I am not sure that the pushing matters that much--it is entirely possible that people, for reasons beyond financial reasons, want to own houses in particular and real estate in general. Two things stick with me:
(1) I was talking yesterday with a developer here in India who is trying to building market-rate affordable housing. He faces a number of hurdles, one of which, he said, is "Indians' obsession with homeownership."
(2) Years ago, when I was in Madison, the guy who cut my hair loved to talk about the rental property he owned in Florida. It would, he said, be the source of his retirement income. I asked him why he was so undiversified--why he didn't sell his place and put the money in an index fund. His reply was that he didn't trust Wall Street, and that he needed an investment that he could "touch" as well as control. I told him his mistrust of Wall Street was misplaced--shows what I knew at the time.
The point is that there is something about real estate that reduces agency problems, One may not be able to control markets, but one can control the management of real estate that one owns. I do remember when I left the rental market for the owner market, I was very happy, not because I thought I would make out financially (house prices in Madison had been stagnant for years), but because I disliked my landlord, and was relieved that I would no longer have to write a check to him.
This is not to say putting people in houses they cannot afford is a good idea, and I have long been dubious of very low downpayment schemes (I do think homeowners who put no equity into their houses are not really owners). But it is a little too easy for people who own houses (or have the choice to do so) to say it is not important to make the option available to others. Freedom to some extent means the ability to take control of one's own life, and to avoid agency issues as much as possible.
I freely confess that this is all supposition based on informal observation. Some work has been done on how ownership solves some agency issues, but I think it is an understudied phenomenon. If anyone wants to help me think about how to model such things formally, I would welcome the assistance.