Moreover, in the first half of 2025, there was a sharp decline in real nonresidential investment in structures. This includes things like factories, warehouses, office buildings, and shopping centers. Yet it is likely that there was a big increase in investment in data warehouses, which would be included in the larger category of structures. If so, that implies that other investments in structures were even less and that investment in gen AI was even more impactful. The weakness of investment in structures was likely due to businesses postponing decisions about the location of facilities because of tariff uncertainty.
If investment in gen AI is so massive, this might explain the frothiness of equity prices at a time when there is reason to expect a weakening of the overall economy. In fact, the so-called Magnificent Seven tech-related companies accounted for about half of the increase in the S&P 500 index of US shares in 2024.
A quarter of a century ago, there was the so-called “dotcom bubble,” in which shares in tech-related companies surged dramatically, ultimately leading to a market correction as the profitability of investments in dotcoms came into question. Moreover, that correction contributed to a mild recession. In today’s case, a market correction is a potential scenario to consider. Tariffs and immigration policy are likely to slow the US economy. When that slowdown becomes apparent, it will likely result in lower nontech equity prices. Tech prices could come down as well if investors liquidate positions to cover other losses.
Also, the other big area of concern related to gen AI is electricity. The International Energy Agency (IEA) predicts that, in the United States, “power consumption by data centers is on course to account for almost half of the growth in electricity demand between now and 2030. Driven by AI use, the US economy is set to consume more electricity in 2030 for processing data than for manufacturing all energy-intensive goods combined, including aluminum, steel, cement, and chemicals.” In addition, the IEA said that “global electricity demand from data centers is set to more than double over the next five years, consuming as much electricity by 2030 as the whole of Japan does today.”
Moreover, about half of planned increases in electricity capacity in the United States involve renewable energy sources. With subsidies for such investment being cut, this capacity will involve higher costs for energy consumers. In fact, the Deloitte Research Center for Energy & Industrials predicts that these trends will result in a significant increase in electricity costs for US households. That, in turn, could have a negative impact on consumer demand.
While investment in gen AI is playing a big role in driving economic growth and asset valuations in the United States, there remains uncertainty as to when this investment will pay off in terms of productivity gains. Theoretically, gen AI should eventually have a big positive impact on labor productivity, thereby generating faster economic growth and improvements in living standards. On the other hand, it will significantly disrupt labor markets; eliminating some jobs while creating others.
Yet what we know from history is that there is a lag between the introduction of new technologies and their impact on productivity. It takes time for new technologies to become fully embedded in an economy. In part, this is because it takes time to figure out the best ways to utilize new technologies. Although gen AI is already having an impact in some industries and some processes, it is not yet fully integrated into the economy in a way that drives overall productivity.
However, the CBO also published alternative scenarios based on different assumptions. One assumption that is very important is the predicted rate of growth of total factor productivity (TFP). This is a measure of the increase in output growth that comes after accounting for increased supplies of labor and capital. It reflects the impact of technological innovation and improvements in processes. If the introduction of generative AI is successful in the years ahead, it will likely boost TFP growth.
Although the CBO does not explicitly discuss the impact of gen AI on productivity, it does offer an alternative scenario in which TFP grows 0.5 percentage points faster than the baseline scenario for the foreseeable future. Under this scenario, federal debt held by the public would be 113% of GDP by 2055 versus a baseline scenario of 156%. That is because a stronger economy would generate faster growth of revenue. In fact, under the faster productivity growth scenario, real GDP per person would be 17% higher than under the baseline scenario by 2055.
Meanwhile, one might expect that, if government borrowing turns out to be less than currently anticipated, this would mean lower bond yields. On the other hand, if the economy grows faster (all other things being equal), that implies higher bond yields. Meanwhile, faster productivity growth would likely mean lower inflation than otherwise.
Of course, no one knows how much faster productivity will grow due to the introduction of gen AI, or even if it will grow faster at all. If productivity grows even faster than the CBO’s alternative scenario suggests, then theoretically, the budget deficit could go away. We simply don’t know. We know that, historically, productivity growth has been uneven and unpredictable. We also know that there has generally been a large lag between the introduction of radically new technologies and their impact on productivity. This was true of computers, which were widely introduced in the 1980s, but where productivity acceleration did not take place until the late 1990s.
Another factor that could influence the future trajectory of deficits and debt is immigration policy. The amount of immigration affects the size of the labor force and, consequently, the ability of the economy to grow. In recent decades, immigration has played a very significant role in population growth in the United States and an even bigger role in labor force growth. Currently, foreign-born people are roughly 14% of the US population and 19% of the labor force. Because the birth rate has been very low, the nonimmigrant population is barely growing. The CBO estimates that, if immigration ceases, the US population will start to fall in 2033.
