Household consumption: Heading for a rebound in the Eurozone and a slowdown in the United States?

Household consumption remains the main driver of growth in the Eurozone (where it has accounted for 54% of real GDP growth on average since 2015) and in the United States (where it accounted for 75% over the same period). However, in the Eurozone, its growth has slowed significantly since the health crisis and has remained subdued. In the United States, on the other hand, the adverse effects of the COVID-19 pandemic were quickly erased (see Chart 1).

As a result, the household savings rate in the Eurozone has remained exceptionally high. It peaked at 15.5% in Q2 2025, its highest level since the COVID pandemic, and is still 2.9 percentage points (pp) above its Q4 2019 level. Conversely, in the United States, the savings rate fell back below its historical average at the end of 2021 and has only seen a modest increase since then.

The reduction in inflation, coupled with the recent fall in real interest rates, have led to an increase in purchasing power in the Eurozone. However, the recovery in real wages has not compensated for the decline observed in 2022-2023, leading to persistently sluggish consumption in the Eurozone. In contrast, consumption in the United States has been significantly more vigorous, with wages rising faster than inflation since January 2023.

Nevertheless, other factors may explain why household consumption trends in the United States and the Eurozone have diverged so significantly over the past five years. This disparity, however, could diminish in the upcoming quarters.

Our analysis shows that the same variables – real gross disposable income, real housing wealth, real financial wealth and real 10-year interest rates – account for the changes in household consumption in both the United States and the Eurozone during the pre-COVID period. In Europe, the weakening of these fundamentals is the primary reason for the sluggish growth in consumption. In the United States, on the other hand, consumption has grown more than these fundamentals would suggest, largely due to Americans' increased propensity to consume housing wealth and the impact of the Biden stimulus. Over the coming quarters, we expect to see a correction in the underperformance of European consumption, although the extent of this correction will vary by country. In the United States, outperformance is expected to come to an end, but without any underperformance.

THE SAME VARIABLES PROVIDE A CLEAR EXPLANATION FOR CONSUMPTION IN THE EUROZONE AND ACROSS THE ATLANTIC DURING THE PRE-COVID PERIOD

In order to understand why US and European household consumption has differed so greatly during the post-COVID period, we conducted an econometric analysis focusing on the main theoretical factors influencing consumption. The findings (see box[1]) show that changes in household consumption in both the United States and the Eurozone can be attributed to the same variables: real gross disposable income of households, real housing wealth, real financial wealth and the real 10-year interest rate[2] .

Over our estimation period (Q1 2002 - Q4 2019), which predates COVID, our model essentially explains the trends in consumption in both the United States and the Eurozone, as shown in Charts 3 and 4.

RECENTLY, THE PERFORMANCE GAP BETWEEN THE EUROZONE AND THE UNITED STATES HAS WIDENED

THE DETERIORATION IN FUNDAMENTALS LARGELY EXPLAINS THE SLOWER GROWTH IN CONSUMPTION IN THE EUROZONE

Unsurprisingly, our model has encountered greater challenges in accounting for household consumption during the pandemic, owing to the extreme volatility caused by COVID-19. Nevertheless, its performance has improved since then.

The recent slowdown in household consumption growth (+3.1% between Q4 2019 and Q4 2024), in contrast to the period preceding the health crisis (+8.1% between Q4 2014 and Q4 2019), is primarily due to the poor performance of the underlying variables (see Chart 5): real financial wealth growth has slowed significantly compared with pre-COVID levels, as has housing wealth growth. In addition, our model has slightly overestimated consumption in the post-COVID period (by nearly 0.5 pp on average between Q4 2019 and Q4 2024). Nevertheless, this overestimation by the model is gradually diminishing (see Chart 4), indicating that fundamentals are increasingly accounting for consumption.

As was the case prior to the pandemic, real gross disposable household income continues to be the principal factor affecting consumption in the post-COVID period. However, its impact has diminished in comparison to the preceding period (4.0 pp from 2019 to 2024, as opposed to 4.8 pp from 2014 to 2019). Meanwhile, the influence of wealth effects remains less significant than that of disposable income and has also declined in the post-COVID period. However, the influence of financial wealth (+0.3 pp compared to +1.6 pp) is now less significant than that of housing wealth (+0.5 pp compared to +1.2 pp). Finally, 10-year interest rates[3] have had a negative impact in the recent period (-0.5 pp), in contrast to the pre-COVID period, during which they made a strong positive contribution (+0.9 pp). Consequently, the rise in interest rates in response to the inflationary crisis has resulted in a bolstering of savings.

