US Economic Forecast has been saved
The unexpected results of the presidential election have introduced uncertainty into both the US economy and our forecast. Plenty of questions remain, but it’s likely that we’ll see shifts in trade, infrastructure spending, taxes, and regulation—with a range of possible impacts for business and consumers.
It is rare for economic forecasters to change outlooks substantially over the course of a few months. Most economic news tells us only a little about how things are changing—one release is positive, the next negative, and the net difference is not all that large. But this time really is different.
The election of Donald Trump as president qualifies as a very significant change to the economic outlook. There is no simple template for the policy mix to be expected from his new administration. Presidential candidates generally don’t present policy proposals that are terribly detailed, and Trump’s proposals tended to be less fleshed out than usual.
In the long transition period, this mainly takes the form of uncertainty. And US economic policy isn’t determined by the president waving a magic wand. It requires buy-in from Congress, and Trump’s relationship with Congress is itself difficult to predict. Future policy outcomes range widely as a result.
Our baseline reflects a relatively successful implementation of Trump’s key economic proposals, although at more modest levels. These include:
The baseline also includes two key assumptions about expectations in the short run:
We do not judge that the supply-side elements of the program, such as regulatory relief, will have a large impact in the five-year forecast horizon. This doesn’t mean that they might not be important for the economy’s long-run growth. But over the next few years, the impact on economic aggregates (such as GDP, employment, and inflation) will be dominated by the large demand-side impacts of this policy mix.
All economic forecasts include assumptions about policy. But this forecast’s accuracy will be particularly sensitive to the policy choices and accomplishments of the new administration. Each of the main macroeconomic policy interventions—the infrastructure spending plan, the tax cuts, and raising the trade barriers—can be carried out in larger or smaller measures. The resulting policy mix will almost certainly differ from the assumptions here. How it will be different is a big question.
The result of our assumption is a few quarters of slow growth starting in 2017 because of the unusual level of uncertainty (particularly around international supply chains). By early 2018, that uncertainty may give way to faster growth because of the demand-side impact of tax cuts and infrastructure spending. By 2020, the infrastructure program likely begins to wind down, reducing GDP growth.
With the economy now relatively close to full employment in the model, the demand stimulus creates inflationary pressures. That’s made worse by the trade barriers, which may raise prices for consumer goods as well as increase the cost of doing business. This can lead to much faster Fed movement and much higher long-term interest rates in the baseline simulation than in our previous forecast.1
Confusion abounds when it comes to describing the reasons how a program such as infrastructure spending or tax cuts will affect the economy. Such policies affect both the level of aggregate demand (demand side) and the economy’s capacity to produce (supply side).
The demand-side impact is simple: Additional demand for spending induces businesses to produce more, and GDP rises. The extra production generates additional jobs and income, which creates more demand and production—the multiplier effect. The demand-side impact is often associated with British economist John Maynard Keynes, who advocated for demand-side policies to solve the economic problems of the 1930s and famously suggested that even useless government spending might be good for the economy:
If the Treasury were to fill old bottles with banknotes, bury them at suitable depths in disused coalmines . . . and leave it to private enterprise on well-tried principles of laissez-faire to dig the notes up again . . . there need be no more unemployment and, with the help of the repercussions, the real income of the community, and its capital wealth also, would probably become a good deal greater than it actually is.2
Of course, it does matter what the money is spent on, which leads us to supply-side policies. While demand-side policies attempt to put underused resources to work, supply-side policies attempt to increase the capacity of the economy—to create more potential resources. Such policies include inducements to investment (such as cutting taxes on capital), encouragement of innovation, making labor markets work better, and reducing the regulatory burden on businesses.
Many policies have both demand- and supply-side components. For example, the proposed infrastructure spending program would put people to work (demand side) while increasing the economy’s capacity by adding to the public capital stock (supply side). Business tax cuts typically have short-run demand-side impacts (as the demand for investment goods rises) and long-run supply-side impacts (as the amount of capital grows).
In the short run—for example, in the five-year horizon of the Deloitte forecast—demand-side impacts tend to be much larger than supply-side impacts. A $100 billion rise in investment translates to an immediate demand-side impact of that amount, plus more from the multiplier effect. If the capital returns 10 percent per year, the economy’s capacity is increased by only $10 billion in a given year, however.