The US administration has introduced a significant shift in immigration policy. This has resulted in nearly zero entries into the United States by undocumented foreigners in recent months. In addition, the administration has taken an aggressive stance on deportations, with a goal of 3,000 deportations per day. If this number is achieved, labor supply would fall by a million people per year. A combination of immigration restrictions and increased deportations would disproportionately affect industries that employ large numbers of undocumented foreigners. For example, the US government estimates that 43% of crop workers, 14% of construction workers, and 10% leisure and hospitality workers are undocumented. A shortage of such workers would lead to higher wage increases and, consequently, bigger increases in prices. It would also restrain growth of output.
Overall, it has been estimated that, if legal immigration remains unchanged while illegal immigration ends and deportations increase, the US will face higher inflation, slower employment growth, and slower growth of real GDP. That, in turn, would mean slower government revenue growth and, consequently, larger deficits and debt.
Meanwhile, the index of selling prices of services had been rising faster than the index for goods for several years. Yet, in the United States, in the most recent few months, the index for goods prices rose much faster than for services. This probably reflected the impact of tariffs on goods prices.
The survey by S&P Global found that many US companies attributed the increase in their selling prices to the impact of tariffs on import prices. Other factors were the decline in the value of the dollar and rising wage costs, which were attributed to shortages of labor.
S&P Global reported that “the strength of the price growth reported in the United States remains broadly indicative of consumer price inflation running close to 4%, according to historical comparisons.” This is roughly consistent with the prediction by the Yale Budget Lab that tariffs will add 2 percentage points to existing inflation in the coming year.
If inflation accelerates significantly in the months to come, it could create a challenge for the Federal Reserve. It has a dual mandate to minimize inflation and maximize employment. If inflation is rising and employment is weakening, it will have to choose whether to target one or the other. The Fed leadership has stated previously that failure to control inflation ultimately leads to a weaker job market. Thus, the Fed’s first focus ought to be inflation. On the other hand, if it believes that inflation will be a one-off event, it might be less focused on it.
The separate survey of establishments found that, in July, there were 73,000 new jobs created. Yet, there were 73,300 new jobs created in health care and social assistance. This means that, excluding that category, there was no job growth in July. Moreover, overall job growth in May and June was revised down to 19,000 and 14,000, respectively. Thus, in the past three months, only 106,000 jobs were created; in the three months before that, 380,000 jobs were created.
In July, there was a decline in employment in mining, manufacturing, information, professional and business services, and government. In most other categories, job growth was very slow.
Also, the establishment survey found that average hourly earnings were up 3.9% in July versus a year earlier, the highest since March 2025. As such, despite a significant weakening of demand for labor, compensation growth for workers is not changing. This might be due to the decline in immigration. This has had the effect of keeping the labor market relatively tight despite weak growth. For now, real (inflation-adjusted) wages are rising faster than prices, thereby providing workers with improved purchasing power. This could change, however, if inflation accelerates significantly due to tariffs.
In response to today’s employment report, US equity prices fell sharply. In addition, the value of the US dollar fell sharply against the euro and the Japanese yen. The yield on the US government’s ten-year bond fell as well. Evidently, investors interpreted the jobs report as signaling a slowdown in economic growth. Indeed, the futures markets show that investors now see a 42% probability that the Federal Reserve will cut interest rates three times by the end of this year. That is up from 8% yesterday.
In this case, there has not yet been substantial retaliation, but the increase in prices and the commensurate reduction in purchasing power will likely have a negative impact on US growth. The Yale Budget Lab estimates that real GDP growth in 2026 will be 1 percentage point below the baseline forecast and, in the years to come, annual growth will be 0.5 percentage points below the baseline. This assumes that the current tariff rates remain in place. The reduction in growth will likely be driven by weakness in both manufacturing and construction. Although durable goods manufacturing is expected to improve, other areas of manufacturing may worsen, especially advanced manufacturing.
Leaving aside the Yale Budget Lab forecasts, the reality is that an 18% tariff applied to last year’s US$3.3 trillion in imports implies roughly US$600 billion in additional annual government revenue, or about 2.1% of GDP. Think of that as an annual tax increase of 2.1% of GDP, more than offsetting the impact of the tax cuts in the bill recently passed by the US Congress. All other things being equal, that will likely have a sharp negative impact on real GDP growth. Moreover, even if the tariffs are not fully passed onto consumers, then businesses will see a sharp decline in profits. Either way, there could be a negative economic impact.
On the other hand, such a large increase in the prices of imported goods would likely lead to a decline in the volume of imports, thereby reducing the net impact on GDP. If there is a substantial decline in the volume of imports, that will be a negative shock to the exports of other countries. Moreover, if businesses begin to see the new level of tariffs as permanent, it will likely lead to a redesign of supply chains as well as significant shifts in global trade patterns.
Finally, one might argue that, given the large US budget deficit, a tax increase might be welcome. Yet the challenge with tariffs as a source of revenue is that they have different impacts on different industries, shift production toward lower value-added processes, hurt business investment, exacerbate income inequality, and could ultimately damage productivity and growth.
