The United States March 10-15th

• Amidst fears of a weakening economy, private payroll employment rose 140k, in line with recent trends
• But the unemployment rate increased to 4.1% and the workweek failed to recover
• Tariffs were implemented and then partially reversed, but overall tariff risks are rising
• We forecast a 0.3%m/m (3.2%oya) rise in core CPI.

The key uncertainty at the start of this week was whether tariffs would be implemented on Canada and Mexico, or delayed again until early April. Unfortunately by the end of the week outcomes were still uncertain: tariffs were first enacted and then removed for some goods, though with some confusion as to the exact share of US imports still being tariffed. Canada struck back with a first wave of retaliatory tariffs, and as of Friday afternoon President Trump was signalling new tariffs on Canadian dairy and lumber. Almost lost amidst the headlines was the next wave of tariffs on all Chinese imports, which when combined with increases in early February yields a 20% jump in tariffs on China within a month. China imposed new retaliatory tariffs in response, though the dollar magnitude was small.

Whipsawing policy changes make forecasting growth and inflation this year an especially fraught exercise. We imagine many businesses that need to make investment decisions face an even bigger conundrum, and indeed this uncertainty factor is starting to crop up in business surveys. That alone is likely to negatively affect growth. Moreover, the fact that some Canada and Mexico tariffs went into effect suggests an increased risk of other tariffs hitting by early April. While we have not made any forecast changes this week, other than to downgrade 1Q GDP growth from 1.5% to 1.0%, we think a further growth downgrade and inflation hike for the remainder of the year has an elevated risk of coming soon given the direction in which policy is traveling.

In terms of the Canadian and Mexican tariffs, while the announced 25% tariff (10% on Canadian energy) went into effect on Tuesday, it was almost immediately taken away except for non-USCMA compliant imports and Canadian energy. The problem here is that there is uncertainty about what share of imports are, or can become, USMCA compliant. In particular, importers who might have previously not bothered to validate USMCA compliance because all tariffs were low may now decide to do so. Our views on how these tariffs might impact the economy are still evolving.

Next week will be light on data, with CPI and PPI being the main event. We expect a 0.3%m/m increase in both headline and core CPI, which would leave over-year-ago core inflation at 3.2%, similar to recent months.

Calm before the storm

This week’s headline data release, the February jobs report, showed a 151k increase in nonfarm payrolls, with an uptick in the unemployment rate from 4.0% to 4.1%. This outcome should essentially guarantee a hold on interest rates at the March FOMC meeting. Private sector job gains were 140k, exactly in line with the prior six months. Amidst a recent wave of more negative business and consumer sentiment there were fears that job growth could already be downshifting, but in fact the report was rather normal.

Moreover, with jobless claims back down around the 220k range at the start of March this month’s report could once again be unremarkable, at least for private-sector workers. Still, there were some soft details within the report, even as any slowdown stemming from recent policy shifts or uncertainty would not likely have had too significant an effect yet. Within the establishment survey, the main disappointment was in hours worked. The workweek had declined in each of the prior two months, and we had attributed the January drop to especially cold weather.

While the weather didn’t fully reverse in February, we had thought the workweek would bounce back—but in fact it held steady. That has created a wide gap between the recent trajectory of employment and hours worked, with employment rising 1.5% annualized over the last three months even as hours worked dropped 0.7% (Figure 1). Meanwhile, the three-month average annualized increase in labor income slowed to 3.9% last month. While this is the lowest number of this boomy cycle, it would still be enough to support decent real spending gains, provided purchasing power isn’t frittered away on import taxes.

Government employment grew just 11k, a third of the trend in recent months, with federal employment falling 10k as a hiring freeze struck that sector. Through the week ending March1 the number of federal workers claiming unemployment benefits has remained quite small, but government job cuts are set to further shrink federal employment in coming months. Challenger reported over 60k announced government cuts in February, and government memos reported in the press have called for cuts of up to 80k at Veterans Affairs and 45k at the IRS. We wrote more about government job cuts here.

As mentioned earlier, the household survey showed some softness. The unemployment rate narrowly remained at 4.1% (rising to 4.14%), while the participation rate dropped from 62.6% to 62.4%, marking the lowest level since early 2023. This decline was seen among younger and older workers, whereas the participation rate for prime-age workers (ages 25-54) remained stable. The participation rate for workers aged 55 and above has declined by 0.5 percentage points since October, reaching its lowest level since the pandemic. The aggregate unemployment rate has fluctuated between 4.0% and 4.2% since last May, largely reflecting modest changes in job finding and separation probabilities since last summer (Figure 2), and permanent job losers in the unemployed pool were little changed from a year earlier. Multiple job holding also moved up, which could include some government workers using deferred resignation (Figure 3).

1Q not looking so hot

This week we downgraded our 1Q GDP forecast from 1.5% to 1.0%, in response to a widening trade deficit and out of further concern that policy uncertainty may cause a slowdown in activity this quarter. We took that step ahead of the release of the February jobs report, which as noted continues to show quite reasonable job growth despite fewer hours worked.

With respect to the trade deficit, this week helped clarify how much of the dramatic January surge in goods imports was coming from nonmonetary gold, which must be stripped out when doing GDP accounting. Normally gold movements are small, but tariff concerns had caused a dramatic surge in imports in December and January. While total nominal goods imports rose 12%m/m in January, the increase was “only” 6% when excluding gold. But that is still a large number, and consequently even if the trade deficit returned to its average in the months prior to the election, the 1Q goods deficit could blow out at a $250bn annualized pace this quarter—enough to subtract a percentage point from growth.

While a wider deficit subtracts from GDP in an accounting sense, an import surge should be offset by higher inventory accumulation or final sales. However, neither of those appeared to be particularly strong in January, which is why some GDP trackers are so low. Some additional data points are suggesting caution on current-quarter GDP growth, although as usual the signals are not all aligned. For consumer spending, the 3% m/m rise in February unit auto sales is a positive sign, and our Chase card spending data indicate a small gain in control retail sales.

On the other hand, other high frequency metrics tracked in the Chicago Fed’s CARTS measure point to a weak month for sales. For businesses, measures of activity and new orders reported in the final PMI and the ISM for February were generally soft (Figure 4), especially for services and ISM manufacturing. While the new orders component of the manufacturing PMI continued to increase, the press release noted that this strength could be temporary as firms engaged in front-loading ahead of tariffs. Qualitative comments in both surveys frequently highlighted uncertainty stemming from tariffs or government policy in general. The qualitative comments in the Fed’s Beige Book also suggested that economic growth was stalling last month.

The United States March 3-7th

• Disappointing January consumption and trade data pull down our tracking of 1Q GDP growth to 1.5%
• China to get hit with another 10% tariff increase, while 25% tariffs on Canada and Mexico may—or may not—kick in next week
• Core PCE disinflation going a bit slower than originally thought
• Forecasting 175k payrolls and 4.0% unemployment rate for February

A strange mix of January data this week led us to lower our tracking of 1Q annualized real GDP from 2.25% to 1.5%. Even that assumes some bounceback in activity in February and March. The Atlanta Fed’s closely-followed GDPNow tracker went all the way down to -1.5%, underscoring the downside risk to our newly lowered forecast. For now we are not inclined to hit the panic button. Much like 1Q22, when GDP contracted at a 1.0% rate, we are not seeing confirmatory signals in the labor market data. This week’s claims data showed a spike but there are some reasons it could be an aberration, and we expect next week’s February jobs report to indicate another strong month of job creation.

This isn’t to say that there aren’t worrying economic growth developments. President Trump has indicated that next Tuesday he will impose tariffs of 25% on Canadian and Mexican imports, and will raise Chinese tariffs by another 10%-points. As with other Trump tariff announcements so far, it’s hard to know if this is a bluff or a genuine turn in policy. However, if it were to be realized it would create a significant new head-wind to economic activity, as well as an upside support to consumer prices. Alternatively, at least the Canada and Mexico tariffs could get delayed again until after April 1, when the administration is set to publish the trade report the Trump asked for in his first day executive order. We discussed the conflicting objectives for tariffs in a note this week, as well as how much proposed policies could raise the overall tariff rate.

The uncertainty around trade policy may already be affecting business and household sentiment. This week saw a number of regional manufacturing and service surveys released, almost all of which showed a deterioration in expectations about the economic outlook. Moreover, expectations for capital outlays in all five regional Fed manufacturing surveys declined in January a concerning development insofar as this is where one would expect to see policy uncertainty most directly weigh on activity. Consumer confidence has also softened lately, some of which appears to be related to an increase in expectations for consumer prices.

The news this week led markets to price in just shy of three cuts by the end of the year, about one more cut than it was pricing this time last week. Next Friday Fed Chair Powell will speak at the 2025 US Monetary Policy Forum; this speech will come four hours after the February jobs report and just ahead of the communications blackout before the March FOMC meeting.

The jobs report will be the data highlight of next week (unlike the normal schedule, this month the JOLTS report will come out the week after the nonfarm payroll report. We expect employment to increase 175k, which is near the trend of the last six months, and that the unemployment rate will hold at 4.0%. Slightly warmer weather in February should act as a support to payrolls, with some possible offsets from slightly lower federal government employment and lower immigrant labor supply. We will also get the ISM surveys, with the services survey being of particular interest. The services PMI has weakened materially between December and February, and we will want to see if the ISM sends a similar signal.

A peculiar month

The downward revision to our 1Q GDP tracking owes to two reports: a weak January real consumption figure, and a sharp widening in the January merchandise trade deficit. Beginning with the consumption report, real consumer spending fell 0.5% in January (Figure 2), a bigger decline than expected and the largest monthly decline since early 2021. Almost half of the decline was accounted for by a slump in motor vehicle purchases. However, softness was fairly widespread, including in real service spending, which increased by only 0.1%, the weakest gain in over a year.

Channel checks by our industry analysts indicate vehicles sales should rebound in February. Moreover, the Chase card data point to modest growth in retail sales in February. However, even with a better February and March, the math on 1Q real consumer spending growth got a lot more challenging after the January spending figures, and we now expect consumer outlays will only expand at a 1.3% annual rate this quarter. Keep in mind though that this comes on the back of rapid 4.2% growth the prior quarter, and some of the weakness could be payback for earlier months.

Turning to trade, advance figures showed the deficit widened by $31bn in January (Figure 3), which when scaled by imports is the largest jump in four decades. Imports surged 12%. Certainly part of the story here should be a classic case of front-loading surged, up 63% annualized over the last three months.

However, we also think there is a possibility that a good chunk of the import surge in in January stemmed from imports of nonmonetary gold (which is gold not used for official reserves), as traders have been moving it in advance of possible tariffs. While we don’t know the exact product composition of January imports, there was a record 33% jump in industrial supplies, which is on top of a 19% rise in December (Figure 4). More detailed December data show much of the rise that month coming from gold. Because most gold isn’t used in production processes, the BEA strips it out of import statistics when calculating GDP. Consequently, while the headline Census figures at first blush suggest the annualized trade balance could widen by more than $200bn in 1Q, the actual figure could be much less. We will be able to better understand what is happening here next week when we receive the full January trade report, which adds the product detail.

Moderately sticky inflation, steady growth in wage income

As long as the labor market continues to hold in the Fed will likely remain patient in pursuing further rate cuts, especially with policy uncertainty from the White House, a recent pop in some measures of inflation expectations, and the gradual pace of disinflation. Per this last point, the core PCE deflator was revised up slightly in 4Q and then rose 0.3%m/m (0.28%) in January. The over-year-ago rate dropped from 2.9% to 2.6%, though it almost rounded up to 2.7%; prior to the 4Q revision we projected that the rate would be 2.5%. The over-year-ago rate is now no lower than it was in the middle of last year. Even as consumer spending slumped in January the pace of underlying wage growth held quite steady at 0.4%m/m. Disposable personal income spiked 0.9% as Social Security’s annual cost-of-living adjustment took effect and some forms of non-wage also popped. That meant the saving rate, which fell 2%-pts between January and December last year, reversed half that decline as it rose back to 4.6%.

Watch out for claims

Last week initial jobless claims surged 22k to 242k, and are now the highest since early December. If this is sustained it would point to a slowdown in employment growth, but we are cautious about jumping to conclusions. Weather was particularly cold last week, there was a snowstorm, and the week contained the Presidents’ Day holiday, which could cause seasonal adjustment challenges. Also, claims around the DC region were not a contributor, so this doesn’t look related to job losses for government contractors. Our expectation is that claims will start to fall back in the weeks ahead, though if they do not it will raise concerns about the health of the labor market. As a reminder, the headline claims figures do not cover federal workers, which are reported in a separate line. Federal worker claims will likely increase in coming weeks in response to recent layoffs.

Picking up nickels and dimes

While none of the executive orders mention revenues from tariffs, Trump has explicitly argued that import duties could replace income taxes, and has called out to the late-nineteenth century as a model. Meanwhile, some Republicans have sug- gested tariff revenue might help smooth what is increasingly looking like a challenging negotiation over budget reconciliation by their party in Congress this year.

However, customs duties account for only a small portion of receipts in the federal budget: just $77bn for FY24, or less than 1.6% of the $4.9tn in total receipts of the federal government last year. And that is up from 1.0% of receipts in 2017, before the Trump I tariffs. Yet some estimates suggest that the Trump II tariff proposals could raise enough money to make a sizable dent in projected deficits over the next decade, and potentially offset some of the lost revenue from extending or adding tax cuts. Table 1 lists a number of external estimates for the potential revenue that could be raised.

A few key takeaways from Table 1: first, estimates tend to be fairly similar for similar types of tariff proposals. Second, unsurprisingly, universal tariffs tend to raise more revenue than targeted tariffs, even at lower tariff rates. That reflects both a broader base of imports and less opportunity for substitution to non-tariffed goods. Given these are nominal estimates and tariff revenue is assumed to increase with the size of the economy over time, the near-term annual increase in revenue might run from the low tens of billions to somewhat more than an additional $100bn or so. These amounts are still only in the low single digits of percentage shares of overall annual fiscal receipts.

