U.S. Economic Outlook*

Tonight, I'll begin by talking about the economy and monetary policy. Then I'll discuss housing markets, which currently are an important part of our policy deliberations and a topic of particular interest to this audience.

So let's start with economic growth. Real GDP—our broadest measure of economic output—increased at a 2.6 percent annual rate in the second quarter of the year. We'll get the third-quarter estimate in two weeks—the latest consensus is for the number to be 2.3 percent. So growth over the past six months has been slower than over the previous few years. But much of the slowdown is part of the natural process of the economy shifting from a high-growth, recovery mode to what is a more sustainable path for the longer run.

A number of factors should support sustainable growth. Over the past six months, employment has been increasing at about 120,000 per month, a bit above its long-run sustainable pace. The unemployment rate is low—4.6 percent. Productivity trends remain solid as technology continues to advance and firms learn more and better ways to use it. Furthermore, recent declines in oil prices should give household budgets a boost, and solid economic growth in other countries should increase demand for our exports. Still, there are factors restraining economic activity. The most notable one is weakness in the housing sector, and there also is uncertainty about how important this factor will be for overall growth. Putting these considerations together, over the next year or so I expect the economy to expand, on average, somewhat below its long-run sustainable growth rate—which many economists estimate as roughly 3 percent per year. Of course, that's an average—and I expect to see some volatility in the numbers on a quarter-by-quarter basis.

Of course, the Federal Reserve is charged by Congress with fostering financial conditions that support two objectives: maximum sustainable economic growth and price stability, or low and stable inflation. Many policymakers employ a comfort zone for inflation—this is a range for inflation that they feel is consistent with their view of price stability.

By my standards, inflation has been too high. I prefer to see it between 1 and 2 percent. But the 12-month change in price index for personal consumption expenditures excluding food and energy, also known as core PCE, has been running at or above 2 percent for 29 months, and in August it was 2-1/2 percent. In part, core inflation has been elevated as businesses passed on earlier increases in energy costs to the prices of their products. High levels of resource utilization also have added more generally to inflationary pressures.

Looking ahead, it's likely that core inflation will come down somewhat over time. The recent declines in oil prices clearly are a positive factor. And the expected deceleration in economic growth will help avoid sustained pressures from resource constraints. Still, there is a risk that core inflation could run above 2 percent for some time. We could be wrong about reduced pressures from resource constraints, or we could see further cost shocks. And perhaps most importantly, if actual inflation continues at high levels, it could cause inflation expectations to run too high. If firms and workers expect inflation to be high, they will want to compensate by raising prices and wages or building in plans for automatic increases. In this way, high inflation expectations can lead to persistently high actual inflation.

Taking all of the factors into account, my current assessment is that the risk of inflation remaining too high is greater than the risk of growth being too low. Thus, some additional firming of policy may yet be necessary to bring inflation back to a range consistent with price stability in a reasonable period of time. But that decision will depend on how the incoming data affect the outlook.

At our August and September meetings, the Federal Open Market Committee held the funds rate target at 5-1/4 percent. The Committee's decision to pause gives us more time to gather information on a number of important developments and assess their implications for the outlook for growth and inflation. And we were able to pause because inflation expectations have been contained. Nonetheless, we have to be vigilant in monitoring these expectations. If they did increase, it would be incumbent on the Federal Reserve to adjust policy to affirm our commitment to price stability.

Certainly one development that we are assessing—and one that could test the economy's resilience—is the slowdown in housing. As you all are well aware, after posting sold gains for nearly a decade, housing starts and sales of new and existing homes began to drop sharply early this year. We also have seen a marked deceleration—and by a couple of measures, outright declines—in house prices. Clearly, these developments impact many of you through your business interests in real estate—not to mention the value of your own homes. And as I noted, they also raise important questions for the economy as a whole. Will the declines in housing markets coincide with more general economic weakness—as we saw during the recessions in the 1970s and 1980s? Or will the overall economy continue to grow at moderate rates despite a contraction in housing, as we saw in the mid-1960s?

There are simply too many factors in play to be able to give clean answers to these questions. However, what I would like to do is step back and look at recent developments in the context of the economic fundamentals underpinning residential investment. Such a perspective can help us gauge how the current situation may play out and can highlight some important sources of uncertainty to the outlook.

Let's start by looking at some facts about the role of residential investment in the macroeconomy. Residential investment includes spending on the construction of new single- and multi-family homes as well as renovations and improvements to existing homes. On average, over the past 50 years residential investment has accounted for a bit under 5 percent of GDP. That's about a third more than businesses spend building factories, offices, and commercial space, but in the same ballpark as how much households spend on recreation, which includes items such as golf clubs, football tickets, and, as the Commerce Department likes to call it, legitimate theatre and opera.

Five percent of GDP sounds small; but residential investment also is highly volatile, and over shorter periods of time it can influence GDP growth a good deal. For example, residential investment increased more than 25 percent and accounted for a full percentage point of GDP growth in 1976. But it fell at close to a 20 percent rate during the twin recessions in 1980 and 1981 and subtracted nearly a percentage point from growth in each of these years.

