I am going to react to three things: Nick Rowe talks about the fact that housing transaction volume is higher when prices are rising; Scott Sumner’s latest attempted swindle; and the paper by Steven Gjerstad and Vernon L. Smith in What Caused the Financial Crisis. This book, which is a based on this issue of Critical Review, is my choice for the best book of analysis of the financial crisis–by a wide margin.

One point that Gjerstad and Smith make is that bubbles can appear in asset markets but not in ordinary markets for goods and services. That is because with an asset your willingness to buy depends on your expectation for price changes going forward.

Scott tries to twist the labor market data to raise doubts that labor demand was weak prior to 2008. I am sorry, but I am using the same labor market indicators that I used years before the crisis, and I think that the decline in labor demand prior to 2008 is pretty evident. See this post.

Back to the main issue, which is the behavior of house prices. Why do price appreciation, construction, and transaction volume move together? For price appreciation and construction, I would give a Tobin’s q answer. That is, when the price of housing rises above replacement cost, you get a lot of new construction. When the price of housing falls below replacement cost, not so much.

For transaction volume, the issue is more difficult. If we take it the housing is an asset, then nominal price stickiness, which is one of the explanations Rowe considers, seems odd. I would not rule it out, but I would look for other explanations first. Here is one possibility:

A lot of transactions volume in the housing market is the so-called “trade-up” market. Some households, when they find that their house has appreciated, decide that they would like to use part of the profits to buy a bigger house.

So then the question is, why is there not a trade-down market, meaning that when household are in houses that have depreciated, they want to sell and buy a smaller house? Iin either case, there will be threshold effects. Because of transaction costs, you need to have accumulated a big change in wealth, income, or household size in order to want to move up, and conversely.

I think that if incomes are rising secularly, the trade-down market will be limited. If you lose some wealth because your house depreciates, you don’t trade down, because you have higher income. (Keep in mind that back in the twentieth century, people who bought houses made down payments, so if your house declined in value by 15 percent it was still worth more than the mortgage.)

With secularly rising incomes, a little bit of house price appreciation is enough to get a good trade-up market going. But an equivalent decline in house prices would not create as much trade-down activity.

What I want to suggest is that, because transaction costs are high, it takes a large cumulative price change to make people consider trading up or down. For example, imagine that a homeowner will trade up if the house has gone up 20 percent and will trade down if it has gone down 20 percent. Suppose that the general trend of house prices is a gradual increase. So, after five years, the typical home owner might have a house that is worth 10 percent more–not enough to want to trade up. From this point forward, if the price goes up 10 percent, that homeowner will trade up. But if the price goes down 10 percent, that will put the owner back where he or she started, so there is no reason to trade down, unless prices fall another 20 percent.

I think the story could be stronger if we introduce a secular increase in real income and a positive correlation between income and demand to trade up.