>I'd like to see that. Of course, the rebates
>are only one piece of the tax cut.
>DAILY FINANCIAL MARKET COMMENT 10/16/03 Goldman Sachs Economics
>
>*The latest retail sales report confirms outsized strength in
>third-quarter growth in real GDP. In fact, it opens up some
>upside risk to our 6.5% (annualized) estimate.
>
>*We continue to estimate a 4.0% growth rate for the fourth
>quarter on the view that tax cuts drove last quarter's surge
>in spending. In fact, rough estimates indicate that consumers
>have spent about three-quarters of the tax cuts quite quickly
>after (and possibly before) receiving them. With the impulse
>of tax cuts on disposable income growth now largely behind us,
>and with few reliable signs that labor market improvement is
>generating an offsetting stream of wage and salary income, a
>slowing does appear likely.
>
>*Moreover, a slowing of the dimensions we envision is hardly a
>rare event in the history of volatile GDP growth from one
>quarter to the next. In fact, it would follow a comparable
>speed up from the second to the third quarters, and it is
>indicated at the moment in the sharp slowing of retail sales
>in September.
>
>Tax-Driven Spending and GDP Growth: What Goes Up Can Come Down
>
>The latest retail sales report provides added confirmation, if
>any was needed, that the US economy soared in the third quarter.
>Although month-to-month sales changes in September were weaker
>than we expected after seeing the chain store sales reports,
>upward revisions to sales gains in July, and especially in
>August, more than offset the September disappointments. Accordingly,
>upside risk has emerged against our recently revised
>estimate that real GDP rose at a 6.5% annual rate in the third
>quarter. However, the risk is not large enough to warrant a
>change in this estimate.
>
>The big question now coming our way is how the US economy can
>possibly decelerate as fast as we are expecting -- to an annual
>rate of 4% for the fourth quarter. Some even ask how we can be
>so confident of this. While confidence is not exactly the first
>adjective that comes to mind regarding forecasts in today's fluid
>environment, we can be confident that such a forecast is well
>within the realm of possibility on both analytical and
>statistical grounds.
>
>The analytical argument starts with the observation that tax cuts
>drove the third-quarter surge in spending. It is almost as if
>consumers watched President Bush sign the tax cut into law on May
>28 and then went out immediately to start spending the money,
>even before it showed up in their paychecks at mid-year. In
>June, July, and August, retail sales of items other than vehicles
>rose 1.2% (month-to-month, not annualized). This was a stronger
>and more consistent pattern of growth than had occurred earlier.
>
>Roughly speaking, it appears that taxpayers spent slightly more
>than three-quarters of the tax cut in those three months. This
>is an extraordinarily high percentage for such a short period,
>especially given that some of the tax cut was paid out in
>rebates, which normally gives rise to smaller increases in
>spending. The estimate is based on the following calculation: In
>July and August, the Bureau of Analysis estimates that the tax
>cuts reduced personal tax and nontax payments by $21.4 billion
>cumulatively (not at an annual rate). If we assume that the
>surge in spending in June occurred in anticipation of this cut,
>then comparisons of the levels of personal saving in June, July,
>and August to the level in May should indicate how much of the
>tax cut was saved. This calculation yields a cumulative saving
>of only $5.0 billion, or 23% of the tax cut, suggesting that the
>remainder was spent.
>
>Of course, this calculation overstates the response to the extent
>that June's increase in spending was not entirely motivated by
>the tax cut. However, two other factors work the other way. The
>first is that data on total spending and saving do not yet
>reflect the latest retail sales data, which will shift the mix
>toward spending. Second, while spending appears to have dipped
>in September, this dip probably did not match the setback in
>disposable income (yet to be reported) following the end of the
>tax rebates. Thus, when the calculation is repeated following
>the release of data on personal income and spending in September,
>it is apt to confirm a much larger initial response by consumers
>to the tax cut than the 50-50 spend/save split we had assumed.
>
>If this analysis is right, then US consumers need to develop an
>alternative source of financing soon, as the tax cuts have
>already exerted their maximum effect on disposable income growth;
>otherwise, they will either have to slow spending or dip deeply
>into saving. The logical source would be wages and salaries,
>however the signs are not promising. Although payrolls ticked up
>in September, it remains to be seen whether this is the onset of
>a new trend or a fluke. (We assume the former, though we do not
>expect a vigorous rate of job creation.) Meanwhile, the up-tick
>appears to have generated little income growth, as hourly
>earnings slipped 0.1% and workweeks were flat on balance in the
>private nonfarm economy.
>
>In fact, the retail sales data suggest that spending did slow in
>September, providing the first of three statistical
>demonstrations that a 2.5-point slowing in real GDP growth is
>entirely possible. Retail sales fell 0.2% overall, and they rose
>only 0.3% excluding vehicles. If we focus on the component of
>this report used in estimating total spending, which also
>excludes building materials, the increase in September was only
>0.1% following gains averaging 0.8% per month from June through
>August. As a result, the level of such spending in September
>points to only a 5.8% annualized increase for the fourth quarter,
>as compared to 11.4% in the third. Admittedly, this calculation
>is tenuous, but it is not biased one way or another. If the main
>part of consumer spending on goods can slow this sharply, why
>can't GDP slow in a similarly abrupt way?
>
>The second point is that it not only can, but often does. GDP
>growth rates are quite volatile from one quarter to the next. For
>example, since 1947 the standard deviation in quarter-to-
>quarter changes in (annualized) real GDP growth has been 4.8
>percentage points. Against this standard, a 2.5-point slowing is
>only half a standard deviation off the norm, and the probability
>of slowdown this big or bigger is a nontrivial 30% (with another
>30% reserved for speedups of 2.5 points or more). If we limit
>the period to the last twenty years, during which time the Brave
>New Business Cycle has resulted in more stable growth rates, the
>standard deviation drops, but only to 2.6 points. This still
>leaves about a 16% probability for the expected slowdown -- with
>no reliance on the analytical reasons advanced above.
>
>The final point is that GDP has just done in reverse what many
>are now questioning for the remainder of the year. By our
>reckoning, real GDP growth speeded up from a reported 3.3% annual
>rate in the second quarter to a 6.5% rate in the third quarter.
>Although this latter figure is just our estimate, many other
>economists have estimates in the 5.5%-to-7.0% range. If GDP
>growth can pick up this quickly, what's to say that it can't slow
>down just as quickly? In short, what goes up can come down.
>
>Ed McKelvey