It's important to remember, though, in the early 1980s mortgage rates were almost 20 percent and the unemployment rate reached nearly 11 percent. The drop in housing then largely was the result of these macroeconomic influences, and not the cause of them. That is, the recessions did not occur because residential investment fell so much; rather, the factors that caused the recessions also resulted in sharp declines in residential investment.

Of course, lots of factors affect residential investment both on a quarter-by-quarter basis and in terms of the longer run.

Now investment bankers like to talk about stocks and bonds. But economists are boring—we like to talk about stocks and flows. Residential investment is a flow—it's the stream of new housing units being built. This flow adds to the capital stock of housing—it increases the total number of housing units in the economy. Of course, it's not just the number of houses that matter, but also their size and quality—the capital stock of housing is an overall concept that encompasses the number, size, and quality of homes in the U.S.

A number of longer-run factors underpin the demand for housing. One is demographics; as the number of households increase, so does the demand for housing. Another is long-run income prospects; as these improve, people will demand bigger and better homes. And the investment value of a home—as influenced by taxes and other factors—also underpins the demand for housing.

Suppose a change in some factor—such as income prospects—causes a one-time increase in the demand for housing. Residential investment (the flow) will have to increase to build the new, bigger houses. But once they are built, growth in residential investment will drop back to roughly its old value. In other words, a one-time step up in the demand for the stock of housing results in an up-and-down cycle in residential investment. Economists talk about this as the stock-adjustment cycle—the natural fluctuation in residential investment in response to changes in the desired housing stock. But as the fundamentals stabilize, the long-run growth rate of the housing stock will settle down to some sustainable rate. This rate also anchors the flow of residential investment because, averaged over long periods of time, the stock and the flow must grow at the same rate.

That said, people have a good deal of discretion in determining exactly when they purchase a new home. So even if underlying fundamentals do not change, over the short run residential investment can fluctuate a good deal more than what's due to stock adjustment considerations. In particular, people will react to the short-run swings in housing affordability caused by cyclical movements in income, unemployment, and interest rates. As a result, residential investment can fall sharply during recessions as households delay purchases because they are concerned about near-term prospects. And it can then surge early in a recovery as households that had delayed spending jump back into the market when conditions turn more favorable.

The 1980s provide an example of how both long and short-run developments simultaneously influence residential investment. With regard to the long run, there was a marked decline in the rate of growth of households in the U.S. as the baby boomers finished moving out of their parents' homes to set up their own households. This reduced the underlying demand for new housing units and helped generate a stock-adjustment cycle. In the end, the average growth rate of the housing capital stock fell about a percentage point between the 1970s and 1980s. But this longer-run development often is forgotten because of the dramatic short-run cyclical swings that occurred back then. I already mentioned the plunge in residential investment during the twin recessions, but it also surged more than 50 percent in the first two years of the 1983-84 recovery.

As an aside, the story is somewhat different when we consider the cyclical swings in total construction—the sum of residential and nonresidential investment. Overall, nonresidential investment generally is less volatile then residential construction, and the swings in growth of the two series are relatively independent. For example, while residential investment fell by more than one-third during the twin recession in the 1980s, nonresidential investment changed little on net. So total construction spending was off just 20 percent during that period. Indeed, we are seeing such differences today. For example, payroll employment in residential construction has changed little, on net, so far this year, but employment in nonresidential construction has expanded at a healthy clip.

Let's now consider housing during the 1990s. Longer-term fundamentals changed again. Household formation rates slowed further, but other factors boosted housing demand. Starting in 1995, the growth in the housing capital stock began to increase, eventually rising to a bit above 3 percent over the past couple of years. In order to achieve this, residential investment rose even more—it averaged an 8-1/2 percent annual growth rate between 2001 and 2005.

What developments contributed to this? Part of the pick up reflects an increase in the share of households who own their homes. After changing little over the previous 10 years, the homeownership rate rose steadily, from about 64 percent in 1994 to an all-time high of 68.9 percent in 2005. Because an owner-occupied house typically is bigger than a rental unit, the shift towards homeownership raised the growth in the housing capital stock.

Much more significant, though, is the fact that people are demanding bigger and better houses. Between 1995 and 2005 the median size of new homes increased nearly 20 percent, and the new homes were coming with more bathrooms and other amenities. Spending on home improvements also rose noticeably.

An important factor supporting these increases was the marked pick up in the underlying rate of productivity growth of the economy that began in the mid-1990s. This raised the wealth and long-run income prospects of American households, some of which was spent on bigger and better houses.

In addition, housing has become a more attractive investment. Borrowing rates fell and changes in the tax code favored housing relative to other investments. And, importantly, financial innovations lowered borrowing costs and increased access to mortgage markets. The momentum of these changes continued to increase through the 1990s.

One financial innovation was securitization. Most single-family mortgage originations now are bundled together and sold as mortgage-backed securities. Bundling reduces risk by averaging defaults and prepayments across a large number of loans, and securitization spreads the remaining risk among the large number of investors in the pool. Thus, both lower the risk premia built into mortgage rates.