In the long run, however, the demand-side impact fades away as the economy returns to long-run equilibrium in which demand and capacity are balanced. The supply-side impact, on the other hand, is permanent (less depreciation). And an additional $100 billion of new investment over 10 years will eventually raise the capacity of the economy—and long-run GDP—by that amount.
It’s important to keep this distinction in mind when judging the impact of potential policy. Unfortunately, boosters for any particular policy are often content to use the confusion to their advantage.
Our scenarios are designed to demonstrate the different paths down which the new administration’s policies might take the US economy. Foreign risks have not dissipated, and we’ve incorporated them into the scenarios. But for now, we view the greatest uncertainty in the US economy to be that generated within the United States.
The baseline (55 percent probability): Uncertainty restrains business investment in early 2017, but tax cuts and infrastructure spending push up GDP in 2018 and 2019. With the economy at full employment, the faster GDP growth creates some inflationary pressures. Increased trade restrictions add to the price pressures. The Fed moves aggressively to prevent inflation from picking up too much, and long-term interest rates rise substantially. Growth rises to 2.5 to 3.0 percent for a couple of years before falling off as the impact of the stimulus fades.
Recession (5 percent): Sudden policy changes in the United States, including a large tariff on Chinese goods, trigger a global financial crisis. The crisis is exacerbated by the sudden change in global supply chain cost structures from higher US trade barriers, as well as retaliation from China. The Fed and the European Central Bank act to ease financial conditions, and growth starts to pick up as businesses adopt to the new global costs and restructure their capital to reflect the new global cost structure. GDP falls in the last two quarters of 2017 and recovers after 2018.
Slower growth (30 percent): The infrastructure program and tax cuts stall in Congress. And the administration has placed significant restrictions on US imports, raising costs and disrupting supply chains. Businesses hold back on investments to restructure their supply chains because of uncertainty about future policy. Global growth also slows because of the supply chain disruptions, reducing demand for US exports. GDP growth falls to 1 percent over the forecast period, and the unemployment rate rises.
Successful policy takeoff (10 percent): The administration takes only symbolic action on trade. With supply chain disruptions off the table, businesses focus on tax cuts that are designed to increase investment spending, and the opportunities available from the infrastructure plan. Tightening labor markets attract many people back into the labor force, and the high labor-force participation rate helps to moderate the impact of faster growth on wages and inflation. Growth remains above 2 percent for the next five years.
The household sector has been the bright spot of the US economy for the past year or two. While business investment turned negative and housing stalled, consumer spending has grown steadily. But that’s not surprising, since job growth has been quite strong for the past few years. Even with relatively low wage growth, those jobs have put money in consumers’ pockets, enabling them to spend while keeping savings at a high (for the United States) level.
Consumer confidence has remained strong, although both major measures fell in October. Was the long, divisive campaign to blame? It’s hard to say. Observers will be watching closely to see if the Trump upset has an impact. For nearly every consumer shaken by the result, there is a consumer who is happy with it. There’s no reason to expect confidence to fall much—so long as job growth continues.
American households still face some obstacles in their pursuit of the good life. They have (mostly) recovered from the over-borrowing of the 2000s, though many remain “underwater,” with houses worth less than what the household owes on the attached mortgage. And there is the problem of growing income and wealth inequality. The possibility of higher inflation would be welcome for reducing the number of underwater households, since their borrowing is in nominal dollars. But it’s unclear whether—much less how—the new administration intends to tackle the inequality problem.
Many US consumers spent the 1990s and ’00s trying to maintain spending even as incomes stagnated. After all, excitable pundits kept assuring them that the technology transforming their lives would soon—any day now—make them all wealthy. But now they are wiser (and older, which is another problem, as many Baby Boomers face imminent retirement with inadequate savings). As long as a large share of the gains from technology and other economic improvements flow to a relatively small number of households, overall US consumer spending is likely to remain relatively restrained.
Real consumer expenditures grew over the summer at about 0.3 percent per month, which is a healthy pace. The savings rate was steady at around 5.7 percent, which is relatively high for the United States.
Headline retail sales were strong in September and October, bouncing back from weakness in the late summer.
Consumer confidence has generally remained elevated, although October saw a decline (to still-high levels). Continued strong job growth and some stirrings of wage gains are likely keeping consumers buoyant.
Every year, thousands of young Americans abandon the nest, happy to leave home and start their own households. But more than usual stayed put during the recession: The number of households didn’t grow nearly enough to account for all the newly minted young adults. We expect those young adults would prefer to live on their own and create new households; as the economy continues to recover, they will likely do exactly that—as previous generations have.
