For the chart show, Sandy Struckmeyer, Steve Kamin, and I will be referring to the packet of materials that was distributed to you. While putting together the Greenbook, two of the questions we asked ourselves were, first, “Where is the economy now?” and, second, “What are the forces that should sustain above- potential growth over the next two years?”
With regard to what is sometimes called “forecasting the present,” your first exhibit shows a variety of indicators that have informed our judgments. The top left panel plots average monthly changes in private payrolls. Like most observers, we were disappointed by the payroll employment report for December. However, that reading came on the heels of more-favorable reports for the preceding two months. Averaging through the monthly ups and downs, as I’ve done in the bars on your chart, you can see that, after a protracted period of job losses, employment has been on an uptrend since midyear, albeit a weaker one than is typical for this stage of the business cycle.
In contrast to the payroll numbers, most indicators of final demand have been quite favorable. Sales of autos and light trucks (the top right panel) and other consumer goods (the middle left panel) point to continued strength of household demand, while the orders and shipments figures (the middle right panel), which here are plotted only through November, tell a similar story for business demand. And as I mentioned earlier, this morning’s orders and shipments release provided further confirmation of an ongoing pickup in business spending.
Thus, in putting together our near-term GDP forecast, we once again were faced with a tension between lackluster payroll employment numbers and pretty stellar incoming data for final sales and production. And, once again, we’ve resolved that tension by writing down sizable increases in productivity (the bottom left panel). Sandy will be discussing productivity in greater detail later. The table at the bottom right adds it all up. We now estimate that real GDP grew at a 4.8 percent annual rate in the fourth quarter of 2003 and will expand at a 5 percent pace in the current quarter. The bulk of the increase in GDP over this two-quarter period comes from robust gains in final sales; but with stock–sales ratios so low, inventory investment is expected to make a noticeable contribution as well. Let me note that this exhibit reflects the figures that were in the Greenbook—before we got the data on shipments and orders this morning. So perhaps we might take a tenth or two off Q4 and add a tenth or two to Q1.
Turning to the outlook for the next two years, the subject of exhibit 2, the question is, “Will the current rapid growth be sustained and why?” As you can see from the top panel, we think that economic growth will be sustained at rates faster than potential through the end of 2005. The remaining panels highlight some of the key contributing forces. Fiscal policy has been quite expansionary in recent years, and as can be seen in the middle left panel, we expect it to provide additional impetus to growth through the end of this year. However, on our assumptions, fiscal policy swings to slight restraint in 2005 as the partial-expensing provision expires and the growth of defense spending slows. Monetary policy also has been quite supportive of growth over the past few years. And with the funds rates projected to remain relatively low over the forecast period, it will continue to be so. Another factor with favorable implications for growth is the ongoing strength of productivity gains, which should influence household perceptions of permanent income and business perceptions of profitability. Given the surge of earnings lately, firms have been able to finance their desired spending while keeping a tight lid on their borrowing. Indeed, as shown in the bottom left panel, debt ratios have fallen sharply in recent years and are now either at or near the lowest levels in the past two decades. With the marked improvement in business financial conditions, risk spreads—as Dino noted yesterday—have dropped dramatically on both investment-grade and junk bonds. Returning to the top right panel, the stock market is assumed to rise at a 6¼ percent annual rate over the projection period—maintaining risk-adjusted parity with the projected yield on long-term Treasury securities. In our forecast, we see the higher stock market as providing some support to household spending after several years in which it was a considerable drag.
Your next exhibit focuses in greater detail on the household sector. Personal income (the dark bars in the top left panel) is expected to rise smartly over the next two years, supported in part, as I mentioned earlier, by the strength of productivity. The strong income growth provides a considerable lift to consumption outlays (the yellow bars). In addition, last year’s tax cut continues to boost spending in 2004. The top right panel shows the implications of the PCE and income projections for the saving rate. As indicated by the thick line, the BEA currently estimates that the saving rate was only 2.3 percent in the third quarter of last year, and we think it fell another ½ percentage point in the fourth quarter. In our forecast, the saving rate climbs about a percentage point over the projection period.