Other innovations increased the efficiency of mortgage origination. Automation has made the process—including credit scoring—much less costly. In addition, homebuyers can now easily shop the internet for a wide range of competing mortgage brokers for the lowest interest rates. These changes have contributed to a drop in initial fees and charges on mortgages from about 2-1/2 percent in the mid-1980s to under 1/2 percent in recent years.

Finally, instruments such as sub-prime mortgages, interest-only loans, and hybrid ARMs have opened up financing to borrowers who previously could not obtain it at all or could not borrow as much as they would like. True, these instruments are riskier than traditional mortgages. Still, to the extent that both borrowers and lenders understand the risks involved and markets have priced this risk properly, they represent a net gain to society. Here, there is a role for public policy: On the part of the Fed, we are promoting financial literacy efforts for borrowers and supervising lenders with regard to both the disclosure of terms and costs to borrowers and to the risks of carrying such non-standard loans on their books.

Together, these financial innovations likely have been an important factor in boosting the desired stock of housing. A strong piece of evidence is the dramatic increase in the homeownership rate that I mentioned earlier. Many people who were not able to access or afford sufficient credit now are able to get a mortgage and purchase a home. And homeownership rates are up across nearly all demographic groups and income categories. The increases have been particularly impressive for Hispanics and Asians. So, although more can be done, minorities and others who in the past had difficulty obtaining credit are sharing in the gains. This is a quite positive development.

So where do we go from here? As I said earlier, I can't answer this question definitively. But the analysis I just discussed can help structure our thinking.

Some factors clearly point to soft residential investment—indeed, we've already seen it post a double-digit decline in the second quarter, and it likely fell a good deal in the third quarter as well. Part of the swing reflects the natural stock-adjustment mechanism I described earlier. Growth in residential investment and the residential capital stock must average the same rate over the long run. So the 6 to 12 percent gains in investment we experienced the past few years clearly are not sustainable against the backdrop of the 3 percent growth rate in housing capital. Moreover, housing stock growth might settle out somewhat below this 3 percent rate. Some of that growth likely reflected one-time adjustments, such as the extraordinary rise in the rate of homeownership. Furthermore, demographics point to slower growth in the working-age population in the years ahead, which would tend to reduce household formation.

In addition, there is the issue of the large increases in house prices that we have seen over the past 5 years. Now, it's important to remember that the factors I just mentioned that caused fundamental increases in the demand for housing also should be reflected to some degree in higher house prices. That said, in some locales, such as Las Vegas, house price increases seem to have defied gravity. To the extent that factors unrelated to demand fundamentals are boosting house prices in some regions, there is a risk that they could unwind and reduce residential construction. This risk is much smaller in other cities, particularly many here in the Midwest, which have experienced only modest house price appreciation.

Furthermore, there are some very important fundamentals that should continue to support housing demand. First, productivity growth should continue to bolster gains in wealth and income—some of which will be spent on housing. Second, the financial innovations that have led to more efficient mortgage markets and increased access to credit are here to stay. Third, financial conditions are not very restrictive—after all, 30-year mortgage rates currently are under 6-1/2 percent. Furthermore, there are no "extra" factors reducing the supply of credit—contrast today's environment with the 1960s and 1970s when there were regulatory ceilings on interest rates or with the early 1990's when there was the "capital crunch."

So, where does this leave us? On balance, we at the Chicago Fed expect some further weakness in residential construction as markets work through the stock adjustment process and there is some unwinding of unusual price movements. Some declines in residential investment could be large, and could contribute to some volatile numbers for overall GDP growth on a quarter-by-quarter basis. But currently we do not see the slowing in housing markets spilling over into a more prolonged period of weakness in the overall economy. So on balance, we expect GDP growth to average somewhat below potential over the next year or so. There are risks, of course. For example, negative wealth effects from softening house prices or the impact of higher interest rates—manifested in part through the repricing of ARMS—could reduce consumption more than anticipated. Such developments deserve monitoring.

But we cannot forget the plus side of the ledger. Outside of housing, there do not appear to be any imbalances that would foreshadow large adjustments in spending in other sectors. And, as I noted earlier, income growth and spending should be supported by the trends in productivity, recent oil price declines, and improved growth prospects for many of our trading partners.

With regard to inflation—it has been too high. Looking ahead, I think core inflation will gradually come down somewhat over time. But there are substantial risks to that forecast. Persistently elevated inflation rates could show through in inflation expectations. If this occurs, we will have to act accordingly.

Looking ahead, I am optimistic about the fundamentals of the US economy. Our nation's core economic values—our belief in free markets and competition, our use of technology and innovation, and our openness to trade—give the economy the ability to weather short-term challenges and provide a solid foundation to expand over time. And the Federal Reserve's commitment to price stability means that we will be achieving this in an environment of low and stable inflation.

---

*The views presented here are my own, and not necessarily those of the Federal Open Market Committee or the Federal Reserve System.