Ultimately, in recent times, the four explanatory variables mentioned above have led to an overestimation of European household consumption by nearly 0.5 pp. This overestimation, which can also be seen as underconsumption by European households during this period in relation to their typical consumption determinants, can, however, be attributed to the diversity among Eurozone countries.

A significantly greater increase in the savings rate in France

In the main Eurozone countries, the household savings rate remains well above its pre-COVID level (whether considering 2019 or the average for the 2010s). This trend can be explained, in particular, by a more favourable shift in the purchasing power of gross disposable income (GDI) compared to wages alone. This dynamic has had redistributive effects that have increased the savings rate. Indeed, GDI has benefited from a form of implicit indexation of social benefits, particularly pensions. However, pensioners have a higher savings rate than the European population average – nearly one-third of GDI in France, according to the Banque de France. Furthermore, the increase in property income in 2022-2023 primarily benefited the wealthiest households, whose propensity to consume is lower than that of the rest of the population.

The periods 2020-2023 and 2024-2025 show different dynamics. The former was characterised by a general increase in savings, observable across all countries and for most household categories. However, over the past two years, the savings rate has stabilised somewhat in Germany, Italy and Spain.

France is an exception: the savings rate there rose by nearly one and a half percentage points, a phenomenon entirely attributable to an increase in financial savings (savings invested in financial assets). The savings rate in France is now higher than in Germany. This increase in the savings rate in France also accounts for a hysteresis effect from the inflationary shock that is more pronounced than in other countries[4]: although inflation has fallen significantly, the shock remains fresh in the minds of French households.

A survey conducted by the European Commission among households revealed that the intention of French pensioners to save has continued to rise after the end of the inflationary period, even surpassing the levels observed during the pandemic.

Furthermore, this trend has not reversed in the past six months. A study by INSEE, which is based on survey data, concludes that the revaluation of pensions in 2024 (which is higher than current inflation, as it was indexed to the 2023 inflation rate, which was significantly higher) accounts for the increase in the savings rate among pensioners[5] . This increase is particularly pronounced among pensioners in the first income quintile, whose financial resources are more reliant on pensions, while wealthier pensioners have also benefited from higher property income. The shift in French households' disposable income towards social benefits in 2024-2025, along with the resulting increase in the savings rate, explain the underperformance of household consumption in France.

In Germany, the underperformance of consumption is equally notable (with a growth rate of 0.3% in 2024, compared with 0.4% in France), but the underlying causes differ. The intention to save has increased most significantly among workers, in an environment where inflation exceeds the European average (by 2 percentage points in cumulative terms between the end of 2021 and the end of 2024). More than elsewhere in Europe, wages have lost purchasing power[6], which accounts for the more moderate growth in household consumption compared to the period before COVID. This weakness may have been exacerbated by the decline in industrial employment. While employment in the service sector has held up better and the rise in unemployment has remained limited, the industrial sector (which accounts for nearly one-third of German GDP, including construction) continues to weigh on overall economic momentum.

In Italy and Spain, the underperformance of consumption is less pronounced. On average, the savings rate of Italian and Spanish households has increased less relative to its pre-COVID level than in the rest of Europe. In both countries, GDP and income growth remain close to pre-COVID levels, and households' propensity to consume has held up better. The reliance on labour income rather than property income has provided better protection during the inflationary period, particularly in Spain, where inflation was lower due to a more moderate rise in energy prices (the Iberian exception).

OVERCONSUMPTION HAS APPEARED IN THE UNITED STATES. WHY?

In contrast to what we are seeing in the Eurozone, household consumption growth has remained steady in the United States (+14.2% between Q4 2019 and Q4 2024, compared with +12.9% from 2014 to 2019).

