New Home Sales fell sharply in December 2016, from a seasonally-adjusted annual rate of 598k in November to 536k. That wasn’t unexpected given the behavior of interest rates since August in particular. It might suggest further declines in new sales as well as construction of new homes in the months ahead.

In the bigger picture, interest rates just should not have this much effect on any part of the home market. With an unemployment rate at or below 5% for each of the past fifteen months, that would normally indicate conditions where a relatively minor shift in the mortgage rate is a secondary or tertiary factor for real estate. These not being anything like normal times, the pace of sales, construction or resales are thus a more intense reflection of conditions in the real economy as quite different than the unemployment rate.

In terms of resales, as noted earlier this week, this disparity is captured in terms of for-sale inventory, meaning the growing lack of it. If demand for homes in general were constant or increasing, then without sufficient volume in resales there should be enormous upward pressure in new home construction to fill in for that missing volume.

Instead, new home sales appear to be growing steadily if irregularly. Even in the historical context with data reaching back to 1963, new home sales have moved up into only past the bottom of the historical range (SAAR).

But this is another instance where the positive numbers of what looks like a recovery are deceiving. Like the headline labor market statistics, the upward track since the bottom takes no account of the size and scale of the decline which preceded it. In cyclical terms, that wouldn’t matter as recovery would be taken literal. Instead, if we rescale the housing market and the new homes component by population, we see that recovery has been quite minimal.

In population terms, the housing bubble wasn’t so far overdone, but the housing recovery has been dramatically underwhelming. That would suggest factors other than housing stock and demographics.

Apart from mortgage availability (meaning either lending standards or balance sheet capacity of mortgage providers), the primary factor in determining the demand for real estate is employment. That includes the pace of income expansion on an individual level as well as overall (national income). From this view, the seeming disparity in new home sales after the crash isn’t a disparity at all, it is in keeping with the same shrunken proportions as the overall labor market.

In other words, the housing market has only recovered for where the economy has, leaving an enormous and unprecedented proportion of the economy out of most interpretations. When the NAR, for example, opines as to why the housing market isn’t growing faster or farther, or why there isn’t more volume being offered given the unemployment rate, it is doing so as if the current economy was normal, and not negatively unique as a durably, even permanently, altered economy. With far less potential labor participating in the recovery, it is very much consistent that far less housing would constitute the same.

Like resales, however, there are signs that beyond the shriveled proportionality the housing market has cooled. The median sales price of new homes hit a record high in December 2016, but it did so at a decelerating pace. Prices have been more volatile throughout 2016, and the average annual gain for the median sales price is down to just 2%; the lowest since early 2012 and the start of this partial rebound.

That average is down sharply from early 2015, when prices were on average rising by about 9%. Taken together with the lower for-sale inventory of existing homes, which under normal economic conditions would contribute to more rapidly rising new home prices and/or sales/construction, the housing market suggests overall a great deal more caution the past two years. That would explain the potential outsized reaction to mortgage rates (with mortgage applications down sharply as if late 2013 all over again).

Fewer Americans working means lower proportional demand for houses; thus, prices can rise and precipitously while volume does not. This is not a condition that most people are unaware, unlike economists who have convinced themselves none of this matters (Baby Boomer retirement, or something). The housing market, like the overall economy, looks somewhat healthy with its positive numbers but only in isolation relative to those positive numbers.