United States Forecast Summary
Since mid-summer new data (and revisions to old data) have forced us to reformulate our scenario for the U.S. economy. Our new plot line is less bright, both for the recent past and for the near-term future.
Combined with data revisions released in July, the initial National Income and Product Account estimates for the third quarter have caused us to revise our relatively optimistic view of the recent trajectory of the economy. That view had two parts: (1) In mid-2013 the economy had (finally) broken out of the “new normal” pattern. Growth had accelerated for a subpar 2% to close to 3%. This acceleration was relatively balanced, except for weakness in housing. (2) Prospects for further improvement were good. Housing had room for significantly better performance. Lower energy prices could lead to somewhat faster growth in both consumption and investment, and also a reduction in the trade deficit.
Data for 2015 including the new data for the third quarter are inconsistent with this optimistic scenario. Third quarter growth in real GDP came in at only 1.5%. A reduction in inventory accumulation cut third quarter growth in half, but this was mostly a correction of excessive accumulation in Q1 and Q2. The average for the past year is a “new normalish” 2%.
Leaving aside the inventory component third quarter final sales grew at a 3% rate, which exceeded our August forecast of 2.4%. Most of the error was due to stronger spending by consumers than we expected. Business spending on equipment was also above our forecast, but total business investment (that is, including spending on intellectual property and on structures) fell short. Housing was below our expectation, as were both sides of the trade balance. Government purchases grew by 1.7%, more than double our forecast, due to a second consecutive quarter of strong state and local expenditures.
Taking a longer perspective brings our earlier excessive optimism into clearer focus. A year ago our forecast for the past year was for final sales growth of 2.9% compared with an actual result of 2.2%. This came in spite of consumer spending and residential investment that were well above our forecast. These were offset by weakness in business spending on equipment and structures, and by a very large deterioration in the net export balance.
Our hopeful scenario was derailed by two main factors:
- We were overly sanguine about the impacts of lower energy prices. To begin with, the price decline has far exceeded what we expected a year ago, and this has produced a much more drastic cutback in spending by domestic oil producers than we foresaw. This largely explains the shortfall in investment spending. In the past quarter domestic oil production (XOIL in our model) has begun to decline.
A number of factors combined to cause a large shift in international trade. First, several unforeseen events have caused a large appreciation in the exchange value of the dollar – by 15% over the past year for the index in our model (RFXBRD). This is almost six times more than our year-ago forecast. To begin with, there has been a divergence in relative monetary policies. For the U.S. there has been a continuous expectation that the Federal Reserve would (finally) begin to raise interest rates. Meanwhile, abroad the opposite has been true. Both the European Central Bank and the Bank of Japan have engaged in quantitative easing, while the Chinese have lowered interest rates and the exchange rate of the yuan. The Chinese also engaged in a misguided effort to first encourage investment in their stock market and then to cool down the resulting “irrational exuberance.” The result was a mini financial panic that induced a flight to (dollar) safety. Second, the market for our exports has weakened independent of the exchange rate effect. Chinese real growth has continued to slow, with spillover effects on their trade partners, especially on commodity producing countries. Europe and Japan are also struggling. Finally, the reduction in domestic oil production requires more imports to make up the difference.
- Overall, the NIPA data suggest that the recent past is fully consistent with the pattern of the past six years – slow overall growth, with a few positive items (solid spending by consumers and the housing sector) and a few that have been disappointing (business investment and international trade).
Recent Monthly Data
As is usually the case, recent monthly data have been a mix of positives and not so positives. Friday’s release of labor market data for October leads the former list. The 271 thousand increase in employment far exceeded expectations. The increases were concentrated in services, with large increase in retail trade, professional and business services, and health care. There was also a solid increase in construction. In the household survey the headline unemployment rate decreased from 5.1% to 5.0%. [Calculated to additional decimal places the decline was actually insignificant – just 0.014%.] The labor force increased by 313 thousand, nearly reversing its decrease in September.
Consistent with the NIPA data, both disposable income and consumption were up in September. Auto sales last month were again above an 18 million annual rate. September and October are the best two month period on record.
Other indicators are less encouraging. Consumer sentiment declined in October. Housing starts have been essentially flat over the past six months, although up substantially from a year ago. Industrial sector indicators show consistent weakness. The total industrial production index has fallen in eight of the nine months this year. This includes activity in the mining industry, which includes oil production. The IP index for manufacturing is, however, not much better with only three of nine months showing gains. Its year-to-date gain is just 0.3%.
The ISM indices are fully consistent with the previous two paragraphs. Their manufacturing index was down for the fourth straight month in October, while their non-manufacturing index rebounded. For manufacturing the level of the index is balanced between expansion and contraction. The non-manufacturing measure, on the other hand, indicates strong expansion.
Except for the current quarter our new forecast is less optimistic for the next three years than our outlook three months ago. For 2016 and 2017 growth in output averages 2.4%, down 0.3% from our August forecast. For the current quarter our new estimate is 3.0% compared with 2.4% in August. However, nearly all of the difference comes from inventory change. In our August forecast we had a decrease in inventory accumulation that reduced output by one-quarter percent. In the current forecast the situation is reversed. Essentially, in August we thought there would be an inventory correction spread across both the third and fourth quarters. Now we think that it was mostly accomplished in the third quarter.
The slower output growth translates into less increase in employment. In 2016, for example, our current forecast has payroll employment rising by an average of 177 thousand per month. That is 35 thousand below our August forecast – a difference of over 400 thousand jobs for the year. Beyond slower output growth, part of the employment change is due to a higher estimate of the unemployment rate that represents full employment in our model (the variable THHUR). In August we had this set at 4.6%. We now have it at 4.8%. With actual unemployment now at 5.0% and falling, we see full employment being reached by the end of next year. Once this happens, employment growth must more or less parallel growth in the labor force, which means monthly increases of 150 thousand or even less.
A couple of additional aspects of our forecast scenario warrant mention. First, we have (again) adjusted our assumed path for crude oil prices. Reflecting recent prices for West Texas Intermediate the new path is a little lower for this quarter and for 2016:1, but then shows a more rapid increase. By the end of 2018 the new trajectory is close to $61 per barrel, compared to just $53 in August. We think this is roughly consistent with the marginal production cost of U.S. shale production.
Second, after the October employment report we now expect the Fed to begin increasing its target for the federal funds rate at its December policy meeting. Our path for rates after that has rates increasing less than one percent per year. Out confidence about this, however, is close to zero.
Our forecast definitely has a glass half full /glass half empty property.
On the optimistic side, it shows continuing output growth, continuing increase in employment, and (at least through the next year) a continuing decline in unemployment. Indeed, output growth over the next year matches that for the past two years and exceeds the rate of the past year. The employment gain over the next four quarters (2.3 million) falls short of its rate the past two years (2.8 million per year), but this is starting from an unemployment rate of 5.0% compared with 7.2% at this point in 2013.
On the half empty side our outlook is gloomy compared with three months ago and even more so if you go back to earlier forecasts. The growth we see rests primarily on continued spending by the household sector on consumption and housing. Business investment and government are both weak relative to longer-term historical standards. The international sector is a drag on domestic growth. Finally, the risks we see are almost entirely to the downside, primarily from potential international shocks or from a significantly distorted financial market situation. Put differently, considering the range of plausible outcomes for the next year our baseline forecast falls closer to the high end than to the low end.