This is all true. However, it is likely that these companies attempted to maintain market share on the expectation that tariffs would be temporary. They evidently believed that the tariffs imposed by the United States were meant to compel changed behavior by other countries, after which tariffs would be reduced. Yet that is not what has happened. Instead, the United States has compelled other countries to make certain commitments. In exchange, the United States has agreed to not impose steep tariffs but has left in place historically high tariffs. My expectation is, once companies recognize the permanence of tariffs, they will raise prices accordingly.
Indeed, one large consumer products company announced that, in the coming year, it will boost US prices by 5% to offset the US$1 billion impact of tariffs. Also, investors evidently expect tariffs to boost inflation. The five-year breakeven rate, which is a measure of bond investor expectations of average inflation over the coming five years, has lately increased by about 20 basis points as the final level of tariffs become clear. The breakeven rate had declined after the April tariff announcements as investors worried about a recession that would weaken inflation. Yet now, investors are not changing their view about growth, only their view about inflation.
On the other hand, many investors evidently believe that tariffs will weaken growth and that the boost to inflation will be relatively short-lived. As such, the futures markets indicate that investors expect the Federal Reserve to cut the benchmark interest rate two to three times before the end of this year.
Meanwhile, the US government released monthly data on personal income, consumer expenditures, and the Federal Reserve’s favorite measure of inflation, which is the personal consumption expenditures deflator, or PCE-deflator. This measure of inflation accelerated modestly in June, with prices up 2.7% from a year earlier and up 0.3% from the previous month.
However, the most notable part of the report concerns prices of durable goods, many of which are imported. The PCE-deflator for durables increased 0.5% in both April and June versus the previous month. This was the biggest gain since mid-2022. In addition, the PCE-deflator for durables was up 0.9% from a year earlier in June, the biggest gain since December 2022. Normally, prices of durables decline. Plus, the government also reported that the real volume of durable goods purchased by consumers fell sharply.
This is early evidence of the impact of tariffs. Moreover, these increases do not yet fully reflect the impact of tariffs as businesses have cut prices to partly offset tariffs. The result has been weakened profits. This will not likely continue, especially as it is now clear that high tariffs are here to stay. Thus, expect inflation to accelerate in the months to come. The Yale Budget Lab estimates that, if businesses fully pass through the impact of tariffs, inflation will nearly double in the coming 12 months.
The issue that has been of most concern to the US administration is the large merchandise trade deficit the United States has with the European Union. The United States also has a large services trade surplus with the European Union. When the two are combined—plus a few other things—the result is the current account balance, in which the United States has a modest deficit with the European Union.
Thus, with the goal of reducing the trade imbalance, the United States got the European Union to commit to purchase US$750 billion in energy from the United States over a three-year period. This will be in the form of oil, gas, and nuclear fuel. The European Union also committed to US$600 billion in new investments in the United States.
Yet it remains unclear how the massive increase in energy purchases will take place. Last year, the European Union imported US$436 billion in energy, including US$75 billion from the United States. The new deal commits the European Union to import US$750 billion in energy products from the United States in the next three years, or about US$250 billion per year. This would require a massive shift in imports from other countries to the United States.
In part, this might be achieved by the European Union’s continuing reduction in dependence on Russian oil and gas. Still, that would only account for a modest share of the transition. Moreover, the European Union cannot simply compel private sector companies in member states to shift their purchases. Thus, it is unclear what mechanism would allow this transition to take place. Plus, if this transition happens, the United States will either need to significantly boost output or shift exports from other countries to the European Union. Either way, it is not clear how this would happen, especially without leading to a big increase in oil and gas prices in the United States, which is the opposite of what the US administration desires.
Member states mostly expressed relief that the European Union will not face a much higher tariff. Moreover, the German government expressed relief that Germany’s automotive exports to the United States will soon face a tariff reduction. Still, some business leaders in Germany said that the 15% tariff will be painful. In addition, France’s Minister of EU Affairs said that “the current situation is not satisfactory and cannot be sustainable.”
In response to the announcement of the EU-US trade deal, European equities initially increased as investors were relieved that there was a deal. However, later in the session, European equities fell as the full implications set in. In fact, equity prices of German automotive companies fell sharply.
Meanwhile, the US dollar rose against the euro. This is interesting. Economic theory tells us that, when a country increases tariffs, the value of its currency rises, which is what happened today. Yet it didn’t happen earlier this year when the US announced major tariffs. Rather, since then, the dollar has fallen as global investors became less enamored of holding dollar-denominated assets.
In any event, the decline in the euro, if it continues, could be inflationary for Europe, thereby creating a challenge for the European Central Bank. The decline in the euro reflects two factors. First, the US tariffs will reduce demand for euros, thereby putting downward pressure on the euro. Second, the trade deal is expected to weaken the European economy by reducing export growth, thereby reducing demand for euros.
As for the United States, it imported roughly US$600 billion in goods from the European Union in 2024. Thus, a 15% tariff will probably have an impact on import prices and import volumes. The most important US imports from the European Union are pharma products and automobiles. Given that tariffs will be similarly high on many other countries, it is unlikely that the tariffs on the European Union will lead to trade diversion.
Complete a brief survey to provide your views on thought leadership content.