Third, estimated revenue tends to be lower under “dynamic scoring,” which attempts to take into account how tariff impacts feed back into GDP growth and inflation. This is an important concept. However, once a broader set of likely feedback effects are considered, larger reductions in collected revenue seem plausible. As we have noted previously, trade policy uncertainty alone can be a significant drag on growth. With markets still viewing tariff threats as merely a negotiating tool, there could be large risk off moves in financial conditions if implemented. Moreover, tariffs of this size may have even larger, nonlinear impacts on behavior than modelled. If the basis for existing trade relationships and global supply chains is called into question, outsized impacts on employment and capital investment may result. The pandemic-era disruptions to supply chains give a historical example of the larger-than-anticipated transmission of such shocks. Retaliation would likely further reduce revenues.

Additionally, as tariffs rise, they discourage imports and the amount collected per additional percentage point increase in the tariff rate slows. At some point tariffs succeed in discouraging imports so much at there is no additional revenue col- lected—and, beyond that point, revenue actually declines. This nonlinear relationship is a type of Laffer Curve that is often referenced in tax policy discussions—and, of course, tariffs are simply a tax on imports.

This is not just a theoretical curiosity: Figure 4 illustrates the empirical relationship between tariff rates and tariff revenues for the escalation during the Trump I trade war with China. Looking at quarterly data since 2018, total tariff revenue collected from China slowed sharply near the current average tariff rate. Consequently, the most recent 10% additional tariff is likely to yield much smaller marginal revenue gains than earlier increases. Imports from China peaked at around $540 billion in late 2018 as the trade war intensified. After a temporary surge during the pandemic, imports began to decline and are currently around $430 billion. While it is not certain that the maximum has already been reached, the slope of the curve from a polynomial fit suggests we may be close to it. This is another reason to be skeptical of the revenue projections in Table 1, especially for scenarios with 60% China tariffs.

Import substitution and transshipments of Chinese exports through other countries may help explain why the empirical Laffer Curve for imports from China looks to have peaked a relatively modest tariff rates given the threat of much higher rates. But these considerations just serve to increase the attractiveness of a broad-based, universal tariff regime—in addition to those motivations discussed earlier.

A final reason to be skeptical of revenue estimates is that the net amount may be much less as both the spillover effects of US tariffs and the more direct effect of trading partners’ retaliation create strong demand for government offsets. One significant unintended consequence of the Trump I trade war was the impact on US farmers, as Chinese retaliation diverted demand for US agriculture exports to other producers, such as Brazil. In fact, estimates suggest that fully 92% of money raised by the tariffs on China went to farmers in “relief” payments. That even larger tariffs won’t generate correspondingly larger demands for relief seems rather unlikely, in our view.

Finally, note that to succeed as a significant source of revenue, tariffs have to fail on almost every other objective listed previously. Relaxing tariffs as part of a successful negotiation reduces revenue. Reducing imports or reshoring production reduces revenue. Creating carve-outs to exempt certain goods reduces revenue. Relief payments to adversely impacted US producers reduce revenue.

Meanwhile, the revenue that is collected will come through either lower US corporate profit margins or higher US consumer prices (or both). The former is likely to undermine hiring and investment; the latter will weigh on consumer spending, even if inflation turns out to be “transitory.” Some have suggested ramping up tariffs via multiple rounds of smaller increases in the rate, but that just risks prolonging the price adjustment and unanchoring inflation expectations—which some recent survey data suggest could already have begun.

Who’s keeping score?

One important caveat to the above revenue discussion is that in the process of marking up a budget reconciliation bill, Congressional Republicans need only reach agreement on what amount of tariff revenue they expect to be generated to support their tax cut and other fiscal proposals. There is nothing in the reconciliation process that forces full offsets to higher deficits (that, say, result from extending the TCJA or adding additional tax expenditures) within a 10-year budget window.

Outside the 10-year window, the reconciliation process requires no additional increases in deficits relative to the CBO’s budget baseline. That baseline has been based on “current law,” meaning estimates of tariff revenue would be included only if they are enacted legislatively by Congress. If instead they are implemented via the Executive Branch, they would not be part of the baseline. But more formally, CBO’s role here is advisory, and as Congress sets the rules for itself, it can decide to instead rely on outside estimates of the budgetary impacts of various policies—as was the case previously with the TCJA and ACA amendments. In the end, simple majorities in both the House and Senate have to be convinced to vote for a budget reconciliation bill.

Whether those assumptions about tariff revenue (as well as any additional assumptions about spending cuts, dynamic impacts of tax cuts on boosting GDP growth, etc.) will be convincing to voters by the midterm elections, or to markets in the meantime, remains to be seen.

One big, beautiful tariff

The overarching conclusion from our analysis is that there are strong incentives for the Trump administration to favor some sort of universal tariff policy (such as its “reciprocal” plan). That does not eliminate many of the trade-offs cited previously, and as noted it carries additional risks for inflation and retaliation. The counterfactual belief that price increases would be borne by foreign exporters might lead policymakers to risk a more aggressive stance; it is unclear whether a temporary increase in US inflation would be viewed as an acceptable trade-off to achieve other goals. Broad-based and persistent tariffs also offer a stronger implicit subsidy for domestic goods-producing firms, potentially slowing the secular decline in manufacturing at the cost of lower productivity.

On the retaliation front, being a large economy with diversified import sources could provide some leverage and help mitigate adverse effects. This might be especially true if trading partners heavily depend on the US as an export market. Empirically, this seems to be the case. The top five exporters to the U.S. (i.e., Mexico, Canada, China, Japan, and Germany) account for about half of the overall goods imports, and the top 20 exporters account for around 80% (Figure 5). In contrast, some of these partners, such as Mexico and Canada, send more than 75% of their total merchandise exports to the US, giving the US extraordinary leverage. Viewed from this perspective, it may not be surprising that Canada and Mexico were the first targets of significant tariff increases, even before China.

A cautious Fed to lean dovish

US consumer expectations have also stepped down in both the Michigan and Conference Board survey. Expectations can be volatile and tend to follow rather than lead consumer spending. But it is notable that both surveys show a significant rise in near term inflation expectations. This increase was not present during the 2018-19 trade shock but was a feature of the 2021-23 pandemic recovery. The rise in expectations over this later period played a role in the Fed tightening cycle, contributing to the accelerated the pace of 2H22 hikes.

We believe that tariffs will keep US inflation close to 3% and keep the Fed cautious in considering further rate cuts. However, the risk of a substantial acceleration in US inflation this year (to above 3.5%) is viewed as modest in an environment in which dollar strength and wage gains are moderating. We believe the Fed will also display a heightened sensitivity to downside growth risks. While a faltering in labor demand is not in our forecast, the Fed will not hesitate to cut policy rates with elevated inflation if it sees a material threat to the expansion. Risks to the expansion are getting priced into the market as the US 2-year real yields declined 45bp this month to their lowest level since August 2022 even as consumer inflation expectations have risen.

ECB easing path looks easy

The Euro area has struggled to grow but solid private sector income gains and healthy balance sheets have created resilience to negative shocks. A pickup in the PMI in recent months bolsters the case for modest growth and this should be reinforced by a rate cut at next week’s ECB meeting. All indications point to a deceleration in wage and price inflation to support further cuts. Start-of-the-year price resetting looks modest after a downward revision to seasonally adjusted January core HICP and national reports point to a soft February print. Core inflation has been running effectively at a 2%ar since last summer. While the ECB has some concern about a tariff war generating upward inflation pressures, it is more concerned about the adverse growth implications. Pressure of higher defense spending may also be slow-moving in terms of boosting domestic demand. In whole, risks to our 1.75% terminal rate forecast are skewed downwards, with questions mainly about the speed of delivery.

“Braking” news: Bundestag challenges

While the risks of a more divided and extreme election out turn in Germany were avoided, the incoming government faces considerable challenges. Rebooting the economy, upgrading infrastructure, financing a green transition, and boosting defense spending will be difficult for the narrowly held coalition government. The centrist parties have a majority but face a fiscal challenge as the current debt brake creates pressure to cut the deficit. Reform or a new special fund is essential but needs a two thirds majority in the Bundestag. This will require Left Party support, which is uncertain if used to finance defense and would come with other conditions. Centrist parties are now exploring a constitutional change before the new Bundestag sits, but this raises legal and political questions. Efforts are also ongoing at the EU level, which Germany could end up supporting if domestic constraints cannot be resolved.

Japan’s reflation blossoms

Incoming data show Japan’s reflation is on track. Following last week’s strong January nationwide CPI, this week’s February Tokyo CPI showed inflationary pressures in previously stable categories, such as rents and transportation costs. We have expected this shift, along with the pass-through of strong wage increases into prices, to promote further BoJ rate hikes. Market pricing has begun to better align with our forecast that the BoJ will raise rates to a neutral level or even higher (Figure 4). We see very little cyclical support for Japan’s exports and outputs on the near-term horizon, as evidenced by manufacturing PMI and IP at the start of the year. However, we expect domestic demand to remain resilient even in the face of this external weakness, providing cover for continued monetary policy normalization.

China: Limited response to US tariff hikes

China will begin releasing January-February activity data after the Lunar New Year lull next week. We expect the PMIs to move back up to 50, while a front-loading of demand ahead of tariffs should drive a solid (4.6%oya) export gain. Overshadowing the expected constructive dataflow is the prospect of an additional 10%-point US tariff hike, due to take effect just one day before China’s annual NPC meeting starts. Our forecast assumes further US action and a fading of recent policy supports will slow China’s growth below a 3%ar in coming quarters. We do not expect any major shifts from next week’s NPC meeting with focus likely to remain on boosting support for consumers (400-600 billion yuan vs. 300bn last year), including a larger trade-in subsidy program and increases in basic pensions and healthcare). Building downside risks to the government’s 5% growth target could prompt greater policy stimulus (through an additional quota of government bonds, as happened in the past two years) but probably not until later this year.

EM Asia CBs: Easing, one way or another

Both the Bank of Korea and the Bank of Thailand delivered 25bp rate cuts this week, citing increasing growth concerns from a US tariff shock. The BoK also revised down the growth forecast but delivered a cautious message on financial stability risks. Despite this caution we expect the BoK to deliver a further 75bp in easing this year. We also look for 75bp of further easing from Thailand which is more exposed to tariff hikes than Korea and faces structural issues related to intense competition from China. Being more constrained by tighter external financial conditions and a weaker FX, the Bank of Philippines has been resorting to RRR cuts to support the economy; last week it announced a 200bp RRR cut which is expected to inject PHP320bn (1.1% of GDP) of liquidity into the banking system.

Türkiye exits shallow recession

Boosted by fiscal and monetary policy easing, Türkiye’s GDP jumped a larger-than-expected 6.9% annualized in 4Q24 after two quarters of contractions. The rebound was driven primarily by domestic demand and leading indicators point to continued solid gains this quarter. A growth pickup stimulus could reignite concerns over external imbalances even as the widening in the current account deficit to date has been modest (0.8% of GDP in 2024). We expect the CBRT to respond by tightening credit conditions in a bid to rein in domestic demand strength but still see upside risks to our 2.8% GDP growth forecast for this year.

LATAM lament for Brazil and Mexico GDP

We see downside risks to GDP growth in LATAM’s two largest economies in Latin America. In Brazil, our forecast calls for a solid 1H25 driven by robust gains in agriculture and consumption. Strong imports and auto production in January add to the story of firming demand. However, consumer and business confidence have declined, and real income growth is set to weaken in 1Q amid stalling job growth and lower fiscal transfers. This suggests an expected 2H25 slowdown could arrive earlier than anticipated. Meanwhile, Mexico GDP contracted 2.5%ar in 4Q and our nowcaster shows no upward momentum at the start of the year. Our survey-based tracker and this week’s January manufacturing exports bounce was comforting but we still see downside risks to our 2.1%ar GDP growth forecast for this quarter. Looking past this quarter, we turn to next week’s business surveys to gauge the degree to which business sentiment was damaged by tariff threats.

The United States Feb 24-28th

• Tariff threats appear to be rattling business confidence…
• …and contributing to an increase in household inflation expectations
• The Fed is firmly on hold as it warily watches trade and other policy developments
• Up next week: 0.2% rise in January core PCE to ease the year-ago measure down to 2.5%

Over the last few weeks this essay has led with the various tariff threats coming out of the administration. Those threats continued this week. One side effect of this saber-rattling may have manifested itself in this week’s data as a number of business surveys showed a deterioration in the outlook, in some cases directly attributed to the brewing trade conflict. Among these, perhaps the most notable was the flash February services PMI, which slipped back below 50 for the first time in just over two years (Figure 1). And while its manufacturing counterpart improved for the fifth straight month, the press release suggested this stemmed from temporary front-loading ahead of anticipated tariffs. However, business concern was also spreading beyond tariffs, with the respondents in the service survey beginning to highlight federal spending cuts also as a risk. We discussed some of the recent moves to curtail government spending in this week’s Focus.

Trade concerns were also likely a factor in the recent rise in household inflation expectations, as measured by the University of Michigan’s survey of consumers. Longer-term inflation expectations jumped to 3.5% in February, near the highs of the COVID inflation surge after adjusting for recent changes in the survey method. Underscoring the idea that tariffs are an adverse supply shock—bad for both inflation and growth—consumer perceptions of buying conditions also deteriorated markedly last month (Figure 2), after having briefly spiked as people said they were front-loading purchases.

Elevated inflation expectations should also reinforce the Fed’s extended pause in its rate-cutting campaign. The minutes to the January meeting suggest the FOMC was already minded to abide by a wait-and-see strategy even three weeks ago. At this point, forecasts are firmly entrenched that the Fed will be on hold at its next meeting in mid-March. Beyond pure expectations of inflation, there are also some signs that inflation pressures have been building in the manufacturing sector, as indicated by a recent pickup in price indexes in the regional Fed surveys and the latest PMI (Figure 3). Fortunately this is not passing through to the services sector, with the services PMI output price measure hitting a new cycle low.

In Washington, the Senate this week adopted a budget plan that would boost defense and border security spending by $340 billion over 10 years. This plan would leave for future legislation extending the expiring 2017 tax cuts. This is a different approach from the one adopted by the House last week, which would increase defense and border spending, cut other spending, and extend the tax cuts all in one piece of legislation. While President Trump has thrown his support behind the House effort, the Senate approach may be able to achieve quicker agreement on the first bill described above before turning to the more challenging effort of achieving Republican consensus around tax and spending cuts. Looming not too far in the distance are two other fiscal deadlines: an extension of current government funding to avoid a shutdown (March 14) and an extension or suspension of the debt ceiling (likely later this summer).