However, our model cannot account for all of the recent growth, which exceeds our model's estimate by nearly 2.2 pp. According to our model, and similar to the Eurozone, real gross disposable household income is the variable that has contributed most to this change in consumption (+7.1 pp compared with +9.0 pp from 2014 to 2019), followed far behind by housing wealth (+2.6 pp compared with 2.5 pp). Real financial wealth also played a notable role (+2.3 pp compared with 3.0 pp). Finally, the real 10-year interest rate had a negative impact on consumption (-0.2 pp compared with 0.2 pp).

The usual interplay of fundamentals is therefore not enough to explain the dynamism of US consumption. Changes in behaviour have emerged since the pandemic, which are not factored into the modelling.

The role of housing wealth

According to an IMF study[7], this outperformance of consumption in the United States can be primarily attributed to a sharp increase in the marginal propensity to consume housing wealth. According to the authors, this propensity almost tripled in the post-COVID period compared with the pre-COVID period[8]. The increase in the value of American households' real estate assets has enabled them to obtain mortgages, which they have used to bolster their consumption.

To test this conclusion, we adjusted the marginal propensity to consume housing wealth in our model for the post-COVID period to the level indicated in the IMF study. This adjustment led to the elimination of the residual in our modelling (see Chart 9), thereby confirming the role of this variable in the outperformance of US household consumption in the post-COVID period.

This result also explains why the wealthiest American households were the primary drivers of consumption outperformance in the post-COVID period. According to Moody's[9], "consumers in the top 10% of the income distribution accounted for 49.2% of total spending in Q2 2025, up from 48.5% in Q1, marking the highest level since 1989".

The impact of the "Biden stimulus"

The fiscal policy implemented in the United States in response to the COVID-19 pandemic has contributed significantly to the relative outperformance of US household consumption, which continued for several quarters after the lifting of various restrictions. Fiscal data show a total expenditure of USD 4,530 related to this policy, representing more than 10% of GDP in 2020 and 2021.

This initiative was implemented in two major stages: first, the CARES Act was enacted in March 2020 to provide support and reassurance amid the pandemic outbreak. A year later, in March 2021, the Biden administration began with the enactment of the American Rescue Plan Act, which was primarily motivated by the need to support economic recovery. These initiatives predominantly offered direct assistance to households (USD 815 billion disbursed as Economic Impact Payments, direct and automatic payments), a temporary extension of unemployment insurance benefits, and tax credits.

The combination of these measures, coupled with a less favourable environment for consumption, resulted in a savings increase of USD 2.1 trillion in 2020 and 2021, according to the San Francisco Fed. It posits that this increase, which American households were able to draw on, enabled them to maintain their consumption despite the inflationary environment, before it was exhausted at the beginning of 2024. Furthermore, US fiscal policy continued to be expansionary after the pandemic period.

WHERE ARE WE HEADING?

TOWARDS A SCATTERED CORRECTION OF UNDERPERFORMANCE IN THE EUROZONE

Is household consumption poised for recovery in the Eurozone? Yes, although the answer differs from country to country.

France and Spain are expected to continue to diverge from the European average

France is expected to see very modest growth in household purchasing power and a slight decline in household savings rates. Indeed, GDP growth is projected to slow down in 2026-2027, as it will no longer benefit from the momentum of social benefits (even assuming that no measures are taken to under-index them), particularly pensions. Growth in disposable income would also receive less support from property income. The slower growth in income would be the main reason for the reduction in the savings rate.

However, the savings rate is expected to remain high. By 2027, it is projected to revert to its 2023 level and remain nearly 3 points higher than in 2019. The purchasing power of wages is likely to continue its growth trajectory as observed since 2024, albeit without accelerating, and therefore will not fully offset the losses incurred during the inflationary crisis. With a relatively resilient labour market, household confidence is expected to strengthen. Consequently, household consumption could potentially return to its 2022-2023 growth rate (+1% per year, compared with +0.4% in 2025) in 2026-2027, but without surpassing it. The least affluent households (those without employment or mainly dependent on income from work) will continue to face constraints on their consumption.

Spain is also expected to maintain its savings rate above pre-COVID levels, but for very different reasons and with distinct implications compared to France. GDP and GNI growth dynamics are expected to remain strong, leading to a net gain in purchasing power. As a result, household consumption is likely to remain vigorous, without the need for households to dip into their savings.

