.shtml?ID_S=2273&T=m.

10. The table showing the mains aims of women concerning work.

http://sophist.hse.ru/facts/zhenrab.shtml.

11. Russian Statistics Committee (ROSSTAT). http://www.gks.ru/wps/wcm/connect/rosstat_main/rosstat/ru/statistics/population/motherhood.

12. Data from England. Office for National Statistics. https://www.ons.gov.uk/employmentandlabourmarket/peopleinwork/employmentandemployeetypes/articles/moremotherswithyoungchildrenworkingfulltime/2017-09-26.

13. Commands in Stata.

https://www.stata.com/features/documentation/.

14. `Margins and interpreting results', M. C. Perraillon, 2017. http://www.ucdenver.edu/academics/colleges/PublicHealth/resourcesfor/Faculty/perraillon/perraillonteaching/Documents/week%2013%20margins.pdf.



Appendix

A Study on resumption of labor by women after childbirth

The Study is anonymous!

Questionnaire for the first child

1. A year of birth of your first child: ___________________

2. Were you employed at the time of birth of your first child?

 Yes

 No

3. Have you studied at the time of birth of your first child?

 Yes

 No

If yes, when have you continued your study after the birth of your first child?

I have continued study when my first child was ____ years ____ months old.

4. Your work experience at the time of birth of your first child:

____ years _____ months

5. Did the social norms concerning resumption of labor by women after childbirth affect your decision of returning to work after childbirth (if you were employed at the time of birth of your first child)?

 Yes

 No

6. How old was your first child when you returned to work after his/her birth?

 My first child was ____ years ____ months, when I returned to work after his/her birth

 I did not return to work until the birth of my second child

 I did not return to work after the birth of my first child

7. Did you returned to work after the birth of your first child, if your family might live comfortably without your salary?

 Yes

 No

8. Please, evaluate your husband's/partner's attitude concerning the possibility of long post-natal leave after the birth of your first child:

 Rather positive

 Rather negative

9. Please, evaluate your first child's state of health before your returning to work:

 My child was ill frequently

 My child was ill rarely

Please, evaluate your first child's state of health within first years of his/her life, if you did not return to work after his/her birth:

 My child was ill frequently

 My child was ill rarely

10. Please, evaluate the importance of your salary in family's income at the time of deciding whether you should return to work after the birth of your first child:

 There was financial need in my salary

 My salary was not too much

 There was no need in my salary

11. Please, evaluate the distance from work to home (if you returned to work after the first childbirth)?

 I worked near from home

 I worked far from home

 I worked at home

12. Please, evaluate financial costs on way from home to work and back in comparison with your salary (if you returned to work after the first childbirth):

 Significant costs

 Insignificant costs

 No expenses

13. Did you have the possibility to attend a public kindergarten for your first child at the time of deciding whether to return to work (after the birth of the first child)?

 My first child attended a public kindergarten

 We had such possibility, but we did not use it

 We had no possibility to attend a public kindergarten

14. Did you have the financial possibility to attend a private kindergarten for your first child at the time of deciding whether to return to work (after the birth of the first child)?

 My first child attended a private kindergarten

 We had such possibility, but we did not use it

 We had no financial possibility to attend a private kindergarten

15. Did you have the financial possibility to hire a nanny for your first child at the time of deciding whether to return to work (after the birth of the first child)?

 We hired a nanny

 We had such possibility, but we did not use it

 We had no financial possibility

16. Did you plan early return to work (after the birth of the first child) at the moment of deciding to have the first child?

 Yes, I did

 No, I did not

17. Did your relatives have the possibility of helping you with childcare while you were working at the time when you decided whether to return to work after the birth of your first child?

 Yes, they had the possibility

 No, they had no possibility

Did you use this possibility?

 Yes, I did

 No, I did not

18. Did you have the possibility of combining breastfeeding and work (at the time of your returning to work after the birth of your first child)?

 Yes

 No

Did you use this possibility?

 Yes

 No

If no, why?

__________________________________________________________________________________________________________________________

19. Your age at the time of a birth of your first child: ______________________

20. Your level of education at the time of birth of your first child:
____________________________________________________________

Questionnaire for the second child

21. A year of birth of your second child: ___________________