Next week’s data calendar will feature a number of releases that will help flesh out the trajectory of 1Q real GDP growth, including the January data on core PCE inflation and real consumption, durable goods, trade, and inventories. For core PCE we look for a 0.24%m/m increase, which if realized would take the over-year-ago rate down to 2.5% from 2.8%. Real consumer spending should be soft (-0.1%) given what we know about retail sales and auto sales, though were spending to subsequently return to its average pace over the last year then consumption for the quarter can still increase more than 2%.

Trade wars: still a phantom menace?

Even as tariff developments filter into business and consumer confidence, the drumbeat of new developments continues. This week President Trump suggested that new 25% tariffs would be coming on autos, semiconductors, and pharmaceuticals shortly after the start of April, and on Friday he signed a memorandum directing the USTR to explore methods to hit back at countries that levy digital service taxes. Unlike the tariffs proposed for specific goods, it does not set a deadline. However, it does suggest that the Trump administration continues to accumulate cudgels to use in negotiation with a range of trading partners. As focus in these tariff negotiations looks set to shift from border issues to more structural economic issues like reducing persistent trade deficits, protecting certain American industries, and forcing other countries to change their own tax and regulatory policies, the risk that some tariffs may actually be put in place—and potentially for some time—appears to be drifting up. Recent declines in sentiment measures and a late-week sell-off in equity markets suggests growing recognition of these risks.

Fed: not going anywhere for awhile

This week Fed speakers and the FOMC minutes reiterated the message from Chair Powell’s January press conference and recent Congressional testimony that the Committee is likely to remain on hold for some time, even as it retains its easing bias. The basis for keeping rates steady is based on both the trajectory of the incoming economic data, with labor markets having done well and inflation still above target, as well as the uncertain effects of government policy, particularly stemming from upside risks to inflation from trade and immigration policy.

Where policy uncertainty appeared to potentially lead to earlier changes in Fed policy is on the balance sheet, as noted in the January minutes. “Various” participants expressed support for the idea of slowing or pausing QT over coming months, given the potential for significant swings in reserves around the approaching debt ceiling negotiations. That Powell did not address this issue up at the January press conference suggests the Fed may not be thinking it will be appropriate to consider such action at the next meeting in March. But communication at that meeting could set market expectations for how much additional QT is likely this year. Even as the FOMC continues to express concerns about inflation running above target and attendant upside risks the easing bias is still in place. That is because the “vast majority” of participants continue to believe that the current level of rates remains restrictive, and is thus appropriately calibrated to gradually return inflation to target. On the other hand, were weakness to re-appear in the labor market the committee could resume cutting.

Builders also nervous about tariffs

The latest batch of housing market data, covering homebuilder sentiment, starts, and existing home sales, was largely a continuation of recent trends that have had starts running at decent levels while existing home sales struggle to get any traction. One item that stood out was the change in homebuilder sentiment stemming from the recent tariffs, both the implemented ones on China and the threatened ones on Canada and Mexico. The overall NAHB housing market index fell from 47 to 42 in February, but the press release noted that responses collected before Canada and Mexico tariff pause was announced were 8pts worse than responses collected after the pause. In some sense this is a surprisingly large effect given the small share of imported material costs in overall construction.

According to NAHB estimates, 7% of materials in residential construction are imported, with that number materially higher for some types of products. But their estimate of higher costs from tariffs on China, Canada, and Mexico is under $4bn, whereas they estimated that the total materials bill in 2023 was $184bn. Moreover, Census Bureau estimates of new residential construction spending, which include labor costs, were ~$550bn last year. Turning to the hard data, new single-family housing starts fell 8% in January. Especially considering that cold weather may have taken a toll, things don’t look too bad, with starts still running near the average for last year. But there isn’t much upside, with high rates weighing on demand, inventories becoming a bit elevated, and likely slower population growth ahead. Meanwhile, existing home sales dipped 5% on the back of the rebound in mortgage rates since last September and sales will likely continue to remain low in coming months absent a pronounced drop in rates.

DOGE: There’s smoke, but is there fire?

In 1993 the term “National Performance Review”appeared in 20 articles in The New York Times. If you had forgotten this was the government efficiency effort spearheaded by then-vice president Al Gore you are probably not alone. In contrast, the term “Department of Government Efficiency” has appeared in more than 20 New York Times articles in the last two days alone. This attention level gives the impression that spending cuts, either realized or imminent, are dramatic. The reality so far is more mundane, though that is to be expected since implementing real efficiency improvements within a month is challenging.

DOGE has started reporting cost reductions on its website, where it claims $55bn in savings, including from personnel cuts, contract and grant adjustments, and other measures. While there is no full breakdown of how much each measure has contributed, the largest components are probably personnel cuts and the USAID shutdown, and perhaps NIH funding. For personnel cuts, we assume DOGE will include the 75k workers who have taken deferred resignation, though as we discussed previously some of this likely represents normal turnover. In terms of other layoffs, news articles had suggested that all 200k+ probationary employees were being fired immediately, though so far actual agency announcements have been less than 20k, with another roughly 10k workers on leave at agencies slated for closure (USAID and CFPB) or due to other miscellaneous layoffs.

Using recent CBO figures for federal employee compensation (wages and benefits) and inflation-adjusting them to the present we estimate annual average compensation of a federal employee was around $130k in FY24. Assuming a cut of 100k workers at this compensation rate would yield annual savings of $13bn, with less than $1bn of this for USAID federal employees. The USAID shutdown also has likely caused at least another 5k job losses among domestic contractors, although some estimates are that eventual job losses may be as high as 50k. Total savings reported for USAID could be over $35bn. Based on data from ForeignAssistance.gov, USAID had obligations of $42.5bn in FY23, and while full-year 2024 data are not yet available the partial year total was $35.4bn. The caveat here is that some of this spending may continue under the Department of State, and the shutdown of USAID is still tied up in court cases. For NIH meanwhile, “indirect costs” are now capped at 15% of total grants. In FY23 NIH grants had direct costs of $26bn and indirect costs of $9bn; if direct costs didn’t change then indirect costs would have to fall $4-5bn to hit the 15% rate, though recipients might try to adjust accounting to claim more direct costs.

In terms of detail on other savings, the DOGE website lists individual contracts and real estate leases that have been cancelled with a total reported savings of $7.2bn as of Feb 20, of which $2.0bn is outside of USAID. DOGE notes that some contract cancellations are reported with a lag and are not yet posted. The reported savings look to be calculated as potential award amount less obligated amount. In some cases that could overstate savings, for example the largest reported USAID contract with reported savings of $655mn appears to refer to be a type of multi-vendor contract under which actual awards obligated are stored as separate records and the full potential amount may never be used. In other cases this could understate savings, such as if most funds have already been obligated but the contract would have normally been renewed to support fresh obligations. In this latter case, the total contract value of $12.8bn ($4.3bn ex USAID) could be more useful, though for both total contract value and unobligated spending capacity the amounts in question can span more than a year, so the annualized value should be smaller.

Summing up, outright layoffs (as opposed to deferred resignations or administrative leave) are probably under 30k to date, which is small in an economy that sees 1.7mn layoffs per month, and savings outside of USAID are also small when compared to federal government expenditures of $7tr. Of course more substantive adjustments could be coming, affecting both federal personnel and government contractors. In terms of how to think about contractors, NYU professor Paul Light has estimated there were 5.2mn contractors in FY23, along with another 2.3mn workers supported by grants. Total contract spending was ~$760bn in FY23, or about $145k per contractor, which could be one method of sizing the labor market impact of lower contract spending. On a final note, there had been some concern that DOGE would unilaterally pause or cancel wide swaths of government spending. The Daily Treasury Statement does not point to this outcome so far, though it admittedly would not capture the termination of contracts with future payment dates. Comparing non-interest cash outlays in the current calendar year to last year, there has been no obvious change (Figure 1).

The United States Feb 17-21st

• Trump announced plans for both limited steel and aluminum tariffs, potentially broad reciprocal tariffs, and auto tariffs

• The CPI surprised to the upside, though the PCE implications were not as bad
• The House Budget Committee released instructions for the reconciliation bill that point toward TCJA extension but limited other tax breaks
• Next week brings first February business surveys

As is par for the course these days, this week brought a fresh batch of tariff news. The first, and more limited step, was announcing 25% tariffs on steel and aluminum, which are set to go into effect on March 12. Imports of these products were close to $108bn in 2024 (Table 1) with tariffs averaging 4.3% on that trade. Consequently the aggregate impact of a 25% tariff on GDP is not that large, though it could clearly be important for those two industries, especially aluminum where imports are a larger share of the market. In both cases, Canada, China, and Mexico are the largest providers, especially Canada in the case of aluminum.

The second, broader announcement was an order to begin figuring out how to set reciprocal tariffs on any trading partner. The size of that tariff would be based not only on the tariff rates that trading partners charge on US exports, but also on a wide range of non-tariff barriers, allowing broad latitude to raise tariffs on effectively any trading partner. In contrast to WTO rulings, the president considers the border-adjustment taxes associated with value-added taxes (VATs) as tariffs. This could vastly increase the scope of reciprocal tariffs, as most countries have significant VATs. Separately, late on Friday the president mentioned he would be bringing new auto tariffs around April 2, although he provided no further details, and it was unclear if these would apply to imports from Canada and Mexico.

The White House announcement on reciprocal tariffs stated that the “lack of reciprocity is one source of America’s large and persistent annual trade deficit in goods.” The current account deficit, which in the case of the US is dominated by the trade deficit in goods, is arithmetically linked to the short-fall of national savings relative to national investment. A large part of that shortfall is the federal budget deficit, which counts as negative savings in national income accounting. This week’s news coming from the other end of Pennsylvania Avenue suggests there is little reason to expect that drag on national savings—and on the current account deficit—to reverse anytime soon.

Also this week the House Budget Committee released instructions for the reconciliation bill that will be used to extend the Tax Cuts and Jobs Act, among other things. For more on the outlook for taxes see our note here. The instructions provide for a $4.5tr gross increase in the deficit for fiscal years 2025-2034 stemming from policy changes by the Ways and Means Committee, whose remit includes taxes. The CRFB estimates that extending the personal and estate tax provisions of TCJA during this budget window would raise the deficit by $4tr, so these instructions imply there will be little in the way of further tax cuts beyond TCJA extension. The instructions, which are not binding, also call for an additional $200bn in defense and law enforcement spending, coupled with $1.5tr of cuts elsewhere. The combination of these actions would increase the deficit by $3.3tr, though the debt limit would be raised by $4tr to account for additional interest costs (Table 2).

Also, as noted by the CRFB, the budget timeline includes one year in which TCJA does not need to be extended (2025), and because tax policy is enacted on a calendar year basis the TCJA extension would need to sunset at the end of 2033. Thus the incremental reduction in tax revenue from TCJA extension is only for a eight-year window, and they estimate that a full 10-year extension would cost $5.5-6tr.

Elsewhere on the cost-cutting front the government moved to terminate federal employees who have probationary status and who had not accepted deferred resignation. Probationary employees usually have one year of service or less, lack civil service protections, and would usually not be eligible for severance pay. According to the OPM FedScore database, there were 216,000 federal employees as of May 2024 with under one year of service. However, information about these layoffs is coming via news reports, with some saying the government has delivered clear orders, and others saying it is simply guidance. And while there have been reports of specific terminations, the on-the-ground application is varied, such as at the Department of Veterans Affairs which said it dismissed 1,000 employees but also said more than 43,000 probationary employees were exempt. In any case, we will watch the weekly jobless claims data for federal employees to track what may be happening.

Inflation panic… or not?

On the data front the big news this week was the January CPI report. The headline figure was up 0.5%, lifting the year-ago reading to 3.0%. While egg-flation is getting the headlines, the more surprising development was in the ex-food and energy core category, which increased a firmer-than-expected 0.4% (Figure 1). The high-side reading for January was seen as particularly important because of an early-in-the-year burst of inflation last year. If this were a recurring pattern, it would suggest that the slowdown in core PCE inflation in 2H24, to just 2.3%, was a temporary development, and that the over-year-ago rate of 2.8% is a better guide to underlying inflation.

However, much of the strength in the core CPI was in categories like used vehicles and vehicle insurance, which either have smaller weights or use different concepts in the Fed’s preferred PCE price gauge. Based on this week’s CPI and PPI reports, we estimate the core PCE price measure increased 0.24% last month, which would bring the year-ago reading down to 2.5%, the lowest since early 2021. That would be about half the size of the core PCE increase last January (Figure 1). Consequently, we continue to see an avenue for inflation to moderate slightly this year, and the Fed should remain patient while still having an opportunity to cut later in the year. On this last point, markets priced just one cut immediately following the CPI report, but as the milder core PCE implications became clearer after the PPI report, and as retail sales then disappointed expectations moved back to 40bps, slightly greater than they had been last Friday (Figure 2).

Further on Fed policy developments, Chair Powell went to Capitol Hill this week to deliver semiannual testimony on the Monetary Policy Report. He was generally upbeat on the economy, noting the above-mentioned cooling in inflation alongside healthy labor market developments. He also indicated the Fed is in no rush to cut rates, consistent with the market pricing. This wait-and-see attitude is likely being reinforced by uncertainties around trade policy, though of course Powell was very cautious in discussing this topic.

Stop the spend

The activity data were mixed this week, with a disappointment in January retail sales which fell 0.9%. The “control” measure fell a similar 0.8%. The report didn’t contain any telltale hints that weather was a big drag; the normally weather-immune nonstore category (mostly internet retailers) stumbled last month while restaurant sales, which tend to be more weather-sensitive, actually posted a nice increase. While the January retail report was a genuine disappointment we maintain a constructive view on the consumer, in large part due to the strength of the labor market. This week’s jobless claims report suggests no imminent deterioration in job market conditions. Given the robust trajectory of spending late last quarter, it should still be possible for real consumer spending to rise over 2%q/q saar in 1Q. There may have been some front-loading of sales in 4Q in anticipation of tariffs, and if we average 4Q consumer spending of 4.2% with an assumption of growth around 2% in 1Q then the average pace of ~3% is right in line with the prior year.

