Housing Wealth Effect: Robert Samuelson Never Heard of Inflation

Monday, 17 June 2013 04:01

That's what readers would probably conclude from a column headlined "Americans have record wealth but aren't spending it." The first paragraph begins:

"In the economic history of our time, June 6, 2013, ought to occupy a special place. That’s the day the Federal Reserve disclosed that the net worth of American households — the value of what they own minus what they owe — hit $70 trillion, a record that exceeded the previous peak before the 2007-09 financial crisis. Higher stock prices and a long-awaited housing recovery are slowly restoring Americans’ lost wealth. By all rights, this symbolic crossing ought to improve confidence, prompt consumers to spend more freely and increase the economy’s growth."

Okay, let's first check the story that consumers are not spending freely. If we turn to the most recent data we see that the saving rate for the first quarter was 2.3 percent. That is slightly higher than the 1.5 percent saving rate we saw at the peak of the bubble in 2005 and 2006, but it's not hugely different. (Arguably, because of the statistical discrepancy in the national accounts we should view the saving rate as being somewhat lower at the bubble peak.)

It is possible that the first quarter saving rate was somewhat lower than normal because households were taking time to adjust their consumption to the ending of the payroll tax cut. Also, they may have been spending part of the big dividend payouts that were made in the 4th quarter to beat the rise in tax rates. If we average in the 5.3 percent saving rate from the 4th quarter, we get an average of 3.8 percent for the last two quarters, 2.3 percentage points above the saving rate at the peak of the bubble. So the question is why consumption is not back to its bubble peaks if wealth is back to its bubble peak.

Samuelson tells us:

"Here’s where the process seems to have broken down. Before the financial crisis, says economist Mark Zandi of Moody’s Analytics, an added dollar of housing wealth might produce 8 cents in extra spending, and an extra dollar of stock wealth, 3 cents. The overall effect was about 5 cents per dollar of new wealth, Zandi says. Now, 2 or 2.5 cents 'seems more likely to me.'" 

Okay, let's check that one.

In the most recent data the Fed reported (Table B.100, Line 49) that households had $9,075 billion in housing equity. At the end of 2012 they had $20,609 billion in stock wealth (B.100.e, Line 6). By comparison, the Fed reports that at the peak of the housing bubble in the third quarter of 2006, housing equity was $10,979 billion. At the end of 2006 the Fed put the value of stock was $20,397 billion.

While these numbers may not seem hugely different, they are not adjusted for either inflation or the size of the economy. To make use of Zandi's wealth effect numbers we have to convert these wealth figures to the same year's prices. Since it is standard to use 2005 year prices we will set that as the base. In 2005 prices we get:


    (2005 dollars)
Year    Stock Housing
2006   $19,633 $10,568
2012   $17,530 $7,719


Okay, so in 2005 dollars stock equity was still down by more than $2 trillion at the end of 2012. (Add 10 percent to the 2012 number if you want to get closer to the first quarter figure, although no one would say the full wealth effect is felt immediately.) Zandi's 3 cents per dollar estimate of the stock wealth effect would imply a drop in consumption of $63 billion this year compared with 2006.

Adjusted for inflation, housing equity is still down by $2,849 billion from the bubble peak. (The bubble is not back, yet.) Zandi's 8 percent housing wealth effect number would imply a drop in consumption of $228 billion from lower real house prices. Adding this to the weaker stock wealth effect, Zandi's estimates imply that consumption should be $291 billion lower in 2013 than it was in 2006 as a result of a diminished wealth effect.

Turning back to the income and consumption data, disposable personal income in the first quarter of 2013 was $9,574 billion in 2005 dollars. Zandi's wealth effect numbers would translate into a drop in consumption of just over 3.0 percentage points of disposable income. In other words, Zandi's wealth effect numbers imply that we should be seeing a saving rate of around 4.5 percent currently, somewhat higher than the 3.8 percent average than we have seen over the last two quarters.

In other words, if we take Zandi's numbers seriously and we do our arithmetic right, we should be asking why consumers are spending so much, not why they are spending so little. There is not much room for the pessimism explanation that is at the center of Samuelson's piece.

So remember boys and girls, always adjust your data for inflation, then you won't end up writing silly columns in the Washington Post.