This likely means some positive fundamentals for housing construction in the short run. Since 2008, the United States has been building fewer new housing units than the population would normally require; in fact, housing construction was hit so hard that the oversupply turned into an undersupply. But the hole is shallower than you might think. Several factors offset each other: If household size returns to mid-2000s levels, we would need an additional 3.2 million units; on the other hand, household vacancy rates are much higher than normal. Vacancy returning to normal would make available an additional 2.5 million units—which would fill 78 percent of the pent-up demand for housing units.
But are the existing vacant houses in the right place or condition, or are they the right type, for that pent-up demand? The future of housing may look very different than in the past. Growth in new housing construction has been concentrated in multifamily units. If that persists, we may find it is related to young buyers’ growing reluctance to settle in existing single-family units.
In developing our housing forecast, we assumed that the demand for housing (in the form of the average household’s size decreasing) picks up this year, vacancy rates gradually drop, and household depreciation begins falling after new renters and buyers remove about 2.5 million housing units from the nation’s housing surplus. Slowing population growth suggests that we will have a short-lived housing boom in which starts hit the 1.3–1.4 million level, followed by a period of contraction until starts reach the level of long-run demand. We estimate this to be about 1.0 million units in the medium term. Housing will likely contribute to GDP growth in 2017–18—particularly if economic policy creates a boom—but subtract from GDP growth by 2019 as the pent-up demand dissipates. In the long run, the slowing population suggests that housing will not be a growth sector (although specific segments, such as housing for elderly residents, might well be very strong).3
While economic growth and job creation may point to strong house sales, higher interest rates may moderate any potential housing boom. Higher inflation and a strong Fed response may drive up mortgage rates more quickly than businesses in the housing sector would like.
Housing permits fell in the late spring and then stalled for two months. September saw a significant rise in permits, suggesting (to the optimistic) that housing construction would start to grow again. At an annual rate of 1.2 million, permits are still below the level required to make up for the many years of low housing growth. Much of the volatility in permits was, however, in the multifamily sector. Single-family starts have been growing at an average rate of about 5 percent over the previous year’s level for several months.
Contract interest rates were flat through October but likely rose in late November. Long-term interest rates had begun rising even before the election. House prices continue to move up and are about 5 percent above the previous year’s level (according to the Case-Shiller national index).
Many may blame election-season uncertainty for lagging investment, but it’s easy to find more fundamental reasons for the weakness in this area over the two past years.
First, oil and gas extraction accounted for 6 percent of all nonresidential fixed investment in 2013. That’s a hefty amount (considerably larger than the sector’s value-added share), so shutting down new US oil exploration in late 2014 had a larger-than-expected impact on investment. Indeed, the 2015 data on investment by type show that most of 2015’s decline was due to two specific categories: mining structures and mining and oilfield equipment. Other types of investment—ranging from commercial structures to transportation equipment to intellectual property—have held up much better. It’s likely the continuing low price of oil, not the low state of political debate, that has held down investment.
The weakness in investment spread to other areas in the first half of 2016. However, it’s not hard to find a key culprit in the fundamentals: The rising dollar is not only making US companies less competitive—it’s cutting overseas earnings valued in dollars and therefore reducing margins for US multinationals. And China’s slowing growth is exposing global excess capacity in many industries.
Our baseline scenario does assume a role for uncertainty in the first half of 2017. The problem businesses face is deciding whether they need to rebuild their supply chains. Industries such as automobile production have developed intricate networks across North America and reaching into Asia and Europe. They’ve assumed that materials and parts can be moved across borders with little cost or disruption. These capital-intensive industries are likely to want to postpone easily delayed investments in these networks until the administration’s trade policy becomes clearer.
Beyond the immediate impact of this policy uncertainty, the truth is this: If and when demand finally picks up, businesses will almost certainly be willing to spend on the plants and equipment necessary to meet that demand. In the baseline forecast, suppression of trade accompanies a modest boom in private investment that is induced by higher infrastructure spending and demand from tax cuts.
Real business fixed investment rose 1.1 percent in the first quarter (according to the first GDP release). That’s the second quarter of very slow growth, but better than the 3 percent decline in Q4 2015 and Q1 2016. Structures investment and intellectual-property investment rose at healthy 4 to 5 percent rates, but total business fixed investment was held down by a 2.7 percent decline in equipment investment.
