![\"\"](\"https:\/\/emilkirkegaard.dk\/en\/wp-content\/uploads\/fertility_rr_pooled.png\")
<\/a><\/p>\nThese are just the top and bottom 25 occupations. There are some questions about the causation. The most obvious interpretation is that some jobs are easier to combine with family making than others, so this would represent a causal effect from occupation to fertility. The second most obvious interpretation is that women who want to make families don’t choose their occupations at random, but choose those that are compatible with family life or otherwise preferred by (would be) mothers, and that’s why some occupations are higher in fertility. The third interpretation is that having children caused women to move into specific occupations afterwards, thus violating the stability assumption. This third option is most likely the case for telemarketer, which is presumably something women decide to do to make some money while staying at home with young children (a good working from home\/remote job). However, for many other jobs, changing jobs is not so easy. Working as a nurse or physician (both in top 25) requires a long-ish degree and passing some test, so it is not so easy to just decide to change to that job tomorrow. Regarding family life compatibility, the interpretation is not so obvious for some jobs. In the low fertility group, we see flight attendants, which makes a lot of sense. They keep traveling around and pregnancy while working in mid-air sounds unwise (in fact, often prohibited). On the other hand, web development is easy to combine with family formation, but they are among the least fertile women. In the high fertility group, we see a lot of healthcare jobs, which I guess may have flexible hours, but also many are physical labor intensive (butcher, packaging, machine filling, agricultural workers) that don’t sound like they would make for good jobs compatible with motherhood. Overall, it doesn’t immediately seem that between job fertility rates are caused by the jobs themselves but rather must mainly be due to self-selection of preexisting differences among the women who choose them.<\/p>\n
Aside from eyeballing the top and bottom occupations we can also do some statistical analyses of them. For instance, we may expect female-typical jobs to be more compatible with family formation and attracting the women more interesting in family formation, and the data bear this out:<\/p>\n
The correlation is not fantastically large, however, about .20.<\/p>\n Occupations can also be rated in various ways for which kind of people they attract or skills they require. Many decades ago, scientists tried to come up with dimensions that describe the kind of actions performed at jobs. The resulting model was the RIASEC<\/a> (John Holland’s model), which is just a mnemonic for “Realistic (Doers), Investigative (Thinkers), Artistic (Creators), Social (Helpers), Enterprising (Persuaders), and Conventional (Organizers)”. The dimensions are somewhat poorly named. Here’s some occupations that are high vs. low in each dimension (latest data):<\/p>\n So realistic are jobs involving concrete things, usually physical labor, and conventional really means busywork\/secretary-like. The others fit the meaning normally.<\/p>\n Later some people factor analyzed these 6 RIASEC dimensions since they were correlated (oblique rotation) obtained a higher order 2 factor solution, Data vs. Ideas and Things vs. People. The latter of these have gotten the most interest because the sex difference is very large on interest in jobs involving things or people, about 1 standard deviation<\/a>. One might say it represents the fundamental division of human labor. Anyway, these are the correlations for RIASEC, the 2-dimensional higher order factor model, and complexity:<\/p>\n The correlations aren’t great either, nothing above 0.18. Two of them show nonlinear patterns, social and complexity. I don’t have any particular theory of why social shows the nonlinear pattern. Complexity is very interesting since this represents the intelligence requirement of the job. Linda Gottfredson<\/a> analyzed job descriptions and requirements many years ago and found that they show a strong complexity factor with descriptions like<\/a>:<\/p>\n The fact that female fertility shows a U shaped relationship with job complexity is thus puzzling, given the usual negative correlation with intelligence\/education and fertility. Well, that is, until we look at the correlation in the plot, which is -0.05. There are more jobs along the negative linear slope than those in the relatively high fertility maximum complexity jobs. It appears that the top and the bottom of the job hierarchy are outbreeding the middle.<\/p>\n The models above are just 2 variables, so maybe some of the relationships arise due to intercorrelations among the variables. To determine this, we can fit the regression model to predict fertility from all the variables together (except for the 2-dimensional reduction of RIASEC):<\/p>\n I’ve included both the standardized slopes and the raw slopes. The variables were recoded into 0-1 scale for comparability, thus the raw slope is going from least possible to most possible job of that type, even if no such job exists in the classification with enough data to be included. These values represent the hypothetical effects of varying a particular variable while holding the others constant. As such, when we see that the effect of conventional is -0.42 it means that holding the other job characteristics (including female%) constant, and hypothetically changing a job into the most secretary\/clerk like job from the least, fertility would decrease by 0.42 logRR, which is about 34% lower birth rate. It would appear in other words, that the most fertile women are working in female dominated jobs that don’t involve busywork\/record-keeping, and aren’t artistic. So basically like a nurse or a preschool teacher. Well, that’s what the model says anyway, but it is only 17% r\u00b2 (r = 0.41), so it isn’t that great a prediction model. Clearly, there are some other job features we are missing.<\/p>\n Another issue is that to boost the precision, I pooled the data from 2 waves of ACS. Doing so required merging the changing job codes which resulted in some occupations being dropped (non-matching or merged). The stability across waves looks like this:<\/p>\n Despite having millions of women in the survey, the estimates aren’t that reliable, about 0.60 if we consider it a retest. Since we know the standard errors, we can also surmise that the correlation would be about 0.74 if there was no measurement error. That is to say, a lot of these occupation categories were changing their relative fertility rates. This unreliability of the outcome also means that the model accuracy is severely underestimated but it is difficult to say how much exactly (adj. R2 should be around 31).<\/p>\n Above I had shown the most and least fertile female occupations, but better perhaps is to look at the largest occupations instead:<\/p>\n ~67% of American women work in one of these jobs, so from a sociological perspective, they are the most important. If occupations have any causal effects on fertility of the women in them, then to boost fertility, we should move women out of the jobs in the bottom into some of the higher fertility ones, preferably near the top. I’m happy to see that post-secondary teachers is the bottom one, which is university\/college teachers. Women (as a group) weren’t good for academia<\/a> (fittingly, written by a woman), and it may appear academia wasn’t good for them either.<\/p>\n Another idea is to group occupations into broader categories to get a clearer picture:<\/p>\n Most of the low fertility rates are found in typically male jobs, or those which were until recently, with exception of arts\/entertainment. Most occupations are below average because most women work in higher fertility jobs (cf. the top 50 list above) but there’s a long tail of uncommon occupations with low fertility rates.<\/p>\n Different kinds of teachers:<\/p>\n The 2 main jobs involving working with small children both have elevated fertility rates (pre- and elementary-middle school).<\/p>\n And just for the record, here are the top lists for each wave separated so without dropping occupations that could not be mapped between the waves. Also note that these results are problematic because of the regression towards the mean you would expect from focusing on the outliers (winner’s curse). Wave 2013-2017:<\/p>\n<\/a><\/p>\n\n
\n
\n
\n
\n
\n
\n
<\/a><\/p>\n<\/a><\/p>\n<\/a><\/p>\n<\/a><\/p>\n<\/a><\/p>\n<\/a><\/p>\n<\/a><\/p>\n