Worse Than It Looks? GDP Versus GDI

With the health care vote pending and dominating the discussion I thought I would spend a little time examining why the economy feels a lot worse than the GDP numbers indicate. Anecdotal evidence and the official statistics seem to be at odds. Then there’s the small matter of an unemployment level apparently immoveable by growth.

Buried in the official reports from the Bureau of Economic Analysis, the economy’s scorekeeper, is a lesser known figure called Gross Domestic Income. This will no doubt sound a bit technical, but the two measures – GDP and GDI – are supposed to be exactly the same. Think of them as accounting terms found on different sides of the ledger, with GDI as the credit and GDP as the debit. As is all accounting the two should balance. In reality, since the source data is very different, there is often a ‘fudge’ factor inserted to force this balance. This fudge factor turns up in the monthly GDP reports at the bottom of the page under the heading Addendum as the mysterious ‘statistical discrepancy’.

GDI consists, as its name implies, of all the various sources of income within the US during any particular year. So it includes wages, salaries, various supplements, an adjustment for taxes on imports [since that is not income], and corporate profits adjusted for depreciation. Anyone familiar with a flow of funds statement will recognize those categories and the need for the adjustments.

To put this in context: for the fourth quarter of 2008 GDI was $14,207.9 billion, down from $14,478.4 billion in the third quarter of 2008. That’s an annualized drop of about 7.5%

But GDP dropped by only 5.4% in that period, and the ‘statistical discrepancy’ needed to account for the difference between the two measures shot up to $139.4 billion.

So while the two numbers are supposed to move in lock step, there are occasions when they slip quite a way out of balance and tell different stories.

Without drifting deeper into the data, I think this discrepancy helps explain why the economy feels very different from the story being told by the GDP data: incomes are moving in a more exaggerated cycle, with stronger shifts up and down. Since incomes is what consumers ‘feel’ most immediately GDI is a better measure of the movement that most people have contact with. It therefore probably reflects the source of consumer sentiment and general mood more accurately.

Even this is only part of the explanation: within GDI two different trends are clearly visible. Since GDI consists of all incomes we can tell the trajectory of profits apart from that of wages. When we get down to this level of analysis I think we reveal an even more accurate gauge to assess the economy’s health, especially with respect to employment.

The crucial point is this: during the latter half of 2009, when GDP rose and the recession was announced as being over, GDI stabilized and flattened out. It did not rise for three straight quarters – unfortunately we do not have the final data for the fourth quarter yet. So the story written by GDP and by GDI diverged strongly as 2009 progressed. One told of recovery, while the other spoke to stagnation.

Worse: within GDI there was a dichotomy. Wages and salaries, which account for over half the total, continued to decline all year long, from a first quarter of $7,833.0 billion, to a third quarter of $7,758.5 billion. The total GDI was level throughout the year only because this drop in wages and salaries was offset by a recovery in corporate profits.

Now, I think, the source of the continued sour mood of consumers is revealed. Incomes have yet to recover.

Further the impact of unemployment is fully revealed. Obviously one major source of the decline in wages and salaries is that there are far fewer people employed and therefore fewer wage earners. Even those who are still employed are facing cutbacks in hours and other forms of lost income. Equally obviously, as sales started to pick up, this drop in wage and salary costs immediately bolstered profits. To the extent that profits grew instead of wages purchasing power was transferred from workers to shareholders and corporations. In other words the benefits of the implied increase in productivity was retained within the corporate world and not passed on. This is not unusual during the beginning of a recovery. What makes this cycle different is the massive level of un- and under- employment which will mean that corporations keep the upper hand in wage negotiations for much longer than normal. So we should not be surprised to see the wage and salary component of GDI track down or sideways for a quarter or two more. As it does the sour mood will continue and will be reflected in a continuation of the low levels of consumption and higher levels of debt reduction we see elsewhere in the data.

The divergence between GDP and GDI has spawned renewed attention on the question of which is a more accurate measure of the economy’s health. The Fed has long looked at both, and there have been a few articles and academic papers written seeking to clarify which is the better indicator.

For our purposes the GDI figures reveal clearly that the crisis was both deeper and more prolonged than that described by GDP. The narrative told by GDI is that the economy was worse than it looked, and that the recovery has favored business not regular people. It is a narrative more in accord with anecdotal evidence. And it helps explain why consumer confidence surveys are still turning up a deep pessimism about the economy.

The people being surveyed, the people on Main Street, are not talking about growth, because there isn’t any at street level.

It’s worse than it looks.