In contrast, German and Italian households are projected to see their savings rates negate the post-COVID increases by 2027

In both cases, these developments would coincide with an increase in household consumption.

In Germany, this trend would be bolstered by improved growth prospects, the resulting rebound in job creation and an increase in the minimum wage (+8.5% in 2026 and +5% in 2027). However, there could be a downside risk to GDP growth and purchasing power if the coalition government implements its plan to limit the increase in social benefits (to generate the savings needed to finance economic support measures). Some of these support measures are not exempt from the debt brake rule (including the measure to expedite depreciation of business investments, which would reduce corporation tax) and must therefore be counterbalanced by savings measures or new revenue streams to avoid exacerbating the public deficit.

In Italy, the reduction in household income tax (EUR -9 billion between 2026 and 2028) would help maintain purchasing power growth, despite a slowdown in wage growth. As this tax reduction is aimed at the middle classes, it would have a positive impact on their consumption and facilitate a decrease in the savings rate.

IS THIS THE END OF THE OUTPERFORMANCE IN THE UNITED STATES?

An unfavourable inflation-employment balance for consumption in the coming quarters

The current and anticipated rise in inflation in the United States is expected to negatively impact household consumption. Inflation has already increased (+3.0% y/y in September 2025 [+3.1% for the underlying measure], up +0.6 pp since April). In addition, we expect prices to continue rising due to customs duties. Inflation is projected to reach +3.4% y/y in Q2 2026 (+3.3% for the underlying measure), with no return to target over the forecast period (+2.9% y/y in Q4).

At this stage, survey data shed light on households' concerns about inflation. In the University of Michigan sentiment survey, the extent of the deterioration in households' inflation expectations is comparable in the short term (1 year) to what was seen during the post-pandemic inflationary surge. It is even more pronounced over a 5–10-year horizon, revealing doubts about the ability of inflation to return to a manageable level in the long term. While the latest episode of inflation, unprecedented in nearly 40 years, was followed by robust consumption growth, the situation in the labour market was quite different. The ratio of job vacancies to job seekers reached 2:1 in H1 2022. Such tight labour market conditions bolstered wages, contributing to sustained consumption. Currently, this ratio is just below 1. Under these circumstances, wage growth is likely to slow, adversely affecting US consumption. As a result, expectations of a rise in unemployment over the next year are as high as they were during the Great Financial Crisis, which may result in an increase in the savings rate.

In addition, the post-pandemic labour market has benefited greatly from an influx of workers into the United States, counteracting the trend known as the "great resignation". This shift enabled the labour force to surpass pre-COVID forecasts by 2023 (see Chart 13). This situation arises from both labour supply support and aggregate demand. The shift in migration policy, with the stated goal of "net-zero migration", is expected to lead to a slowdown in population growth, which in turn will slow consumption.

A fiscal policy with no effect on consumption, particularly among the lowest-income households

According to impartial studies, the Trump administration's fiscal policy is not redistributive in nature. The One, Big, Beautiful Bill Act (OBBBA), which extends the tax cuts from the 2017 Tax Cuts and Jobs Act (TCJA), broadens the standard deduction and introduces several complementary measures, including temporary tax exemptions for tips and overtime.

The Congressional Budget Office (CBO) estimates that, compared to legislation where the temporary measures of the TCJA would have expired at the end of the year, the OBBBA results in a benefit that increases with household income, except for those in the bottom two deciles of the income distribution, for whom the impact is negative[10] .

This contrasts sharply with the approach favoured by the Biden administration. Furthermore, the OBBBA primarily serves as an extension and expansion of the 2017 TCJA: (i) the law extends existing tax structures and therefore does not create new incentives; (ii) the tax cuts introduced by the Trump I administration in 2017 did not have a significant positive impact on employment or wage growth[11], which reduces the likelihood that the OBBBA will boost household demand.

The US administration anticipates that foreign companies will absorb the tariffs, which will subsequently fund the OBBBA's tax cuts. However, if the revenue from additional tariffs is not redistributed to groups with a high propensity to consume, it may be viewed, in part, as an inflationary tax that penalises the purchasing power and, consequently, the consumption of these groups, while the consumption of wealthier individuals remains intact or even favoured. Political awareness of this issue has prompted the administration to approve tariff reductions on food items and to introduce a "tariff cheque" worth USD 2,000.