One risk to the household sector forecast is that the saving rate might rise more than we have allowed for in the baseline. The bullets in the middle panel present our thinking on this issue. The first thing to note is that the BEA, in effect, shifted the goal posts in its most recent comprehensive revision, by taking the level of the saving rate down on average over the past decade about a percentage point. This suggested to us that we ought to revise down our thinking about what constitutes a “normal” or “target” saving rate by a like amount. Considerable econometric work suggested the same. That said, the current saving rate still is well below the level that many observers often think of as a more normal rate. It’s important to keep in mind, however, that the target saving rate is a moving target. It varies over time as the fundamental determinants, such as the wealth–income ratio, the composition of income, and real interest rates, change. Indeed, we read the current settings of those fundamentals as pointing to a target saving rate for the next year or two in the neighborhood of 3 percent. In our projection, the saving rate rises to 2.8 percent by the fourth quarter of 2005, eliminating the bulk—but not all—of the gap between actual and target saving. We’re not sure what accounts for the gap that we’re seeing currently, but departures from our saving models of this magnitude are not unusual. Historically, disequilibriums like this have been worked off fairly gradually, but a more sudden and complete adjustment, and a consequent drag on aggregate demand, certainly is a risk.
The forecast for housing starts is presented in the bottom left panel. The pace of homebuilding, fueled by low mortgage rates and robust income growth, is anticipated to remain strong over the projection period, edging down only a bit from the current elevated pace as mortgage rates drift up a bit. The bottom right panel shows our projection of house prices. In our forecast, house prices slow from the 7 to 8 percent gains posted in recent years, to about 2½ percent during 2005. The downtilt primarily reflects the current disparity between rapid house-price inflation and much smaller increases in the data for rents. The econometric evidence suggests that such a disparity typically is closed by house-price increases moving into alignment with rents. Of course, there is a wide band of uncertainty around this forecast, which imparts a wide band of uncertainty—both upside and downside—to movements in household net worth over the next two years, with consequent implications for household spending.
Turning to the business sector, exhibit 4, we expect real outlays for equipment and software—both the high-tech component and the “other” component—to rise rapidly this year, spurred in part by the temporary tax incentive. Indeed, the partial- expensing provision adds about 2½ percentage points to overall E&S growth this year and subtracts close to 6 percentage points next year. The basic contour of the forecast is consistent with the information given recently to the Reserve Banks by businesses in their Districts. As shown in the middle left panel, slightly more than half of the respondents plan to boost their capital spending in 2004—the responses are appreciably more upbeat than those in June. Although the usual accelerator effects were the primary reason given for the improvement, a sizable fraction of respondents also pointed to replacement needs as an important consideration. Moreover, partial expensing finally appears to be on the radar screens of at least some firms. A quarter of the respondents also cited improved cash flow or balance sheet positions, reflecting no doubt strong profits and, presumably, the expected profitability of new investments. As shown to the right, in our forecast the profit share continues to rise through mid-2004 before drifting down as labor costs accelerate and competition puts pressure on margins. Nonetheless, even through the end of next year, profits are still elevated by historical standards.
An important element of our capital spending forecast is strong demand for computers. In part, our forecast for computer spending has been informed by the projections of industry analysts for unit sales of PCs—both so-called desktop boxes (which typically sit on the floor) and laptops (which often sit on desks). [Laughter] The bottom left panel shows one such forecast from a leading consulting group. As you can see, unit sales are expected to rise sharply throughout the projection period, driven by the need to replace machines that have become physically or technologically obsolete. After adding in the effects of quality change, these unit sales translate into real spending increases on the order of 40 percent annually over the next two years—roughly in line with the increases posted during the second half of the 1990s. One concern that has been raised in many circles is the implication for the domestic economy of outsourcing computer production—that is, whether today’s boxes contain considerably less domestic content than the boxes sold a few years ago. The bottom right panel addresses this question. It uses cost shares to measure the domestic content of PCs sold in the United States. Most of the domestic value-added in a PC is for the micro-processing unit—the Pentium or similar chip that lies at the heart of our PCs. This is shown by the black shaded portion of the bars; smaller contributions come from the production of specialized chips that provide basic logic or act like traffic managers inside your PC and, to a lesser extent, from integration and final shipment. The bulk of the foreign value-added is for such things as printed circuit boards, storage devices, and peripherals. The point is that the United States is still the world leader in designing leading-edge chips (a very high value-added activity) and fabricating them (also a high value-added activity). As a result, the domestic content of PCs has changed relatively little over the past few years. Accordingly, the sharp rise in PC sales that we are forecasting should translate into domestic production at about the same rate as in recent years. Steve will now turn the discussion from domestic content to foreign content.