DrQian
I think the premise is worth questioning though: It's true that many people have most of their wealth tied up in their house. But unless they want to substantially downsize, they can't access these savings.
In general it seems bad that it's common for people to have most of their wealth in an illiquid, undiversified investment that they also live in.
1) Green belt policies prevent the expansion of cities outwards. Compare the footprint of London to other major European cities and you'll see it's barely moved at all for decades.
2) Discretionary planning policies that mean local councils can turn down development, even when it adheres to stated rules. See for example this brownfield development in Brighton, a city that has one of the worst housing delivery records in the country: https://www.bbc.co.uk/news/articles/cwygkz57k1vo
The professional planning board recommended it for approval, but it was turned down by the political committee because of local complaints. A well specified zoning system would avoid this. Reduced uncertainty would make it possible for smaller builders to enter the market
3) affordable housing mandates (known as Inclusionary Zoning elsewhere). These specify that some amount of the new building must be offered at below market rates. Although there's a strong moral case for capturing some of the value produced by giving planning permission, this effectively acts as a tax on building homes in places with shortages. Ideally, people would be most incentivised to build where prices are highest, but this policy removes that incentive.
I don't have sources to hand, but a quick google suggests the following:
Although people subjectively rate their performances better than an objective observer would (see, e.g. http://en.wikipedia.org/wiki/Overconfidence_effect), being overconfident can actually lead you to increase your perfomance, at least according to http://econpapers.repec.org/paper/lmumuenec/11885.htm
"Abstract: This paper suggests a potential rationale for the recent empirical finding that overconfident agents tend to self-select into more competitive environments (e.g. Dohmen and Falk, forthcoming). In particular, it shows that moderate overconfidence in a contest can improve the agent's performance relative to an unbiased opponent and can even lead to an advantage for the overconfident agent in absolute terms."
So that isn't overwhelming evidence, but it does suggest that what you wrote might not be correct.