US taxes: A $4 trillion tab, tip not included

• Congress is likely to extend virtually all the expiring provisions of the 2017 Tax Cuts and Jobs Act (TCJA)
• That will increase deficit and debt projections relative to CBO’s current baseline
• Recent House budget resolution instructions leave little fiscal space for anything beyond TCJA extension
• Extension of TCJA alone won’t create new fiscal thrust, but would avert a fiscal cliff if no action were taken
• Some other proposed tax breaks would shrink the tax base and encourage wasteful tax avoidance strategies

Since Inauguration Day almost all the interesting policy May, 2024. 2026 values are identical to pre-TCJA values, except for automatic inflation adjustments to personal action in DC has come from the executive branch. At some exemptions and standard deductions. Dollar values for standard deductions and the personal exemption for 2026 are CBO projections that could change if inflation turns out differently than forecast. Note that some provisions of the income point this year, however, Congress will have to get back in the tax, such as standard deductions, take on different values for “heads of household” (mostly single parents) and taxpayers game of setting policy. That is because several important, and popular, provisions of the personal income and estate tax code are set to expire at year-end. These provisions were put in place by the 2017 Tax Cuts and Jobs Act (TCJA), though subject to a sunset condition. Congress will also likely consider several tax proposals that the president mentioned during his campaign, such as exempting tipped income or overtime pay.

Given Republican control of both chambers of Congress, we believe it is likely that effectively all the expiring personal and estate tax provisions of the TCJA will be extended. (In a separate note we plan to discuss the few expired and expiring business tax provisions of the TCJA). We think the president’s added campaign tax proposals will have a harder time passing, given deficit concerns. Moreover, the recent House budget resolution doesn’t appear to have much room for taking on board these proposals. That said, we think it’s possible we could see some token concessions, such as exempting the first few thousand dollars of tipped income. In this note we discuss three economic implications of policy actions regarding personal income taxes: (i) what they do for aggregate demand, or short-run GDP growth consequences (ii) how they affect the deficit and debt sustainability and (iii) how they alter incentives for various economic choices.

How we got here

The TCJA came into effect at the start of 2018. It built on the philosophy of broadening the tax base and lowering tax rates. Base broadening was achieved through two means. First, some deductions and exemptions were directly trimmed. For example, the cap on the mortgage interest deduction was lowered and a new cap on state and local tax (SALT) deductions was introduced. Second, by almost doubling the standard deduction it reduced the number of tax returns with itemized deductions, thereby neutralizing some other deductions that were politically more difficult to cap or eliminate. Table 1 and 2 further catalog the key base broadening provisions and lower tax rates, comparing the law as it currently stands under the TCJA to our estimate of what would occur in 2026 were the law to lapse, in which case many tax provisions would revert to pre-TCJA rules.

Not everything was about base broadening and rate lowering. For example, the 20% deduction of eligible pass-through business income didn’t have much grounding in economic principles and was apparently more geared toward vote-getting. Another realpolitik consideration was the constraints imposed by the Byrd rule. The TCJA passed under reconciliation, meaning in brief it passed with a simple majority in the Senate. However, to comply with the Byrd rule it couldn’t increase deficits beyond a 10-year horizon. Since the cuts to the corporate tax code never sunset, the personal tax cuts had to sunset inside of the 10-year window for the math to work.

Effects on aggregate demand and debt

In the highly unlikely, but not impossible, event that Congress lets the TCJA expire as scheduled, the tax burden on the household sector next year would go up by a little over 1% of GDP. Given standard fiscal multipliers that could reduce aggregate demand by 0.5-1.0% of GDP. But as noted above, we expect Congress will instead extend most or all of TCJA, implying no net change in how tax policy affects aggregate demand and GDP.

By law, the CBO’s baseline budget projections assume current law, which is that the TCJA sunsets on schedule. Even with that unrealistic assumption, debt as a share of GDP is projected to increase over coming years (Figure 1). Adding in the more realistic assumption that the personal and estate tax provisions in TCJA are extended then the cumulative increase in debt over the next 10 years would be close to $4 trillion according to estimates from the Committee for a Responsible Federal Budget, or just under $400bn per year (see the Low estimate for TCJA in Table 3). That would leave the debt/GDP ratio close to 9%-pt higher than under the CBO’s current baseline by 2035 (Figure 1).

The various other tax adjustments that were mentioned during the campaign would lead to further deterioration, adding another $1tr to $5 tr (Table 3).

Any added tariff revenue wouldn’t be factored into the “scoring” of Congressional legislation. Nonetheless, tariff revenue could help the fiscal picture, though likely not enough to offset the deterioration due to extending TCJA. For example, CRFB estimates that a 20% universal tariff would raise around $3 trillion over 10 years, or $1-2 trillion less than extending TCJA loses in revenue.

De minimis exemptions: big business of low-value trade

• De minimis shipments surged post ’18 trade war, reaching ~1.3Bn entries worth $64Bn in 2024

• China accounted for the majority, highlighting the cost (tariff) savings and availability of CBEC infrastructure to exploit it at scale

• We estimate that de minimis exemptions likely resulted in $5Bn-$7Bn uncollected duties, while saving U.S. consumers significantly on some consumer goods

• Since this exemption creates a margin to avoid higher tariffs, it is not surprising that it became the first target of the Trump 2.0 trade policy

• Tariff exemption removal may narrow China’s price discount for de minimis product from ~40% to ~27%

Big things come in small packages, but in cross-border trade, it’s often cheap things that do. This is the idea behind the “de minimis” exemption under U.S. Code Section 321, which allows duty-free import of goods valued at $800 or less per person per day. This exemption was created to facilitate trade and reduce the administrative burden on importers and customs authorities by simplifying the process for low-value shipments. It has been in place since the 1930s, and the threshold was increased from $200 to $800 in 2016 as part of the Trade Facilitation and Trade Enforcement Act of 2015. Shipments that qualify for this exemption not only avoid customs duties but are also exempt from certain administrative fees and the more onerous paperwork required for higher-value shipments.

De minimis imports are back in the spotlight after President Trump issued an executive order imposing a 10% tariff on Chinese imports, excluding them from de minimis treatment due to misuse for illegal fentanyl shipments. Previously, the Biden administration proposed changes to exclude certain tariffed goods from this exemption. The new order removes the exemption for all Chinese goods, indicating potential broader reforms. On Saturday (February 8), the White House paused the enforcement of this new order for now, giving the government some time to find an efficient way of processing these shipments under new tariffs. Lawmakers have been examining this issue for some time, so it should not surprise many observers. It also suggests broader reform of the de minimis exemption is likely.

Here we highlight the growing role of de minimis imports in the U.S., as the 2018 trade war led to increased use of this cost-cutting method, reaching as much as 9% of overall imports in the case for China. While it reduces costs for businesses and consumers, it also lowers tariff revenue by $5-$7Bn a year. We analyze the bilateral trade data between the U.S. and China to show that de minimis exemptions may have helped exporters adjust, partially offsetting the impact of higher tariffs. We also estimate that this cost saving likely kept inflation lower for some consumer goods, possibly taking off as much as 0.13pt for PCE inflation during the first trade war. Given the sense that de minimis exemptions create a loophole around tariff policy, it looks like an obvious policy instrument to target by the White House. De minimis exemptions, at least the scope and the generosity of the threshold, are likely the first victim of the unfolding trade war.

No longer minimal

The value of goods traded under the “de minimis” category is difficult to quantify. Customs and Border Protection (CBP) collects data from Bills of Lading (BOLs), which include the declared value of shipments and other details, but this information is not consistently made public. Researchers P. Fajgelbaum and A. Khandelwal (2024) compiled annual data on de minimis trade, some of which was obtained through FOIA requests, in a recent NBER working paper. We present the total volume of these shipments into the U.S. and the number of entries in Figure 1.

Data for 2024 is estimated based on values available through the third quarter and suggests that the value of shipments, at $64 billion, may have nearly reached its pandemic-era peak with the total number of packages surpassing 1 billion for the first time. CBP reported that as of late fall 2024, about 92% of all cargo entering the U.S. fell into this category, with approximately 90% arriving as cargo on commercial airline flights carrying international mail.

Incentives behind de minimis trade boom

It is striking that de minimis trade was minimal, around $50 million, in 2012 but surged to over $60 billion in nearly a decade. Increasing the threshold likely provided additional incentives to use this channel, but the significant surge occurred after the trade war with China began (Figure 1). Avoiding extra paperwork and customs duties must have incentivized Chinese exporters to ship directly to U.S. consumers, bypassing rising tariffs. CBP reported the breakdown of de minimis shipment volumes by source country for several fiscal years (2018-2021), showing that about two-thirds came from China during this period. Canada and Mexico combined accounted for only about 10%. This large share from China aligns with the timing of additional tariffs imposed on China and the rise of cross-border e-commerce platforms (like Shein and Temu) in China that predominantly rely on direct shipping to consumers abroad.

Note that these figures are largely absent from U.S. trade data. In theory, the harmonized schedule code 9999.95.00 for “Estimated imports of low valued transactions” should include de minimis imports, but data from the U.S. International Trade Commission shows no more than $5 billion in this category (Figure 2). Assuming China still accounts for about two-thirds of the de minimis value shown in Figure 1, only about 13% of the approximately $40 billion in de minimis shipments from China are reported in official U.S. trade statistics.

Fajgelbaum and Khandelwal (2024) estimate that these goods would typically face a 15.3% tariff rate, compared to the 19.3% rate for all goods subject to tariffs from China in 2024. These rates suggest uncollected tariff revenue of $5.4 billion to $6.8 billion in 2024. This also implies that U.S. consumers likely faced lower costs than the rising tariff rates would suggest. In fact, Fajgelbaum and Khandelwal (2024) show that removing de minimis exemptions would disproportionately affect lower-income consumers, who rely more on direct shipments from Chinese e-commerce platforms than their higher-income counterparts. An end of the de minimis shipment tariff exemption also points to an increase in the U.S. effective tariff rate on China’s exports, by around 1.4%pts on our estimates (including latest 10% universal tariff hike).

The murky waters of de minimis trade

Given the measurement issues discussed earlier, one way to assess the relevance of de minimis shipments is to check bilateral trade data. By comparing U.S. import data from China with China’s export data to the U.S., we can gain insights into the unaccounted de minimis shipments in official U.S. trade data. The difference between these figures was significant but relatively stable before 2018 (Figure 3). This relatively stable discrepancy was likely, in part, due to shipping cost (cif vs. fob) and re-exports via Hong Kong or other third-party economies etc. For example, goods that are exported from China to Hong Kong and then to the United States are likely to be counted in U.S. data as imports from China but in China’s data as exports to Hong Kong.

A reversal in the discrepancy emerged since the beginning of the trade war. There has been a widening gap between China’s goods balance under the SAFE’s BoP current account and China Customs’ merchandise trade data (comment from US Treasury). Aside from accounting standard differences, it can be partly explained by domestic exporters’ shipping to their foreign entities, which may have been booked in customs’ exports but not SAFE’s due to no change in ownership. However, even if we use adjusted China exports to the U.S. based on BoP data, the discrepancy reversal post the tariff war remains in place. It could in part attribute to over-reporting by Chinese exporters to obtain export tax rebate subsidies. We did find similar evidence in the context of China–EM Asia trade. Conversely, U.S. importers might be underreporting to partially circumvent rising tariff costs.

An increased reliance on de minimis shipments could also be a factor. This is because the de minimis imports from China are not fully covered in the U.S. imports while their mirrored exports are counted in the Chinese export statistics. Getting into more granular details at the product level highlights how much consumer goods sold over cross-border e-commerce (CBEC) platforms have played a role in the evolving discrepancy.

According to China Customs, more than 90% of the CBEC exports were consumer goods (~92% before 2022, over 97% as of 1H24), mainly apparel, footwear, bags, home textiles, kitchenware, household and office appliances, mobile phone and other digital products etc. Therefore, we identify more than 70 relevant consumer goods items at the HS-6 product level and compare the discrepancy between China reported exports to the U.S. and the U.S. reported imports from China.

Figure 4 shows a consistent discrepancy reversal after the ’18-19 tariff war across different product groups, most evident for mobile phone and digital products, partly on their relatively high unit value. We also look (Figure 5) at the discrepancy by categorizing the products by the scale of tariff hike they faced (before the February 4, 2025 10% tariff hike). Interestingly and intuitively, the consumer goods subject to more tariff cost increases (e.g. 10-25% tariff hike) saw the largest reversal after 2019. They are well illustrative of the role of de minimis shipments from China to the U.S. amid the rapid development of China’s cross border e-commerce, to partly bypass the tariff cost in recent years.

In conclusion, the discrepancy in bilateral trade data aligns with the increasing significance of de minimis shipments to the U.S. since 2018. The rising tariffs have clearly incentivized Chinese exporters to rely more heavily on these shipments.

How will Chinese CBEC giants react?

China’s cross-border e-commerce (CBEC) exports surged from $92.7 billion in 2018 to $261.7 billion in 2023, and likely exceeded $270 billion in 2024. The share of CBEC exports in China’s total exports increased from 3.7% in 2018 to 7.7% in 2023. Although the U.S. share in China’s overall export markets has declined since the trade war began in 2018, its share in China’s CBEC exports has remained relatively high at around one-third, though data is only available since 2022 (Figure 6). Note the low-value shipment is based on simplified customs procedure, a system developed in recent years, covering only part of the mirrored de minimis shipments to the U.S. Another major destination of China’s CBEC (low-value) exports is Europe (e.g. ~17% for the UK, Germany and France as a whole). Back in 2023, the European Commission already proposed a removal of the long-standing import duty de minimis exemption (with threshold at EUR 150). We see rising challenges for China’s CBEC business, especially in the de minimis shipping space.

For U.S. textile and apparel imports, we found prices for hosiery, shorts, sweaters, trousers etc imported from China are roughly 30-40% lower than those imported from the rest of the world (Figure 7). Our colleagues covering the Chinese e-commerce secto Temu, AliExpress) vs. local retail platform (e.g. Amazon) in their same-product price comparison. Table 1 illustrates the price impact if the de minimis tariff exemption is removed, assuming all the cost increase will be passed through to end-demand consumers in the U.S. Assuming the same dutiable value for goods under de minimis shipment and local retails, the removal of de minimis tariff exemption will lead to a 30% cost increase for goods shipped by Chinese CBEC (or $13.5 in the table), along with $2.62 merchandise processing fee per package, while the incremental tariff cost increase will be only 10% (or $4.5) after latest tariff hike for local retailers as they are already subject to the average tariff cost (~20%). With that, the price discount for Chinese CBEC under de minimis shipment will narrow to 27%, though the discount could be modestly larger if the tariff cost is partially taken by the seller, or smaller if considering additional shipping disruption induced cost increase.