Nondefense capital-goods shipments—an effective high-frequency measure of equipment spending—rose 2.2 percent in September after three months of decline. Much of the improvement was in aircraft, however, as capital goods shipments less aircraft grew 0.4 percent.
Private nonresidential construction fell in September after rising from May through August. Commercial and manufacturing construction both dropped substantially in September, while office construction also fell, albeit by a smaller amount.
Yields on corporate bonds started to creep up even before the election but then jumped, raising the cost of capital. Stock indexes made headlines by reaching new highs after the election, as traders anticipated the impact of tax cuts, higher infrastructure spending, and (especially for financial companies) reduced regulation. Corporate profits fell in the second quarter and have been trending down since the end of 2014. However, profits remain at close to a record share of national income.
Over the past few decades, business—especially manufacturing—has taken advantage of open borders and cheap transportation to cut costs and improve global efficiency. The result is a complex matrix of production that makes the traditional measures of imports and exports misleading in some senses. For example, in 2007, 37 percent of Mexico’s exports to the United States consisted of intermediate inputs purchased from the United States.4
Recent events appear to be placing this global manufacturing system at risk. Brexit, which may affect the United Kingdom’s position in the European manufacturing ecosystem, along with the suggestion that the United States would cancel or renegotiate its position in NAFTA, may slow the growth of this system or even cause it to unwind. The impact would be felt in several ways:
The forecast assumes some retaliatory measures that reduce US exports—but not an all-out trade war.
The current account is determined by global investment flows, not trade costs. The forecast assumes that the current account deficit therefore remains at the level of our previous baseline.
US goods exports grew at rates of 1 to 2 percent per month from June through December. That’s a welcome development, as exports are now contributing to GDP growth. Imports have been flat, or down a bit—and that goes for nonpetroleum imports. As a result, the US trade deficit declined from $45 billion in June to $36 billion in September.
The dollar has appreciated in September and October, and jumped further after the election in November.
The Chinese economy is recording satisfactory growth, though many observers remain concerned about the country’s financial system and continuing infrastructure investment.5 Some analysts focus on signs of strength in China’s consumer and service sectors, suggesting a long-awaited adjustment to becoming a consumer-driven economy. Others point to indications that official Chinese figures may be implausibly high. The country’s future remains a large risk for the global economy.
Europe’s industrial production continues to fluctuate without growing strongly. Brexit’s impact on Europe—both the United Kingdom and the Eurozone—is uncertain, although there is no reason for any immediate impact. Over the longer term, the United Kingdom may face a significant loss of competitiveness and slower growth, although the effect on the United States is likely to be small.
Recently, we started to see a modest contribution of federal spending to GDP. That’s the result of the last federal budget agreement, which raised caps on both defense and nondefense purchases. The new administration, however, may change this if it decides to follow through on the infrastructure spending program on which President-elect Trump campaigned. The forecast assumes a more modest (although still quite large) rise in both federal and state and local investment spending. This is an important driver of the relatively fast growth expected in the baseline in 2018 and 2019. It also drives a decline in growth in 2020 and 2021 as the high level of infrastructure spending declines somewhat.
The rise in spending occurs along with a significant tax cut. The forecast assumes that the president and Congress will decide that these initiatives are sufficiently important to permit a significant rise in the federal deficit. If the deficit becomes an important debating point, passing either or both of these policy packages will become more difficult.
After years of belt-tightening, most state and local governments are no longer actively cutting spending. A portion of the infrastructure program is likely to take the form of transfers to state and local governments (most infrastructure investment in the United States takes place at the state and local level).
The federal deficit ended the year at $587 billion, up substantially from last year’s $439 billion. Receipts ended the year up just 0.6 percent, while outlays rose 4.5 percent.
Federal employment grew an unusual 0.3 percent per month from August to October. State and local employment remains flat.
If the US economy is to produce more goods and services, it will likely need more workers. However, many potential workers remain out of the labor force: They left in 2009, when the labor market was terrible. The stability of the participation rate in 2016, while the labor force is aging, is an encouraging sign. Accelerating production will carry with it an eventual acceleration in demand for workers, along with a welcome mild rise in wages. That should help to bring people back into the labor force.
But a great many people have been out of work for a long time—long enough that their basic work skills may be eroding. When the labor market tightens, will those people be employable? Deloitte’s forecast team remains optimistic that improvements in the labor market will eventually prove attractive to potential workers, and labor force participation will pick up accordingly.
