ECB Consumer Expectation Survey

The European Central Bank (ECB) runs the monthly Consumer Expectations Survey (CES) from Frankfurt. Ipsos, an external firm, manages the What You Expect panel. It covers about 19,000 adults aged 18 and over in 11 large euro area countries.

A 20-minute core runs monthly, plus a 10-minute follow-up. It rotates across housing (February), labour (May), central-bank topics (August) and detailed finances (November).

Piloted in January 2020, aggregate indicators run from April 2020. It tracks beliefs about past and future inflation, income, spending, growth, unemployment, housing and credit. It is not a consumer confidence index. Results support ECB policy analysis, not the views of its decision-making bodies.

Each round opens on the first Thursday of the reference month and closes on the Tuesday before the next round. About 70% of replies arrive in the first 10 days.

Results and a press release appear in the last 10 days of each month, 4 to 6 weeks after the reference month. Four reports plus aggregate datasets accompany each release, and anonymised microdata follow quarterly: January, April, July and October. Timing has no fixed day or time; aggregates are corrected and republished when found wrong.

Replies fall into four themes, mixing numeric and qualitative questions.

• Inflation perceptions and expectations: believed past inflation, expected inflation over 12 months and 3 years, plus a probabilistic 100-point split.
• Economic growth expectations: expected GDP growth over 12 months.
• Labour market and unemployment: perceived and expected unemployment, plus quarterly job-loss, job-finding and job-search odds.
• Household income and spending: perceived past and expected spending growth plus expected income growth over 12 months, with probabilistic income bins.
• Housing and mortgage rates: expected home-price growth over 12 months with its probabilistic split, and the mortgage rate in 12 months.
• Credit access and applications: whether credit felt harder or easier, and credit applications in the past 3 months.

The October 2023 release, published December 5, 2023, put perceived inflation at 7.8%, down from 8.0%. Twelve-month expectations held at 4.0% and 3-year at 2.5%. Expected nominal income growth was 1.1%, down from 1.2%.

Perceived past spending growth was 6.3%, up from 6.2%, while expected spending growth was 3.3%, down from 3.4%. Expected GDP growth was -1.3% from -1.2%; expected unemployment was 11.4% and perceived 10.9% from 11.1%. Job-finding odds rose to 27.7% from 22.9% and job-loss odds to 8.8% from 8.2%.

Expected home-price growth was 2.0%, down from 2.2%, and expected mortgage rates were 5.4%, a survey high. Credit access tightened to new highs as applicants rose to 17.7% from 14.8%, a high.

1. Sample filter: only replies with a completed background questionnaire count, pooled across countries or split by group.
2. Weighting: core weights adjust monthly replies and quarterly weights adjust quarterly ones to mirror the population. Only population weights are used.
3. Net percentage: for up-or-down questions, the increase share minus the decrease share. Weighted replies of increase, increase, same, decrease and increase give 66.7% minus 16.7%, a net +50 points.
4. Winsorised mean: numeric answers are capped at the 2nd and 98th percentiles, then averaged. For example, 2, 3, 3, 4 and 30 cap 30 at 4 and average near 3.3%.
5. Interpolated median: the midpoint between discrete answers resists outliers. Its interpolated median is about 3.4%, between the raw median of 3% and the mean.
6. Spending brackets: Don't know replies take the median answer inside each bracket.
7. Probabilistic bins: respondents split 100 points across inflation, home-price or income ranges, giving an implied mean and implied uncertainty per reply. Results aggregate as sample medians. One or two bins use a triangle, and three or more a beta fit.
8. Publication thresholds: series under 20 replies stay unpublished, 20 to 49 are flagged "u" (unreliable), and 50 or more carry no flag.