Beyond the illustrative example we present in Table 1, there could also be a non-trivial aggregate impact on inflation as de minimis imports were sizable over the past couple of years. As our example suggests, there are multiple moving parts in the de minimis trade, including additional administrative fees that would have been incurred if the same shipments were to enter subject to tariffs. In fact, Fajgelbaum and Khandelwal (2024) analyzed this counterfactual scenario and estimated that de minimis trade likely saved somewhere between $11.5Bn and $18.1Bn for U.S. consumers. This translates into 0.08 to 0.13 pt increase in PCE inflation at its peak in 2018. In other words, de minimis trade and the cost saving consumers had, likely kept inflation a few tenths lower amidst the first trade war.

Our China e-commerce equity analysts suggested that platforms like Temu are anticipating higher tariffs and are adjusting their business models accordingly. Large Chinese cross-border e-commerce platforms, such as Temu, are reportedly investing in U.S. warehousing infrastructure to prepare for tariff increases on direct shipments from China. This strategy was prompted by the Biden administration’s September 2024 announcement outlining plans to address the misuse of de minimis exemptions. It seems likely that these platforms will continue to gain market share as they expand their warehousing and retail networks in the U.S., despite higher tariffs.

Notably, ocean freight costs are much lower than air freight, and with a larger presence in the U.S., the absence of the de minimis exemption may not significantly hinder these platforms, provided that tariffs on consumer goods typically sold on these platforms do not become excessively high. Ironically, this shift might result in a reported increase in imports from China as de minimis shipments decline.

The United States Feb 10-14th

• Tariffs on Canada and Mexico were temporarily postponed, even as new 10% tariffs on China went ahead

• We expect to see more tariffs, or the threat of them, in coming months
• The labor market remains strong, as jobs increase 143k and the unemployment rate fell to 4.0%
• Next week we look for 0.2%m/m rise in core CPI, and continued solid gains in control retail sales

The week that started with a 30-day postponement of the planned tariffs on Canada and Mexico even as new 10% China tariffs went into effect, continued with more headlines related to federal job cuts, and ended with a decent January employment report. Actually that wasn’t quite the end, as on Friday afternoon President Trump announced plans for reciprocal tariffs on all US trading partners next week. However, this news may just refer to a campaign proposal to craft legislation that gives Trump that option.

On trade, the not-entirely-surprising delay in the Canada and Mexico tariffs came after both countries announced steps to address border issues. Another delay seems likely, though we doubt that will go on indefinitely, nor do we think the 10% increase in tariffs on China is the end of that story. On his first day in office Trump signed the American First Trade Policy, which calls for a report on trade practices by April 1. More significant steps could follow. In the case of China we expect that means higher tariffs by 3Q. For Canada and Mexico the end-game may be a renegotiation of the USMCA trade agreement. But the implementation of the China tariffs alongside the threat for further trade restrictions has elevated uncertainty and could adversely affect US economic growth.

However, we are not making any changes to our growth forecast for 2025, as the economy is entering this bout of trade uncertainty on solid footing and the actual outcome of some of the proposed tariffs is far from certain. That being said, the uncertainty alone from various trade headlines could cause companies to re-think planned investments, and we see risks to the capex outlook building.

On deck for next week—in addition to any news out of the Capitol—will be the January inflation reports and retail sales. Inflation has been moderate in recent months but a lingering question is whether we could see a step-up in 1Q similar to last year. Our assumption is that the 1Q24 jump doesn’t represent residual seasonality that will repeat this year, and we look for a 0.2%m/m (3.1%oya) gain in core CPI. On retail sales, a plunge in auto sales as well as weather-induced declines in sales at building material stores and restaurants should depress the headline (-0.7%m/m), but we think the control category should post a solid gain (0.5%).

Labor market in good shape

The January employment report showed nonfarm payrolls rising 143k last month, and a net upward revision of 100k to the prior two months. Payrolls have now risen by 178k per month over the last six months. While the January print was slightly below that level, overall the labor market continues to appear quite good. The unemployment rate also dipped from 4.1% to 4.0%, the lowest level since May. In fact, the unemployment rate would have fallen two-tenths (from 4.2% to 4.0%) had new population controls introduced in January also been used in prior months. For the Fed, while there are a lot of risks swirling around, the hard data are coming in close to their Congressionally-mandated goals. With an economy that ain’t broke there’s nothing to fix, and so the FOMC will likely be doing nothing for the foreseeable future.

The details of the report suggested that weather—primarily the cold spell last month, but possibly also the LA wildfires—could have reduced payroll growth in January. This impact was evident both in the industry mix of employment changes and in the workweek and hourly earnings. Leisure and hospitality employment fell 3k, about 35k below the six-month average trend, while construction employment rose just 4k, which is 10k below trend. Transportation and warehousing was also almost flat after averaging gains over 20k in November and January. Rather than weather-related, this slowdown likely reflects difficulties in seasonally adjusting package delivery employment around the holiday months.

The workweek fell from 34.3 in November to 34.2 in December and then again to 34.1 in January, a new cycle low. While normally worrisome, the fact that it came with an acceleration in hourly earnings growth along with the fall in the unemployment rate points toward weather. Average hourly earnings jumped 0.5%m/m (4.1%oya), and over the last three months the estimate of aggregate nominal labor income grew close to 5% annualized, double the rate of PCE inflation in that period. One final sign of a weather effect came from the count of people in the household survey who were absent from work because of it: that was 591k, about double the median of 284k in past January’s.

Even more job news

The January employment report also featured the annual benchmarking of the establishment survey and new population controls in the household survey, with the latter used to calculate the population size for the country overall and for demographic groups. The March 2024 level of seasonally-ad-justed payroll employment was revised down 589k (0.4%): a large move but not quite a bad as we had feared. The monthly changes after March were also revised to account for updates to the birth/death adjustment and new seasonal factors. Those led to employment growth being revised down over much of the year, but revised up in the last three months of 2024 (Figure 1). While the period after March 2024 could also eventually see a downward shift following the next annual benchmark, the trajectory within the year does show a mid-year slowdown giving way to faster job growth in more months.

The population update to the household survey was more noteworthy than usual this year as the Census Bureau revised up its population estimate by 3.5 million to account for the immigration surge in recent years. This translated to a 2.0mn increase in household employment, and raised the participation rate and unemployment rate 0.1%-pt relative to what they would have been without the adjustment. The combination of the upward revision to the household employment and the downward revision to establishment employment mean the two surveys now agree much more than before: over the last four years, the gap between the change in establishment employment and the household survey’s payroll-adjusted concept of employment shrank from 3.1mn to just 311k.

The abating productivity tailwind
The solid productivity growth that has been a key support for economic growth recently appears to be abating. In 2023, Q4/Q4 growth in nonfarm business productivity increased 2.7%, but that decelerated to 1.6% in 2024, matching the years immediately before COVID (Figure 2). The rollout of AI could lead to a renewed acceleration in growth, though in practice this is probably still some way off given the early stage of adoption at many companies. While AI gets the headlines, the overall NIPA measure of intellectual property product investment has been soft, and equipment investment is also moderate.

Productivity has decelerated faster than wage growth, so unit labor costs accelerated between 2023 and 2024, rising from 2.1% to 2.7%. While a bit firm, this is probably not high enough to be alarming to the Fed, especially since these can be quite volatile. Also, business margins have been strong since the pandemic, which reduces the need to immediately respond to wage growth with corresponding price increases.

Household sentiment responds to tariffs

The fact that tariffs news is dominating the airwaves is not lost on consumers. The preliminary February release of the University of Michigan survey saw sentiment fall and inflation expectations rise on the back of tariff-related concerns. After rising steadily from last July, the sentiment index has declined more than 6pts over the past two months to 67.8, nearly back at that recent low. Both present conditions and expectations fell this month. Meanwhile, one-year ahead inflation expectations jumped a full percentage point to a 14-month high of 4.3%, while five- to ten-year ahead expectations broke above their recent range to reach 3.3%. The 1.5% surge in one-year inflation expectations over two months is large but not unprecedented. The survey was taken over a period that culminated in Trump threatening but then postponing tariffs on Canada and Mexico, so these figures could ease in the final release. Still, with new tariffs frequently in the headlines, these measures may stay elevated for a while. While sentiment fell across all political affiliations, the gap in inflation expectations between Republicans and Democrats widened to its highest recorded level (Figure 3).

Focus: Federal jobs

The last few weeks have brought a flurry of headlines regarding job cuts for federal employees and contractors. How will this affect labor market indicators such as monthly nonfarm payrolls and weekly jobless claims? In short, we don’t think the effects will be visible right away, both because of the limited scope of cuts announced so far and because of the definition of employment. Recent headlines relating to cuts in federal employees or contractors include:

• The offer to many federal civilian employees of deferred resignation, which allows them the option to stop performing government duties while still being paid through September 30, or between October 1 and December 31 for employees retiring in that span. Postal workers and those in immigration enforcement or national security roles are excluded. The offer originally had to be accepted by February 6, by which time 40,000 employees had done so, though a court extended it through February 10.
• Shuttering USAID and retaining only about 300 employees. While newspapers have quoted an employee count of at least 10,000, the FedScope database shows only about 2,500 personnel in the US as of March 2024. An unknown number of outside contractors could also be affected, though presumably many would be outside the country.
• On February 2 Bloomberg published an article headlined “Musk Says DOGE Halting Treasury Payments to US Contractors,” which was based on statements Mr. Musk made on X. However, the only specific item that the article mentioned was a “Lutheran charity,” and since then the Treasury has stated that DOGE has “read-only access” to payment data, implying that DOGE employees would not be unilaterally stopping contractor payments.
• On February 6 the WSJ reported the White House was working on an executive order that could come out as soon as next week to fire “thousands” of employees at the Department of Health and Human Services. The White House has denied such a plan. There have also been stories of planned large-scale cuts at the General Services Administration, Office of Personnel Management, and National Science Foundation, which collectively employ less than 20k people.

So far these totals appear small given the federal government directly employs 3 million workers (Table 1), as well as a large number of contractors. Moreover, these moves are unlikely to make a noticeable dent in total monthly payroll changes. In particular, employees on deferred resignation will still be paid through at least September 30, which means they will still be counted as employed. The deferred resignation program also allows workers to find a new private sector job even as they receive deferred compensation, so many of these workers could register as having two jobs, thereby boosting reported tota employment until September 30. Also, since these workers would be willingly leaving their jobs, they would not be eligible to file unemployment insurance claims. As for the USAID layoffs, the number of domestically based employees is small.

In addition, a number of the 40k employees who have taken the deferred resignation offer may have been planning to leave even without it. In recent months close to 25k employees per month were choosing to leave the Federal government (Figure 1). These employees would be incentivized to accept this offer, since it will give them months of pay without the government preventing them from taking another job.

Given the points above we don’t see large government payroll reductions by February, though layoffs could grow over time. Any layoffs would likely be initially targeted at workers outside the Postal Service, Defense, and Homeland Security departments, and possibly also Veterans Affairs. While different data sources don’t fully agree on how many workers that number is, the government’s FedScope personnel database shows it to be 800k as of March 2024, or close to 1,300k including Veterans Affairs. Contractors could also be affect- ed, and we plan to explore that group further in future work.

The United States Feb 3-7th

• Trump’s threat to impose 25% tariffs on Canada and Mexico appears more real as of publication time

• This would weigh on growth and push up inflation; the magnitudes will depend on the details
• Heading into this possible trade war, core inflation continues to moderate alongside solid GDP growth
• The Fed met this week; like the rest of us they are waiting and watching

In a normal week this essay would start with a discussion of the Goldilocks-like data received this week, which featured solid growth alongside tame inflation. But that would ignore the much more important development in the mounting risk to the global economic order. Late Thursday afternoon, Trump reiterated his intention to impose 25% tariffs on Canada and Mexico this coming weekend. Since then, the head-lines have continued to bounce around, first suggesting tariffs would be delayed until March 1, then that the Canada and Mexico tariffs would be imposed on Saturday, as well as an additional 10% on China.

As of the time of publication, it is still unclear what trade policy will be in place next week between the US and its three largest trading partners. Pending more concrete information about the nature and extent of the threatened tariffs, as well as any possible retaliation, we have not changed our economic outlook.

As we have noted previously, such an outcome would constitute an adverse supply shock that would both boost prices an hit growth, although the magnitude and timing of each impact depends critically on the details of the plan. In a research note from last week we discuss some of the potential impacts of the types of tariffs that could be implemented this weekend. The growth estimates are particularly difficult to pin down, with estimates ranging from 0.2-2.0% of GDP.

Amidst the headlines emanating from Washington this week it was business as usual at the FOMC meeting, where as expected the Committee left rates unchanged. In the press conference Chair Powell stressed the FOMC was in no hurry to cut in the near future, that they could keep policy on hold if inflation stays firmer than desired, but that could also ease rates if there was undesired weakening in the labor market. Not surprisingly Powell was asked a number of questions on Trump administration policies, which as expected he largely avoided except to note that it was too soon to know how policies on trade and immigration could affect the outlook.

Elsewhere on the policy front the administration briefly announced and rescinded a pause on federal loan and grant programs, which we discuss in the Focus. While this is off the table for the moment, this may be a first step toward eventually challenging the constitutionality of the laws that prevent the Executive branch from impounding Congressionally approved spending. And in other news, the administration seemingly encouraged a large number of federal workers to resign, a step that was accompanied by the type of FAQ not normally published on OPM stationery. Predictably there is some uncertainty about whether the move, which offered extended paid leave, is in fact legal.