Significant immigration restrictions and/or deportation might have a marginal impact on the labor force. According to the Pew Research Center, undocumented immigrants make up about 5 percent of the labor force.6 Removal of all such workers would clearly have a significant impact—but that is unlikely to happen. A more realistic assumption might be the return of about half a million undocumented workers annually. That would create labor shortages in certain industries (such as agriculture, in which 17 percent of workers are unauthorized, and construction, in which 13 percent of workers are unauthorized)7, but it would likely have little significant impact at the aggregate level.
Initial claims for unemployment insurance are holding steady, in the 260,000 range. Job openings have been flat, in the 5.5 million range. However, this is just under the highest level recorded (since 2002). Quits (voluntary separations) have also been stable at around 3.0 million—also a high number. The large number of voluntary quits suggests that labor demand remains strong.
Payroll employment rose at an average rate of 176,000 per month in August–October. That’s almost twice as much as would be required by the growth of the labor force, although slower than the rate recorded in 2014–15.
Interest rates are among the most difficult economic variables to forecast because movements depend on news—and if we knew it ahead of time, it wouldn’t be news. The Deloitte interest rate forecast is designed to show a path for interest rates consistent with the forecast for the real economy. But the potential risk for different interest-rate movements is higher here than in other parts of our forecast.
Global financial markets are now in a highly unusual state. About $10 trillion in sovereign debt is now trading at negative interest rates (meaning borrowers are paying for the privilege of loaning money to these countries). The existence of negative interest rates is unprecedented, and the fact that even large countries (such as Germany) are borrowing on these terms indicates that global financial markets have not fully recovered from the problems of the previous decade.
Despite this, the forecast sees both long- and short-term interest rates headed up—maybe not this week, or this month, but sometime in the future. The combination of tax cuts and infrastructure spending assumed in the baseline is very simulative. The Fed currently is assuming that the economy is near full employment, and would therefore likely react strongly to a pickup in growth such as assumed here.
For that reason, the current baseline assumes an aggressive Fed reaction, with the Fed funds rate at or above 4 percent by 2019. Long-term interest rates move even more, in anticipation of the rise in inflation in the forecast.
Speculation about the future of the Fed will continue to grow over the next year. The new administration has an immediate chance to make an impression on the central bank by filling two open slots on the Board of Governors and appointing a new chairman in February 2018.
Junk bond spreads rose in late October and have remained elevated, while the AAA corporate spread over Treasuries fell a bit after the election. The spread of long-term over short-term Treasuries picked up after the election, as traders anticipated that the new administration would promote economic stimulus that would lead to more government borrowing and ultimately higher inflation.
Stock prices remain elevated. Just after the election, equities plunged, but they rapidly made up the lost ground. Financial stocks did particularly well, as traders anticipated regulatory reform in the financial sector.
It’s been a long time since inflation has posed a problem for US policymakers. The US economy has been below potential since 2008, and even before then there were few signs of significant inflation. With so much slack, neither workers nor businesses had the ability to raise prices.
Many observers believe that the US economy is approaching full employment (although the Deloitte forecast projects some additional contribution from workers returning to the labor force as the economy continues to improve). Thus, the fiscal stimulus of tax cuts combined with infrastructure spending—if they occur as forecast—will likely create some shortages in both labor and product markets, and, as a result, some inflation. A return to 1970s-style inflation is unlikely, but it would not be surprising in those circumstances to see the core CPI running at 2.5 percent to 3.0 percent—or perhaps even a little higher. The forecast expects timely Fed action to prevent inflation from rising too much, but the price (of course) is higher interest rates.
The overall CPI grew at an average rate of 0.3 percent during August–October, and by October was up 1.6 percent over the previous year’s level. That mainly reflects energy prices finally starting to rise a bit. The core DPI was up just 0.1 percent in September and October, suggesting that there is little in the way of inflationary pressure in the economy.
Final demand PPI was up just 0.9 percent in the year ending in October. There is little inflation in the intermediate goods and services pipeline.
The hourly wage grew just 2.4 percent in the year to October, lower than the average growth recorded in the previous recovery. Some measures of pay, such as the employment cost index, are beginning to show signs of acceleration. If that continues, it will indicate that the economy is reaching full employment, and that continued demand growth could create higher inflation.