Estimates from the Yale Budget Lab show a reduction in household disposable income for the first eight deciles of the income distribution, due to the combined impact of the OBBBA and customs duties, with the effect diminishing as income rises (Chart 14).

Consumption and real estate: towards a continuation of the trade-off

Surprisingly, the recent outperformance of US consumption has occurred despite the most extensive and rapid monetary tightening since the early 1980s. The Fed raised its target rate by +525 bp between March 2022 and July 2023, resulting in tighter financial conditions. This led to an increase in mortgage rates from 3% at the end of 2021 to 7-8% in 2023 (on average over 30 years). However, this shift in the interest rate environment has only resulted in a modest increase in mortgage debt servicing costs, which have risen from 5.6% to 5.9% of disposable income since Q1 2022 (i.e. their pre-pandemic level). Although homeowners have managed to sustain their consumption levels, house sales have plummeted since early 2023, falling significantly below the volumes seen after the onset of the 2008 financial crisis, which explains why the burden of mortgage debt servicing has increased only marginally.

However, the inclination towards consumption may continue by default. It seems unlikely that the current monetary easing (we anticipate two further rate cuts [-25 bp each] before mid-2026, following the total -50 bp cuts in September/October 2025) will provide a significant boost to the property market through lower borrowing costs. This is due to the fact that 10-year US Treasury yields, rather than the Fed Funds target, act as the main benchmark for mortgage rates. However, the transmission of Fed rate cuts to the long end of the yield curve has been inconsistent in recent quarters: despite a total 125 bp cut since September 2024, the 10-year yield on US Treasury bonds was over 40 bp higher than its pre-easing level in October 2024. According to our forecasts, this trend is likely to continue, with the 10-year yield projected to reach 4.5% by the end of 2026, despite a key interest rate of 3.5%; this aligns with a 30-year mortgage rate of 6-7%. This does not bode well for a rebound in residential property, particularly in light of increasing pressures on the labour market.

*****

Projections for household consumption are subject to significant uncertainty. This is particularly true in the United States, where consumption is more reliant than in Europe on wealth effects, which could be affected by a market downturn, particularly concerning the valuations of companies involved in the development of artificial intelligence. In Europe, these uncertainties seem to be more evenly distributed. However, the latest economic indicators confirm a notable improvement. Although still elevated, European households' concerns regarding the labour market have recently diminished and may continue to do so, as the labour market is proving resilient. On the other hand, the modest recovery in real wages (following a decline during the inflationary period) and the ongoing relatively high level of uncertainty could weigh on the prospects for a rebound in household consumption in Europe.

Article completed on November 3^rd 2025.

[1] The authors would like to thank François Faure for his contribution to the econometric analysis.

[2] Neither the unemployment rate (which may lead to precautionary savings) nor inflation (which can generate savings to offset the effects of real cash balances) emerges as an explanatory variable in the long-term relationship. These two variables sometimes appear significant and are therefore included as explanatory variables in short-term dynamics.

[3] The average yield on 10-year German bonds denominated in euros was used.

[4] See our Chart of the Week for 23 July 2025, Good news for the ECB: household inflation expectations have returned to normal, L. Barette.

[5] June 2025 Economic Note, ”L’épargne des ménages au sommet”.

[6] See our Chart of the Week for 8 January 2025, Has household purchasing power returned to its pre-inflation surge level?, L. Barette and H. Baudchon. Since then, the catch-up has improved by a few decimal points, except in Germany and Japan.

[7] Mai Dao, La-Bhus Fah Jirasavetakul, and Jing Zhou. Drivers of Post-COVID Private Consumption in the U.S., IMF Working Papers 2024, 128 (2024), accessed 31 October 2025, https://doi.org/10.5089/9798400281044.001

[8] The marginal propensity to consume housing wealth rose from 0.06 over the period 1990-2019 to 0.15 over the period 2020-2023.

[9] Top 10% of Earners Drive a Growing Share of US Consumer Spending - Bloomberg, September 2024.

[10] CBO, 2025 – Link

[11] Brookings, 2021 - Link