For the week, and weekend, ahead, it seems safe to say that the most important development will be determining how the White House wants to alter our most important trading relations. It will also be a busy week on the data calendar, with the key report coming Friday with the January employment report. We look for an 150,000 increase in nonfarm employment with the unemployment rate holding steady at a low 4.1%. That assumes some small drags from the California fires and cold weather nationally, so the underlying trend is likely firmer. Over the last six months job gains have averaged 165k, though some strong prints recently raise the chance that the trend is above 175k. This release will also include the annual employment benchmark, which we discuss further in the employment data preview. That will lower lagged job growth materially in the establishment report but raise it in the household survey, while the unemployment rate should see little impact.

That was then…
As noted above, the data received this week left little to be desired. Growth in economic activity remains solid and the labor market appears to be on good footing through mid-January. Even so, core PCE inflation continues to moderate, with the December reading increasing only 0.16%. While the year ago reading was unchanged at 2.8%, the three- and six-month annualized readings of 2.2% and 2.3%, respectively, show an improving trend. Meanwhile, moderating wage inflation suggests that—absent tariffs—these better inflation reading should continue.

In particular, the 4Q employment cost index (ECI) increased at a 3.6% annual rate (3.8% over year-ago), which is neither too hot from an inflation perspective nor too cool from a consumer spending perspective (Figure 1). The marquee data release of the week was the 4Q GDP report. Real GDP increased at a 2.3% rate last quarter. While a little slower than the 3.0% pace averaged over the prior two quarters, much of the deceleration owed to a sharp slowing in inventory accumulation, and final sales rose at a 3.2% annual pace. Private domestic final purchases (PDFP), consumption and investment, also increased at a 3.2% rate.

The 4.2% annualized increase in real consumption last quarter was led by a 12.1% increase in spending on durable goods, headed by a rise in autos. Some of this vehicle spending may have been front-loaded purchases of EVs ahead of a likely end to tax credits, as well as some replacement of vehicles destroyed in last fall’s hurricanes, and indeed we expect unit sales to fall 7%m/m in January. Even so, spending growth for non-durables and services was also solid last quarter, indicative of the general health of the consumer.

While consumers were humming along in 4Q, business capex stumbled, as the 2.2% pace of contraction was the worst quarter in just over three years. Some of this weakness may have been payback for strength earlier in the year, particularly in the volatile aircraft segment, though there were declines in a number of other categories as well. Combined with strength in earlier quarters though the pace over-year-ago of 3.5% is in line with post-pandemic trends, if not particularly strong in the longer run (Figure 2).

Prospects for business investment are front and center for two of the biggest news developments this week. First, the Deep-Seek news that jolted markets at the start of the week raised questions about the outlook for data center investments. As we highlighted in a recent note, capital outlays for data centers has been a big support for overall capex recently. While we have been expecting that contribution to gradually slow, the recent news raises risks for a faster deceleration, though it could also boost the pace of innovation and customer uptake in AI. Even as AI is racing ahead though the NIPA measure of intellectual property product investment is increasing slowly. Some of this category is imputed for the most recent quarter, but over the last year it has still only risen 3.5%, on the low end of what is normal for an expansion (Figure 2).

The second development is that the uncertainty created by the disparate headlines in trade policy could have its biggest effect on business investment. Even if trade policy ends up unchanged, not knowing the future rules of the game could cause companies to delay plans for spending on new plant and equipment. Some Fed research in 2019 had estimated that trade uncertainty around China tariffs took off nearly a percent from US growth, and uncertainty now is at least as high.

The New New Colossus
In last week’s Data Watch we summarized the range of measures the new administration has taken to reduce immigration stemming from those without regular visas, including individuals seeking aslyum or refugee status, or being admitted under humanitarian parole. We noted that this form of immigration may have been adding on the order of 50k per month to employment. Even if this immigration flow falls to zero right away the labor market effect won’t immediately dissipate, since older cohorts of arrivals will be increasing their labor force participation, while new immigrants may try harder to evade the CBP at the border. Nonetheless, we do expect this slowdown to start to affect payroll growth soon, potentially by the time of the February jobs report.

ICE has also raised deportations, with daily figures from ICE showing arrests of around 1,000 per day. In fiscal year 2024 ICE made a total of 146k arrests between its ERO and HSI teams—about 400 per day assuming arrests were made every day of the year. The recent pace could go higher, as the Washington Post reported that a quota of 1,200-1,500 per day has been set. To calibrate what this might mean for employment, consider as an upper bound that arrests were carried out seven days a week (though it could be just five), that 1,500 people per day were arrested, that all of them were working age adults, that 72% of them were employed (the national average of people aged 16-64), and that all those arrested were deported, then employment would be reduced by roughly 24k per month more than during the prior year.

Given the above calculations, employment growth could fairly quickly move closer to 100k per month, assuming no change in labor force participation rates across the economy. On the other hand, this slowdown could be ameliorated if reduced net immigration were to spur higher participation among the rest of the population.

Global Data Watch: Exceptional but not bullet-proof

• Continued positive signals on global growth …
• … are trumped by rising concerns of disruptive US tariff policy
• Expecting a firming in goods and divergence in service price inflation
• Next week: Modest US jobs; tame EA core HICP; BoE, Banxico, RBI, cuts

Incoming economic data bias risk to the upside to our already upbeat current-quarter global growth forecast, while tilting to the downside our forecast for persistent 3% 1H25 global core inflation. However, any shifts in risk arising from the latest economic releases or central bank decisions pale in relation to the rising risk around US policy. We have little conviction in our forecast that the Trump administration balances support for the US business sector with disruptive policies that restrict immigration and trade. On balance, this baseline assumes a modest negative global supply shock alongside a positive US sentiment boost in response to an administration signaling its intent to provide regulatory and tax relief.

Early readings on US sentiment align with our baseline, but it remains difficult to distinguish posturing from policy intent in the varied signals coming from the Trump administration to date. We are, however, increasingly concerned that the policy mix may tilt into an unintentionally far less business-friendly stance. The sustained 25% tariff on Canada and Mexico set to start this weekend is materially different than the tariff increases built into our baseline as it would generate a large negative supply shock concentrated in countries that have far bigger spillovers to the US. Importantly, it threatens to dismantle a multi-decade free trade agreement, disrupting supply-chains and depressing business sentiment—which heightens risks of non-linearities not captured in economic models.

Tariffs on Canada and Mexico, purportedly driven by drug trafficking and immigration motivations, could also signal the broader application of trade restrictions. Specific tariffs may be imposed on countries deemed to be engage in “unfair” trade practices (Brazil, China, and India) and on sectors to safeguard US economic security (commodities, semiconductors, and pharmaceuticals). These tariffs may be calibrated to minimize their impact on inflation and disruptions to supply chains, but the combined effects of this scatter-shot approach could impart a larger growth drag than we envision.

On the immigration front, signals that a ramp up in deportation activity may be in the offing further concentrate a policy-driven supply shock on the US. A slowing inflow of new immigrants is already incorporated in our forecast. While estimates of the incremental drag of deporting a million people this year are modest—about 0.5%-pt off GDP by 2026—this shock could be magnified through its concentrated sectoral impact (agriculture, construction, hospitality) and through tightening labor markets.

There are good reasons why financial markets and US business sentiment continue to send an upbeat signal as the expansion stands on solid foundations and the commitment of this administration to support business seems genuine. However, an ideological bias to disengage the US economy from the rest of the world is strong and we are not confident that a policy balance can be easily achieved. The US expansion’s sustained outperformance is not an immutable reflection of structure. Impressive US growth and wealth creation since the pandemic reflect an interaction of a healthy underlying structure with supportive fiscal and industrial policies, a large immigration boost, and upbeat private sector behavior. We should not underestimate the resilience of the expansion in the face of modest shocks, or the power of large shocks to shake the foundations of US exceptionalism.

Current-quarter growth looks strong Lower than expected US and Euro area GDP gains dampened our 4Q24 GDP tracking. However, we continue to track an above-potential 2.8%ar and see upside risk to our forecast for GDP to expand at a similar pace this quarter. In the US, there is considerable noise to look through in last quarter’s 2.3% gain with offsetting swings in a durable consumption acceleration and inventory stall alongside an up-and-down 2H24 profile of transportation equipment spending.

Through the noise, income generation looks to remain balanced and underlying demand appears to be tracking at close to a 3%ar. While net trade will likely continue to weigh on GDP growth, the broad-based momentum in activity readings at year-end points to upside risk to our current-quarter forecast of a 2.3%ar GDP rise. Next week, the January employment report is likely to show a moderation in hiring to a 150,000 gain with temporary drags from weather and fires temporarily depressing gains.

While it is hard to send an upbeat message around a Euro area expansion which stalled last quarter, we see the latest news as reducing downside risk. Smoothing through noise created by Ireland and the Olympics, the region looks to be expanding closer to a trend-like pace. Importantly, the labor market remained resilient last quarter, credit data are moving higher, consumer spending was likely firm. With the PMI improving in January we are comfortable that growth will approach 1% this quarter.

EM Asia ended 2024 on a strong note as the region received significant benefits from the global pickup in manufacturing. It is not clear to what extent this strength reflects front-loading of activity ahead of the Lunar New Year and early signs from Taiwan suggest January IP and exports could slow meaningfully from a December bounce. Next week’s January PMIs will help guide us through this noise and we expect Taiwan and Korea to post gains. ASEAN PMIs are expected to remain broadly stable vis-à-vis December and consolidate last month’s rise.

In LATAM both Brazil and Mexico slowed in 4Q24, but we are not projecting this weakness to persist into the new year. Despite subdued January sentiment readings, we look for a stronger 4%ar GDP gain in Brazil, driven by a healthy grain crop and continued solid consumption gains. Mexico’s flash 4Q GDP print showed a contraction but like in Brazil, growth looks set to bounce back this quarter on the back of the agriculture sector and solid consumption. Supported by solid outcomes anticipated across the region, LATAM GDP is expected to rebound to a 3.5%ar this quarter.

China advancing but awaits its tariff fate

The latest news of a 10% tariff on China is less than we forecast but it would be a mistake to think this is the final word on US China trade. An executive order signed on Day 1 calls for a review of the progress on agreements made in 2019—of which there has been little. A potential Trump visit to China could delay further tariff hikes. While China waits, its economy is bouncing from the policy stimulus announced in 2H24.

On the heels of a boomy policy-induced 7.6%ar jump in GDP last quarter, we look for a near 6%ar gain this quarter. Government bond issuance was strong in January while the first batch of a trade-in subsidy announced by the NDRC in early January should provide continued support for retail sales. High frequency data also show a pickup in container shipping, international flight, and property sales.

Services inflation sticky but divergent
While we see a global firming impulse placing upward pressure on goods price inflation, persistent business cycle diver-gences are set to drive country variation in service price inflation outcomes. Economies that have generated weak domestic demand and rising unemployment rates are set to see more marked inflation declines and we project low inflation out-comes––within 0.5%-pt of central bank target midpoints––for Canada, Sweden, Norway, and New Zealand alongside EM Asia, Peru, South Africa, and Czechia (Figure 3).

Forecasting inflation is more complicated for countries in which demand and supply side performance have conflicted (Figure 3). In the US, strong immigration flows and productivity gains have modestly eased labor market tightness even as GDP increased at a 2.8%ar over the past eight quarters. At the same time, the past two years have seen little labor market slack open in Mexico, the UK, Euro area, and Australia despite persistent subpar GDP growth. Our bias is to place more weight on demand performance in the inflation outlook for these countries.

Incoming news from the Euro area supports the case for significant further disinflation. National releases point to lower limited start-of-the-year price resetting which we believe will be confirmed in Monday’s Euro area core HICP print with an expected 2.6%oya gain. Separately, the ECB corporate survey on wage expectations showed expected wage growth slowing to 2.7% in 2026, fully consistent with inflation firming to a 3.3%ar in the final three months of 2024 while core PCE inflation cooled to a 2.2%ar. Uncertainty about evaluating the underlying signal from this news will linger but this week’s 4Q24 reading sent a constructive inflation signal as the year-ago gain in private sector wages cooled to 3.6% (Figure 4).

EM CBs: each to their own
EM central banks are delivering on our expectations of varied policy responses amid divergent domestic conditions and constraints from a pausing Fed and US policy uncertainty.

• In LATAM, Mexico’s weak GDP reading leads us to now expect a 50bp cut next week. Banxico’s 2025 Monetary Program suggests that larger cuts could be implemented as the central bank appears comfortable tolerating a weaker exchange rate. We expect a second 50bp cut in March, before Banxico shifts back to a 25bp pace in May. Mean- while, Brazil’s BCB hiked 100bp this week amid rising inflationary pressures. Recent weaker activity readings led BCB to highlight the risk of a more material growth slowdown but elevated CPI forecasts offset this dovish tilt. We look for another 100bp hike in March. Chile’s CBC kept the policy rate stable at 5%, as expected. The statement suggests this pause is likely to be extended throughout the first half of the year.
• In CE3, Hungary’s NBH kept rates unchanged at 6.5%, but turned more hawkish in its communications. The reinforced hawkish stance suggests the resumption of cuts we are expecting from March could be postponed. At the same time, Czechia is likely to resume cuts next week after a more favorable CPI print for December. Only a sharp upside surprise in the January CPI, to be published a few hours before the rate decision, could derail the cut.
• South Africa’s SARB cut rates 25bp as expected, but both the hawkish tone and split decision came as a surprise. Despite sub-target inflation, the SARB cited material upside risks from the external environment. Recent political frictions within the government of national unity have also weighed on the rand. We no longer expect further rate cuts unless core inflation materially surprises to the downside.
• In India, we believe fiscal prudence will win the day in its Budget and pave the way for the RBI to cut rates by 25bp at next week’s MPC meeting. A raft of liquidity measures announced earlier this week should thus be seen as a pre- cursor to next week’s cut.

The United States Jan 27-31st

• The risks around more widespread tariffs increased this week, keeping policy uncertainty high
• Executive orders were about as expected, including measures to lower immigration
• LA fires could be a drag on January job growth
• Next week’s data to show robust 4Q GDP, moderate Dec core PCE, and a rebound in ECI growth

The week was dominated by headlines from Washington following President Trump’s inauguration. The resident quickly issued many executive orders, some of which echoed those from his first term and which were mostly in line with expectations. That left key areas of policy uncertainty open, including on tariffs. On the tariff front the president suggested a new 25% rate could be applied to Canada and Mexico as soon as February 1. We see this as linked to a desire to start renegotiating parts of the USMCA trade agreement and to get more help from the two countries, especially Mexico, in dealing with illegal immigration and drug trafficking. It would not surprise us if some announcement noting progress towards these goals comes out soon and delays the start of tariffs, though if they were to go in as planned we expect they would raise prices and reduce growth in all three economies.

