I did not argue that this situation is "bad," I argued that it's "hostile to people who want to start families."
Now, you can guess based on my emotional language that I think it is good to pursue policies that make careers more open to people who want families. And, as it happens, the US government agrees, in so far as it designates wanting a family a protected class for the purpose of employment discrimination.
My argument, devoid of moral content, is:
If it's your aim to avoid excluding people who want families from a career, you need to regulate the amount of hours people in that career are expected to work, the amount they're expected to travel, and the distribution of lifetime expected earnings (don't unnecessarily gate a prosperous 40s behind penurious child-bearing years). Since this particular career is overwhelmingly government-funded, free-market analyses miss the point - the government decides how many Physics Grad Students there are, what they're paid, and largely what their future career trajectories are. Given that it holds the purse strings it also holds substantially more power to regulate working conditions than it otherwise would.
It's empirically obvious that people who want families are excluded from science, and I'm positing that the above is why.
I also want the society to support more families, especially by the highly educated people (not in the name of some IQ-supremacy, although you could argue that's beneficial to society, but simply to level the playing field - low educated people don't seem to have any problem having tons of kids, while high educated are delaying family creation more and more - and in case of women, often into their less fertile years).
However, my proposed solutions are quite different, and much more direct - simply address the main issues directly - the main issue IMO currently being housing (at least in major cities, near major colleges, in major Western countries). So you could make a simple rule, giving (either literally, or in a form of long-term (10+ years) loan that automatically gets written of under some conditions) young people apartments when they e.g. enroll in a difficult STEM PhD program. Once this problem is "sorted" (housing represents a major factor in how "secure" you feel about your future - you and your kids never going homeless) I think that would encourage many young highly-educated high-achiever couples to start families much sooner.
Housing is not the main issue (grad students can afford that, after all), child care is.
For several years an adult needs to be present 24/7 and if the adult is someone other than the parent that ain't cheap. If the adult is someone other than the parent and it's not business hours that really ain't cheap.
What's a better use of the NSF's dollars? Funding night-nurses for grad students so professors can keep working them through the night? Or just funding day-care and regulating their hours?
You definitely don't want to just give grad students assets that can be converted into liquidity after some fixed term, because then you're just exacerbating the "you get rich eventually, after you're too old for kids" issue. A better asset would be something like a seven-year voucher for day-care that is:
1. Granted after a two-year vesting period (so that people receiving this benefit get at least a Master's degree)
2. Non-transferrable (so students don't just sell it)
3. Not conditional on continued enrollment (so it doesn't tie you to an abusive professor / department)
Again, I think this only really works with regulations on hours (possibly with a shift system of some kind), but once the federal government begins funding such a benefit there's no reason to restrict it to grad students (I'm sure military service-people could use such a benefit as well, and the more government professions receive it the more flexible the voucher is).
Now, you can guess based on my emotional language that I think it is good to pursue policies that make careers more open to people who want families. And, as it happens, the US government agrees, in so far as it designates wanting a family a protected class for the purpose of employment discrimination.
My argument, devoid of moral content, is:
If it's your aim to avoid excluding people who want families from a career, you need to regulate the amount of hours people in that career are expected to work, the amount they're expected to travel, and the distribution of lifetime expected earnings (don't unnecessarily gate a prosperous 40s behind penurious child-bearing years). Since this particular career is overwhelmingly government-funded, free-market analyses miss the point - the government decides how many Physics Grad Students there are, what they're paid, and largely what their future career trajectories are. Given that it holds the purse strings it also holds substantially more power to regulate working conditions than it otherwise would.
It's empirically obvious that people who want families are excluded from science, and I'm positing that the above is why.