Why Europe's Food Inflation Numbers May Be Missing the Real Story of Household Vulnerability

Rising food prices have become a defining pressure on household budgets across Europe, yet the relationship between inflation and actual food consumption is less straightforward than conventional economic expectations suggest. New research covering all 27 European Union member states finds no statistically significant systematic relationship between food-price inflation and inflation-adjusted household food expenditure. Instead, unemployment emerges as a more consistent indicator of changes in spending on essential food purchases.

The findings come from the study Food-Price Inflation and Household Food Consumption in the EU-27: Economic Vulnerability in European Food Systems, authored by Sabina Aliyeva, Zalán Márk Maró and Péter Czine, and published in the journal Agriculture on November 1, 2026. Examining European household expenditure and economic conditions between 2010 and 2024, the researchers question whether rising food prices alone adequately capture the financial pressures experienced by consumers.

Their analysis raises a distinction between what national spending statistics reveal and what households may experience. The absence of a statistically detectable relationship does not establish that food inflation is harmless. Instead, it exposes the difficulty of identifying changing consumption patterns, economic vulnerability and purchasing adjustments through aggregate expenditure data.

Europe's Food Inflation Shock Reveals a Deeper Consumption Puzzle

Europe's recent inflationary episode placed food affordability at the centre of economic concerns. Euro-area food inflation reached approximately 15.5% in March 2023, following overlapping disruptions involving agricultural production costs, energy prices, international commodity markets, supply chains and the Russia–Ukraine war. Although inflation subsequently eased, food-price movements remained uneven across European economies.

The pressure was especially relevant because food represents an essential component of household consumption. Consumers cannot easily eliminate food purchases when prices increase, but they can alter what they buy, where they shop and how they distribute expenditure across necessities. Consequently, higher food prices need not produce a proportional decline in inflation-adjusted food spending.

The researchers approach this problem by distinguishing food inflation from broader consumer-price movements. Their principal measure calculates the difference between annual food inflation and core inflation, which excludes food, energy, alcohol and tobacco. This identifies whether food prices are increasing faster or slower than underlying consumer prices rather than treating every inflationary episode as equivalent.

Using Eurostat data and supplementary FAOSTAT information, the study examines real per-capita household expenditure on food and non-alcoholic beverages. Its baseline analysis covers 361 country-year observations from 2010 to 2023, while limited additional observations from 2024 are incorporated into supplementary testing.

The statistical approach accounts for persistent differences between countries and common annual economic shocks. It also considers disposable income, unemployment, household size and demographic dependency. Separate models investigate whether material deprivation, income and the poverty-reducing impact of social transfers alter the relationship between relative food inflation and expenditure.

Rising Food Prices Show No Reliable Spending Link, but Joblessness Does

The key finding challenges the expectation that higher relative food inflation consistently accompanies lower real food expenditure. The baseline model estimates that a one-percentage-point increase in food inflation relative to core inflation is associated with approximately €1.44 less annual real food expenditure per person, measured in constant 2010 euros.

Although negative, this estimate is statistically insignificant and represents only about 0.08% of average expenditure in the sample. Its confidence interval extends from a reduction of €8.46 to an increase of €5.58, meaning modest associations in either direction remain compatible with the evidence. The research therefore cannot establish either a systematic decline or a complete absence of an underlying relationship.

Unemployment produces a substantially clearer statistical pattern. A one-percentage-point increase in the unemployment rate is associated with approximately €18.83 lower annual real food expenditure per person. Unlike the food-inflation estimate, this relationship is statistically significant and remains negative across the study's principal specifications.

The contrast draws attention to the different ways economic pressure can appear in household expenditure. Food inflation changes the relative cost of purchases, while deteriorating employment conditions can constrain the resources available for consumption. The findings support closer examination of labour-market conditions when interpreting food expenditure, although the observational analysis cannot establish causation.

The researchers also test whether higher income cushions households against food-price pressure. The estimated interaction suggests a weaker negative association at higher disposable-income levels, consistent with the original hypothesis. However, the result lacks statistical significance, preventing a firm conclusion about systematic income-related differences.

Material deprivation and social transfers produce similarly inconclusive results. Neither significantly modifies the relationship between relative food inflation and real expenditure, while both estimated interactions have signs opposite to those hypothesised. These findings do not demonstrate that deprivation is irrelevant or that welfare programmes fail to protect households.

National indicators cannot adequately reveal how economic pressures differ between individual households. A country's average income, deprivation rate or poverty reduction through transfers may conceal substantial variation in financial resources and purchasing behaviour. The study therefore leaves unresolved whether vulnerable population groups experience stronger consumption responses than national averages suggest.

Stable Food Expenditure Can Conceal Changes in Purchasing Behaviour

The research identifies an important limitation in interpreting aggregate consumption statistics. Real food expenditure measures inflation-adjusted spending, not physical quantities consumed, dietary quality or nutritional intake. Similar expenditure outcomes can therefore coexist with different combinations of products, purchasing decisions and consumption patterns.

Existing research examined by the authors indicates that consumers may respond to higher prices through product substitution, discounted purchases, retailer changes and broader household budget adjustments. Some households may preserve food quantities by selecting cheaper products, while others may reduce non-food expenditure to maintain essential purchases.

Such adjustments offer possible explanations for why food-price pressures do not consistently translate into detectable expenditure changes. However, the researchers cannot determine whether these mechanisms explain their empirical findings because the dataset does not track individual purchases, food-basket composition or product quality.

The distinction is relevant for agricultural producers, food manufacturers and retailers. Changes in consumer preferences and substitution between products can influence the composition of demand even when national expenditure figures show limited movement. Identifying such market adjustments would require more detailed purchasing and product-level information.

The researchers reinforce their central result through twelve alternative statistical specifications. These examine different food-price indicators, income measures, additional economic controls, delayed responses and alternative food-supply data. None establishes a statistically significant systematic relationship between relative food inflation and real household food expenditure.

Additional tests address unusually high relative food inflation in Bulgaria, Hungary, Latvia, Lithuania, Romania and Slovakia during 2022. Removing those observations or limiting their statistical influence does not fundamentally change the findings, suggesting the result is not driven exclusively by exceptional inflationary episodes.

Nevertheless, important uncertainty remains. The expenditure dataset is complete across all 27 countries only through 2022, with observations available for ten countries in 2023 and three in 2024. This limits the ability to examine longer-term household responses to the latest inflation shock.

Food Affordability Policy Needs More Than Inflation Statistics

For policymakers, the research supports examining food-price movements alongside employment conditions and broader household resources. Inflation statistics identify changes in prices, but they cannot independently establish whether households are reducing consumption, substituting products or experiencing worsening food-related deprivation.

The consistent negative association between unemployment and real food expenditure strengthens the case for including labour-market indicators in assessments of household economic vulnerability. However, the findings do not establish that unemployment is universally more important than food prices or provide a basis for ranking specific policy interventions.

Social protection presents another unresolved question. The absence of a significant interaction involving national poverty reduction through transfers does not establish whether particular benefits effectively protect food consumption. Evaluating such programmes requires information about recipients, coverage, household circumstances and consumption outcomes beyond the study's national indicators.

The study also identifies a measurement challenge for understanding food-system resilience. National expenditure aggregates can describe broad economic patterns but cannot show whether consumers maintain nutritional intake, alter dietary quality or substitute across food categories. These distinctions remain central to assessing the consequences of price shocks.

More detailed household spending data, higher-frequency observations and extended post-inflation coverage could reveal whether consumers cope with rising prices by spending less, switching to cheaper alternatives or reducing the quantities they purchase.