Working around this policy uncertainty will also be a task for the FOMC, which meets next week. Fed communication has signaled that the committee plans to pause rate cuts at this meeting, so the monetary policy decision should be a non-event. The statement will likely not provide a strong signal about the March meeting, and while Chair Powell probably won’t foreclose the possibility of a rate change in his press conference we think the bar for further cuts by March remains high. Powell will also likely note that each FOMC member is forming their own forecasts about trade or other fiscal policies, and avoid implying a consensus expectation about how these policies will affect the economy. Going through 2025 the FOMC may be hesitant to adjust policy preemptively in response to tariffs, as was the case in 2019 when the first rate cut that year came over a year after the US and China began raising them. A caveat to that point is that inflation concerns are higher now, which could lead the Fed to maintain rates even if tariffs induce a drag, or make them more willing to raise rates if the inflationary effects were to dominate.

The data flow also picks up next week and will include a variety of activity reports for December along with the first release of 4Q GDP, which we expect will show a 2.75%q/q saar increase. Also of note will be the core PCE deflator for December, where we look for a 0.19%m/m (2.8%oya) rise, and the 4Q employment cost index, which we expect to have increased 0.9%q/q (3.6% ar), accelerating a bit from 3Q on the back of higher hourly earnings growth we’ve seen in the last couple quarters. Given that some labor market measures have tightened again recently wage growth could firm again later this year, though because of lags in the economy we could see further deceleration before that. For more on that point see this week’s Focus.

Taking (executive) orders

Among the range of executive orders that President Trump signed in his first days in office there were a number that could affect the economy, including by tightening immigration, freezing federal hiring for many roles, and pursuing energy sector deregulation. On the energy front the president declared a national energy emergency and reduced climate regulation with the intention of boosting fossil fuel production. It may take time to see how this will translate into further production, since as our oil and gas industry analysts note production volumes are already high. They do anticipate though that lifting the moratorium on LNG export licenses will be positive for demand, which is in addition to the pre-existing growth in energy demand coming from data centers, a point that we wrote about recently.

On immigration, the administration has moved to reduce immigration by, among other actions, suspending asylum for people arriving via the southern border; suspending refugee admissions for at least 90 days and possibly longer; terminating some or all humanitarian parole programs; expanding expedited removal to the whole country and applying it to immigrants in the country under two years, as opposed to applying it only within 100 miles of the border and to immigrants in the country under two weeks; restarting the Migrant Protection Protocols program, which will allow some immigrants to be returned to an adjoining country (mainly Mexico) while awaiting immigration hearings; and promising to increase deportations of immigrants already residing in the US who lack a visa, including individuals who are covered under humanitarian parole programs.

These steps will reduce one source of people flowing into the labor force, which could begin to translate quickly into lower values for monthly employment growth. While we don’t know exactly how fast new immigrants who lack visas become employed, presumably the effects could begin as soon as the February employment report and become more noticeable through time. Although border encounters (which include people arriving via humanitarian parole programs and with appointments via the now-suspended CBP One app) have already slowed, they were still mildly elevated at 125k/ month in November and December (Figure 1). Based on DHS custody data we estimate that would lead to roughly 80k adults per month being released into the US.

Given labor force participation rates for immigrants and estimates of how quickly immigrants become employed we fore- cast that recent elevated levels of encounters have been boosting employment growth by at least 50k/month (Figure 2, see our earlier note describing the methodology). This could be an underestimate if higher levels of border encounters are also correlated with increased flows of immigrants who avoid detection. We would expect that the number of adults being released into the US via these encounters will rapidly decline, though there may be some offset from immigrants more actively working to avoid detection.

On top of fewer new immigrants arriving a larger volume of deportations could also reduce reported employment figures. This NY Times article has a useful summary of the number of people who may be targeted, and the relevant legal path for pursuing a deportation. At this time we think it is difficult to accurately map out an exact timeline for deportations, or how many there will be. On balance though, we think breakeven employment growth could slow to around 100k or less later this year or in 2026.

The executive orders on the federal workforce likely won’t have too much of a near-term effect on employment growth. A federal hiring freeze is similar to an executive order that Trump issued at the start of his first term. Gross federal civilian hiring is around 30k/month, though some of these individuals are not subject to the freeze, and net employment growth has been 4k/month over the last year. These are all similar to the period just before Trump took office in 2017, after which federal employment fell by 2k/month over the following year, with the largest fall in April 2017 when employment declined 12k. In sum, were the same pattern as 2017 to play out now the impact on total job growth will be small. The DOGE might increase reductions, as could an order to stop remote and hybrid arrangement: according to a recent OMB report 1.1mn federal workers were eligible to work from home at least part of the time, of which a fifth were fully remote. However, the effects on total employment could again be limited if workers move into private-sector jobs.

What we learned from the data

Among the limited flow of new data this week we got the flash January PMIs, which highlighted a softer service-sector report, and jobless claims that suggest that the LA wildfires could have a larger drag on January employment than we had originally expected. For the PMIs, the service survey’s headline business activity index, which has been consistently strong since late spring, suddenly fell four points to 52.8. The press release did note that “especially adverse weather” was cited by some firms, and while weather usually has less effect on diffusion indexes it is possible that some of the weakness is a temporary weather effect. In any case, the employment index actually surged to its best level in two and a half years, and given the activity index’s noisy short term relationship to GDP growth we aren’t too concerned yet.

On the jobless claims data, country-level data from California shows an unusual pop of 6k initial claims in Los Angeles county (Figure 3), which appears fire-related since it matches neither the typical seasonal pattern in prior years nor the behavior in surrounding counties and the rest of the state. This could imply a material drag on January payrolls: for example claims rose about 5k in Florida after Hurricane Milton and by 15k in Texas after Hurricane Beryl, and both states subsequently had job changes near 50k below trend. We will refine our estimate of the LA effect next week when we get another week of claims data.

Focus: Getting tight again?

The labor market softened significantly in early summer 2024, prompting the FOMC to initiate an easing cycle with a 50-basis point cut. Since then, incoming data have been somewhat mixed, aligning more closely with the expected normalization of the labor market. Although hurricanes and strikes slightly dampened earlier payroll growth, the three-month average rebounded to 170,000 in December. Job openings have stabilized, hovering around 7.9 million since April. With the unemployment rate at 4.1%, labor market tightness, as measured by the ratio of job openings to unemployed, has remained steady in 2H24 (Figure 1). More timely survey measures, such as the Conference Board’s labor market differential, actually point to some firming since September. If this trend continues, it could pose upside risk to our wage inflation forecast, particularly in the latter half of this year.

Assessing labor market tightness can be challenging, especially as the market recovers from a period of high churn. The ratio of vacancies to unemployed individuals is both empirically and theoretically appealing as a proxy for labor market tightness at a given point in time. It provides a concise summary of the relative number of job seekers on both sides of the market. However, some vacancies are filled by already employed workers, potentially creating more vacancies in a phenomenon known as “vacancy chains.” During periods of high churn and numerous job-to-job transitions, measured vacancies may surge. Controlling for quits suggests that the significant drop in vacancies after mid-2022 brought the level of tightness close to 2019 averages, while the traditional vacancy-to-unemployment (V/U) measure has been below its 2019 averages over the past six months (Figure 2).

For policymakers and forecasters, these different measures of labor market tightness may have different implications for the wage inflation outlook. Recent research by Heise, Pearce, and Weber (2024) examines a broad set of labor market tightness measures and constructs an index (the HPW index) that improves forecasts for predicting ECI wage inflation. We utilize an even larger set of labor market measures, including survey-based ones, in addition to those shown in Figure 2, to construct a latent factor that summarizes the degree of tightness across all these data. The resulting measure is broadly akin to the HPW index (Figure 3). We can then use this more timely measure to project ECI wage growth further into 2025. Based on a similar lag structure we used in our 2025 outlook, this method suggests some further cooling in wage inflation early this year to pace that remains above its pre-pandemic level, followed by some firming later in the year to end 2025 around 3.5% oya (Figure 4).

North America’s Fraught Tariff Adventure

• Proposed 25% tariff on Canada and Mexico aims to start USMCA negotiations, win border concessions

• But February 1 deadline likely to be postponed

• Autos and energy, major US imports from CA/MX, will see either higher prices or carve-outs

• US inflation likely to rise; growth drag uncertain but recession risks for CA and MX

Although the flurry of executive orders did not include a Day One announcement of new trade restrictions, President Trump reiterated his intention to impose 25% tariffs on imports from Canada and Mexico, two of the US’s largest trading partners. Trump posited a February 1 start to these levies, but our base case remains that this is largely a bargaining chip to accelerate the re-negotiation of the USMCA trade agreement signed during his first term. However, potentially dismantling a decades-long free trade area could be a significant shock. Trade-related uncertainty remains high (Figure 1), and one lesson from Trump’s first term is that policy changes can be announced—or at least threatened—on short notice.

This is not the first time Trump has agitated for tariffs on Canada or Mexico; he followed through on some but not all such threats during his first term. It thus remains challenging to handicap the likelihood of additional tariffs sometime this year. What is clearer is that tariffs act as adverse supply shocks, pushing up inflation and down growth. There have been a wide range of estimates of these macroeconomic impacts; some point to recessions in Canada and Mexico as a result. Here we identify which sectors account for the most trade among the three former NAFTA economies to better understand the transmission mechanisms. The largest sectors, including energy and autos, are more likely to be excluded from any levies—thereby reducing both the potential costs but also some of the purported benefits of tariffs cited by the incoming administration.

Why this tariff threat now?

Trump has linked higher import duties on Canada and Mexico to his campaign promises to address the migrant crisis that resonated with a number of US voters, citing illegal inflows of migrants and fentanyl from both countries. In the case of Mexico especially, Trump is thus using the threat of tariffs as an extension of his immigration agenda. But the equivalence here is surprising—at least to Canadians—as the traffic is remarkably unbalanced. Over the 12 months through September of last year, US Customs and Border Protection seized 19.5kg of fentanyl at the US northern border, versus 9570kg at its southern border (another 316kg was seized at coastal and inland checkpoints). Further, estimates suggest less than 24,000 illegal immigrants crossed from Canada into the US over the past year, versus over 1.5 million from Mexico.

An additional motivation may be to accelerate the review of the USMCA trade deal ratified late in Trump’s first term. We now see this process starting in 1Q25 rather than 4Q25 as originally planned. There may be any number of reasons this issue has become a greater priority for the incoming administration, but the fact that Canada and Mexico have seen two of the three largest increases in bilateral trade deficits with the US since Trump left office is likely a top concern (Figure 2). How the USMCA review will play out is unclear, but it could readily turn into an outright renegotiation with broader, long-term implications for regional trade and economic activity.

Carrots and sticks

As Trump revived his threat to impose 25% tariffs late last year, both Canada and Mexico took steps to try to mollify Trump’s stated concerns. In December the Canadian government announced a plan to spend an additional C$1.3bn over six years to beef up border security; it currently spend around C$2.2 billion per year. But another irritant for the Trump administration—more spending on NATO defense—may have to wait until after the upcoming national election. Meanwhile, over the past few years Mexico has made progress in dealing with trade concerns expressed by the US, particularly with Asian countries and on key sectors (Table 1). Last week President Sheinbaum proposed a new investment plan ahead of Trump’s inauguration, which suggests movement toward easing bilateral tensions between the US and Mexico after years of a more confrontational approach. The plan is centered on “import substitution,” the need to increase national content,” and a “nearshoring decree,” among others. Sheinbaum also suggested the possibility of matching potential US tariffs on China. Although official numbers still show modest Chinese foreign investment into Mexico, we have argued it has in fact been increasing. But more dedicated action to limit the participation of China in the Mexican economy may not come into effect until 2026 or later, after rules of origin and national content are reinforced in the USMCA review.

By contrast, progress on immigration and drug trafficking has been slower. Since 2019 Mexico has been actively working to halt illegal crossings, but the surge in deportations since 2H23 allowed the incoming Trump administration to revive its focus on border security. US authorities also expect curbing illegal immigration will limit the drug trade. Mexican authorities have announced a number of recent drug seizures, particularly of fentanyl, but acknowledge more needs to be done.

In our view a more open policy toward dealing with organized crime can be used by local authorities to demonstrate that clear actions have also been taken on this front. This week’s decision of naming cartels as terrorist organizations is eyed as a cornerstone action in the ambitious agenda to address the bilateral supply-demand problem of illegal drugs.

At the same time that each country has attempted to offer carrots to the incoming administration, they also have been sharpening their retaliatory sticks. Both have pledged “robust” tit-for-tat tariff responses if the US goes ahead with 25% import duties. Canada reportedly has a list of targeted US products, supported by 12 of 13 provincial and territorial governments. Moreover, a recent poll suggests fully 82% of Canadians favor retaliation if the US imposes a 25% tariff. Similarly, indications are that Mexico is prepared to retaliate in proportion to US actions, including targeted responses if the US puts import levies on specific items or industries.

Where the pain could be felt

To get a better sense of how US tariffs—and the potential retaliation by Canada and Mexico—might affect each of these economies, we turn to an investigation of the bilateral trade flows by industry. We take a look at these data through a few different lenses: (1) what products dominate North American trade; (2) what is the share of North American trade amongst all trade for these countries; and (3) what share of domestic demand is met with North American imports and global demand met with North American exports. The last of these three concepts is the most important for understanding economic effects but is also more complex to estimate. We looked at a variant of this question in earlier work: the exposure of domestic production by industry to global imports.

Looking first at the mix of trade between the US and its two major trading partners, a few points stand out (Figure 3):

• The US import mix is materially different between Canada and Mexico. About two-thirds of imports from Canada are natural resources such as agriculture, forest products, minerals, and energy; or intermediate goods such as metals and chemicals. In contrast, these make up only a quarter of imports from Mexico. Most of the rest are manufactured products, including autos, computers and electronics, and electrical equipment.
• The auto sector in particular has a highly-integrated North American supply chain. Thus transportation equipment is a major component of bilateral trade across the region—although US imports of transportation equipment from Mexico are about 150% larger than from Canada.
• The total size and composition of US exports is quite similar for Canada and Mexico. In both cases a bit more than half of exports represent manufacturing excluding metals and chemicals.

North American imports collectively are also a large part of total US imports at 28% (Figure 4). Among these, a particularly large share of global imports of energy products (57%) and transportation equipment (49%) are sourced from Canada and Mexico. News reports suggest Trump is focused on bringing auto plants back to the US, and auto production is the industry that makes the most use of imported content, largely from Canada and Mexico. As it would be US producers who would thus be directly impacted by the tariffs, some exemptions might be carved out for this industry in the near term. Likewise, it is not clear that levying tariffs on energy products will accomplish much besides raising their price, especially as Canada accounts for nearly 60% of US imported crude oil and 97% of imported natural gas, as well as being a significant source of imported electricity.

Not surprisingly, US exports to Canada and Mexico also account for a large share (33%) of total US exports. Those exports are more evenly distributed across product categories than imports, and range from slightly over 20% of energy product exports going to those two countries to close to 40% for transportation equipment (Figure 5).

Wide range of potential macro impacts

As we noted at the outset, tariffs tend to act as adverse supply shocks, lowering growth and raising inflation. How large these effects are likely to be depends significantly on the details of how the tariffs are imposed and what sort of retaliation occurs. The latter tends to sharply worsen the economic outcomes for the countries that retaliate as well as their opponents in the models we cite below, but US trading partners broadly appear increasingly willing to state—at least as part of their own negotiating stances—that they will incur these costs in the near-term given the repeated-game nature of trade negotiations. Whether they actually do so, and to what extent, could materially change the realized outcomes.

There are two approaches to try to quantify how an 25% U tariff on Canada and Mexico might impact all three economies. The first is to look at history, while the second is to try to model it. The former approach is grounded in realized data but suffers from a lack of a close historical analogue. The latter can be carefully calibrated and matched to prior data, but still depends on simplifying assumptions to be tractable. We consider both, and view them as giving very rough guides to what the actual impacts could be over time.

Lessons from the Trump I tariff experience

In the prior trade conflict with China, US real GDP continued to perform well in 2019, rising 3.4%Q4/Q4 while the unemployment rate fell to 3.5%. Core PCE inflation actually slowed to 1.5% annually from 2.0% the prior year. Part of the the strength in economic activity reflected an offsetting boost from passage of the Tax Cut and Job Acts (TCJA) in Decem- ber 2017: based on the Brookings fiscal impact measure, the TCJA added 0.50-0.75% to growth in 2019. However, business equipment investment fell 2.2% that year despite TCJA measures to boost investment spending. Additionally, manufacturing production dropped 2.7% and manufacturing employment growth halted, the PMI and ISM manufacturing surveys both deteriorated, and export growth slowed (Figure6). This deterioration, with the slowdown in inflation, was enough to prompt the Fed to start cutting rates in mid 2019.

Note that the currently proposed tariffs on Canada and Mexico would apply to somewhat more trade activity than those imposed on China earlier. Canada and Mexico accounted for 28% of US goods imports ($914bn) and 33% of exports ($683bn), respectively, over the past year, with imports equal to 3.2% of US GDP. By comparison, mainland China’s imports were 2.6% of GDP at the start of the last trade war, then began to steadily decline as the tariff rate was boosted from 3% to 21% in 2018-2019, falling to 2.1% just before COVID and 1.5% today. However, these figures likely overstate the extent of substitution away from China given that some Chinese exports to the US were re-routed though third countries. That said, muddling through a trade conflict with China might create the same shocks to sentiment or supply chains as upending a decades-long free-trade arrangement.

Indeed, a more pessimistic growth signal comes from research by Caldara, et al, who estimated that the 2018 spike in trade policy uncertainty alone subsequently reduced US GDP by 1%, excluding any drag from the actual tariffs themselves. Their measure of trade policy uncertainty has spiked even higher now (Figure 1), as have other similar metrics.

For Mexico specifically, there are similarities to 2017-18 under the first Trump regime, where erratic policies from AMLO and the risk of cancelling NAFTA saw Mexico fall into a recession the following year, in 2019. We do not eye a recession at this point for 2025-26, but the possibility of that scenario is clear based on our own analysis of very damaging effects from a permanent 25% tariff on its exports to the US.

Modeling the Trump II tariffs

A comparison of external modeling efforts for the impact of a 25% tariff on all three economies (Table 3) reveals just how uncertain these estimates can be. The three sets of estimates cited below all start with some version of a dynamic general equilibrium trade model, then tweak parameters and inputs to attempt to quantify the growth and inflationary impacts.

What stands out across all estimates is that the US, as a larger and more diversified (and mostly closed) economy, suffers less damage than its major trading partners. This may account for part of the appeal to Trump, expecting that raising tariffs will put the US into an advantageous negotiating position. Estimates from the Canadian Chamber of Commerce specifically model the consequences of Canadian retaliation, which suggests a 0.6%-pt larger drag on GDP for the US and 0.8%-pt larger drag for Canada. At -2.6%, the shock to Canadian output significantly risks a recession there, while the -1.6% hit to US output would push growth below potential. Separate work by our Mexico team concluded that it, too, risks a recession.

The Peterson Institute estimates appear to give a lower-bound for the growth effects. They caution, however, that their modeling approach does not fully account for tightly-integrated North American supply chains, particularly for autos, and thus could be underestimating the drag. This tight integration implies that a product can cross the border multiple times during assembly, which raises gross trade flows relative to final sales and likewise the cost of tariffs relative to final sales. These estimates would not be as downward biased if the tariffs were structured as a value-added tax, but this adds complication—and there is no evidence that the Trump administration is prepared to implement such a scheme, let alone that companies are prepared to comply with it, by February 1.

By contrast, the earlier 2019 Bank of Canada study may over-state the impacts, as it assumes a global 25% tariff (i.e., universal retaliation). Nonetheless, the long-run hit to GDP for both Canada and Mexico is nearly three-times that for the US. An even bigger deviation occurs with inflation, with a sizable cumulative rise estimated for the US but a price level that ultimately settles below the no-tariff baselines for Canada and Mexico thanks to deep contractions. As a result, despite the near-term boost to inflation that likely would follow from large 25% tariffs, we would expect central banks to largely look through what should be a one-off price level shock and potentially shift to more accommodative stances.

We first discussed the implications of Canada and Mexico tariffs last December with a focus on price effects, in which we estimated a 20-40bps increase in the US price level in response to a 25% across-the-board tariff. As expected, uncertainty is high here: when using the same methodology recently to examine global tariffs, we found that some third-party inflation estimates were materially higher.

The bottom line is that the imposition of 25% tariffs by the US on Canada and Mexico would result in a worsening out- look for all three economies, and the impacts could potentially be sizable—especially if they presage the end of the region’s free-trade area. Whether that realization reduces the chances that they will be enacted (and retaliated against), only time will tell. Our baseline remains for these to be used primarily as negotiating ploys.

US: The Fed in 2025

• Next week should be a boring start to a tumultuous year for the Fed

• Navigating the economic effects of trade policies will be tricky once again

• And the new administration may try to use various levers to influence monetary policy

In this note we preview what we expect from the FOMC for next week, and for the rest of the year. After getting through with next Wednesday’s expected non-event of a meeting, the Fed will have to juggle three issues over the remainder of the year. First, it will have to deal with the usual balancing act of trying to achieve full employment and price stability. The policy debates in the second half of last year served as a reminder that even in a relatively benign environment that is not an easy task. Second, the FOMC will have to decide whether and how to incorporate Trump policies, particularly on trade, into their policy deliberations. Third, the Fed will likely have to deal with Trump’s efforts to influence monetary policy, both through appointments and potentially through other efforts to exert more sway on the institution.

The week ahead…

But before getting to those issues, next week the FOMC has the quotidian task of getting through a meeting that is largely expected to be a non-event. Various Fed communications over the past month have indicated policymakers’ desire to step away from rate cuts and take a pause at next week’s meeting; we and the markets also look for no action. In December the post meeting statement adopted the forward guidance that the Committee would consider “the extent and timing” of additional adjustments to the funds rate—a less committal form of forward guidance than earlier in the rate cutting campaign. We think that given the uncertainties in the outlook this watered down guidance remains appropriately bland.

We expect other, relatively insignificant, edits to the statement. For example, the last statement discussed the state of labor markets “since earlier in the year,” wording which no longer makes sense. We expect the updated assessment of current conditions to reference “solid” activity growth, an unemployment rate that is low, and inflation that has made progress but is still “somewhat elevated.” We expect no mention of trade or other policy issues looming on the horizon.

Powell’s post-meeting press conference has often stolen the show on FOMC day in recent years. For next week, however, we expect he will adopt more of a “duck and cover” approach In particular, we anticipate he will indicate that each Committee participant is using their own conditioning assumptions on what trade policies are ultimately adopted, and that the only thing decided at the meeting was the monetary policy statement agreed to next Wednesday. We don’t think he will foreclose the possibility of action at the March meeting—there’s too much that could happen between now and then to make a wager that the Committee will also step aside at that meeting.

…and the rest of the year

The Fed enters 2025 with the economy in a very good place. After rising for much of the first half of 2024, the unemployment rate has stabilized more recently, ranging in the low 4s for the last six months. Some indicators of labor demand, such as job openings, suggest the cooling in the job market has leveled off. Other indicators, such as gross hiring rates, warn against calling the “all clear” on the labor market. Inflation has also been getting closer to the Fed’s mandate. While the 2.8% over-year-ago rate on core PCE is still too hot for comfort for most of the FOMC, the three- and six-month annualized run rate of 2.3% shows welcome further progress; over-year-ago comparisons are about to get more favorable as well.

In this event a Fed sitting on the sidelines looks superficially like it’s following an “opportunistic disinflation” approach. However, the Fed isn’t simply waiting for good luck to bring inflation down further. At 4.33%, the funds rate is about 130 basis points above what the median Committee participant thinks is neutral for the economy and about 70 basis points above the top of the central tendency of participant estimates of neutral. Chair Powell has recently described this policy stance as “meaningfully restrictive.”

If the labor market were to begin softening again, we believe the Committee would quickly move to lower rates to at least these estimates of neutral. More recently, however, questions have arisen as to what gets the Fed thinking about reversing direction and hiking again. Looking back to 2021 may offer some lessons. Fed leadership began the year by noting they expected “transitory” inflation. By November Chair Powell “retired” that phrase, in part because over the intervening several months the labor market tightened significantly, and higher wage inflation was threatening to entrench the rise in price inflation.

Likewise, we think that a pivot toward tightening this year would likely be preceded by a sense that the labor market was again getting out of balance, and thus risking a wage-price spiral. An unemployment back below 4% and a V/U ratio above 1.3 could get the Committee to talk about talking about rate hikes.

The seven-year itch

The above-discussed deliberations will be complicated by the backdrop of Trump’s potential economic policies. We don’t think immigration policies will have first-order implications for the Fed’s mandate. While lower immigration will reduce labor supply and job formations, it should also reduce consumer demand and hence labor demand. On net, we think it should be roughly neutral for labor market slack and inflation. Chair Powell’s remarks in testimony indicate that he shares this view, and the discussion of the staff’s inflation forecast in the latest FOMC minutes did not call out immigration policy.

That inflation discussion did call out trade policy, and we believe that this is the bigger wild card for the outlook. In most economic models tariffs raise domestic prices and depress economic activity, creating a dilemma for monetary policymakers. When the Committee looked at this issue in September 2018 the staff presented a policy option of “seeing through” the rise in prices; in other words, not reacting to what should be, in principle, a one-time increase in the price level rather than an enduring inflationary development. A major caveat is that at that time inflation had been running below 2% for much of the previous half decade, and so unanchoring inflation expectations to the upside seemed a manageable risk. After the experience of the past three years we doubt the Committee will take for granted the stability of inflation expectations.

Even with a backdrop of low inflation, there was a rich Committee debate before the 25bp cut in July 2019—a move largely in response to escalating trade tensions. A subsequent weakening in manufacturing, investment, and global growth motivated two further eases. One lesson from that period that still holds, we believe, is that the FOMC did not act entirely preemptively: the first cut happened well over a year after the first rounds of US and Chinese tariffs and retaliatory tariffs.

The center will hold.

Even the novel risk of a universal tariff—including against our free trading area partners—may not be the most unusual challenge the Fed faces this year. A reminder of this came Thursday when President Trump, speaking to the World Economic Forum, said “I’ll demand that interest rates drop immediately.”

The traditional way that the president exerts influence onmonetary policy is through appointments (Figure 1). The seven governors on the Federal Reserve Board in Washington can only be removed “for cause”—a phrase most legal scholars believe that the Supreme Court still interprets as some sort of criminal malfeasance. This “for cause” protection may limit how much Trump can remake the Fed Board. There’s more legal debate as to whether the president can remove a governor from a leadership position (i.e. Chair or Vice Chair). The recent decision by Governor Barr to step down as Vice Chair for Supervision but stay on as governor was historically unusual but shows a path which, if followed by the two others in leadership roles, would limit the president’s ability to reshape the Fed.

At a minimum, though, early next year Trump can fill the seat to be vacated by Governor Kugler. A few months later that person could be nominated for Chair, as Powell’s leadership term expires next May. Some of the names speculated for this role include Kevin Hassett, Kevin Warsh, and current governor Chris Waller. Other names floated in the media have included Marc Sumerlin, Larry Lindsey, and David Malpass.

More radical change would come via changes to the Federal Reserve Act. For example, Trump’s nominee to lead the Council of Economic Advisers, Stephen Miran, has proposed several changes to the Act, including shortening the terms of governors, nationalizing the reserve banks, and changing the voting structure of the FOMC. Most of these would require 60 votes in the Senate. Not only would it be difficult to get Democrats to go along with these changes, but the Fed has many supporters among Senate Republicans. In the very few times the Federal Reserve Act has been re-opened since 1935, the fundamental governance of monetary policymaking has been unchanged. We expect that will also be the case over the coming four years.

Of course, the president could also try to harangue the Fed into doing his will. That could work. Or that could be counter productive, as the Committee bends against appearing to lose its independence. At any rate, it was hard to discern in Trump I that either the Yellen or Powell Fed departed in an obvious way from acting in accord with its usual responses to economic developments.