Patrik Šmid

Slovenia: purchasing power & housing affordability

Annual price and wage data through 2025; median-income comparisons through 2024.

Real purchasing power of the average net wage: +19% over 10 years, +35% over 20, +72% over 30.

Housing affordability: several questions, several measures

Access to housing has worsened since 2015. Home prices, a standardised new mortgage payment and paid rents all rose relative to average net wages. An alternative median-income denominator also shows deterioration, though the size differs. Existing households’ measured housing-cost stress has fallen. These findings can coexist.

National annual data, retrieved 9 October 2026. Wage-based measures end in 2025; median-income comparisons end in 2024. Higher cost-to-income indices mean worse affordability. They measure changes, not the share of income needed to buy a particular home.

Home price / net wage+38.3%2015 → 2025
25-year payment / net wage+41.9%2015 → 2025 · scenario
Paid rent / net wage+11.8%2015 → 2025 · rent-index proxy
Market-rent housing overburden19.7%2025 · 25.9% in 2015

Low rates improved mortgage cash flow between 2007 and 2021. They did not reduce the deposit needed for a given home price. By 2025 the annual lending rate was slightly above its 2015 level, so financing added to the increase in the price-to-wage ratio. Mortgage repayments are a cash-flow measure; principal repayment also builds equity.

Mortgage assumptions and formula

Monthly payment = loan × i / [1 − (1 + i)−n], where i = annual interest rate / 12, and n = term in months. The index assumes the same loan-to-value ratio in every year (for example, 80%), a 25-year amortising loan, and each year’s ECB rate held constant for the term. A constant loan-to-value cancels in the index; it does not cancel in an actual household budget.

The ECB series combines fixed and variable products and fixation periods. The annual rate is an unweighted mean of twelve monthly reported new-business rates. This scenario excludes fees, insurance, taxes, maintenance, refinancing and rate resets. It describes how financing terms affect a comparable new loan, not payments on the stock of outstanding loans.

The starting year changes the answer

2015 was near a housing-market trough. We also show the earliest available house-price year, 2010 and the low-rate year 2021. A positive change below means greater cost relative to wages.

Change in cost-to-wage ratios, ending in 2025
ComparisonHome price / wage25-year payment / wagePaid rent / wage
2007 → 2025-5.8%-32.2%-6.4%
2010 → 2025+12.5%+8.1%+4.7%
2015 → 2025+38.3%+41.9%+11.8%
2021 → 2025+14.6%+30.0%+16.9%

The deterioration since 2015 is clear across these measures. A claim of continuous worsening over 20 or 30 years is unsupported: comparable national home-price data begin in 2007, and the 2025 purchase and payment ratios remain below that year’s levels.

Sensitivity to the mortgage term
Simulated mortgage-payment-to-wage change
Constant term2007 → 20252015 → 20252021 → 2025
20 years-28.5%+41.3%+27.0%
25 years-32.2%+41.9%+30.0%
30 years-35.5%+42.6%+32.9%

The direction of the conclusion survives 20-, 25- and 30-year terms. Each row keeps its term constant across years. Longer terms lower the monthly payment on a given loan but increase total interest and prolong repayment.

From 2015 to 2024, home prices rose 20.9% relative to median equivalised disposable income, versus 35.3% relative to average net wages. The direction agrees; the magnitude depends on whose income we measure.

Why the two income denominators differ

The wage measure covers employees of legal persons. Eurostat’s median equivalised disposable income covers residents in private households, including transfers and other income, adjusted for household size using the modified OECD scale. It is a median per equivalent adult, not a median household’s total income and not the income of prospective buyers or renters.

EU-SILC’s 2025 income observation refers to calendar-year 2024. We shift the income labels back one year before joining to home prices. We use euros, not PPS, to avoid introducing a separate price conversion. No 2025 median-income value is extrapolated.

From survey year 2015 to 2025, overburden fell from 6.1% to 3.5% for all residents, 11.5% to 2.3% for owners with a mortgage, and 25.9% to 19.7% for market-rent tenants. Renters remain much more exposed. These are percentages of people within each group, not percentages of their income.

Why overburden can improve while buying becomes harder

Eurostat overburden means housing costs above 40% of disposable income, with both costs and income net of housing allowances. Costs include rent, mortgage interest, utilities, regular maintenance, structural insurance, taxes and mandatory charges; mortgage principal is excluded. This differs from the full principal-plus-interest payment scenario.

Existing owners may have bought earlier, repaid debt or locked in lower rates. People unable to form a household or buy a home are not identified as prospective buyers by this indicator. National totals also combine very different tenure groups. We therefore cannot infer acquisition affordability from a low national overburden rate.

These are published EU-SILC point estimates, labelled by survey year. Income generally refers to the previous calendar year; cost components have their survey reference periods. Subgroup sampling uncertainty and changing composition matter. Standard errors are unavailable in this extract, so we do not attach invented confidence bands or claim statistically significant changes.

From 2015 to 2025, home prices rose 23.7% relative to the paid-rent index and 64.9% after CPI inflation. These describe valuation and price pressure. They do not, on their own, measure household affordability or prove a housing bubble.

What each metric can tell us

Separate acquisition, cash flow, ongoing costs and outcomes
MeasureQuestionEvidence and limits
Home price / incomeHow much purchasing capacity is needed?Derived index; average wage and median equivalised income alternatives. A national price index does not give a home’s price in euros or “years of income”.
Deposit / incomeHow hard is entry before borrowing?With an unchanged deposit percentage, this moves exactly like price / income. It is not an independent signal. Actual saving time needs savings, wealth, purchase costs and a target property.
Mortgage payment / incomeCan a new buyer service a comparable loan?Derived scenario; includes principal and interest. Rate, term and loan-to-value assumptions must accompany any absolute estimate.
Paid rent / incomeHas renting become more costly?Derived rent-index proxy. CPI rents reflect paid rents, including existing contracts, rather than a new-lease market index. The national wage is not renter-specific income; utilities are excluded.
Housing-cost overburdenWho actually faces high housing costs?Observed survey indicator, by tenure. Running costs included; mortgage principal excluded. Captures current residents, not all unmet housing demand.
Price / rent; real home pricesWhat is happening to valuation and prices?Context measures. Price / rent is not a rental yield because the price and rent indices do not represent the same dwelling. CPI deflation is distinct from income affordability.
Residual incomeIs enough money left for essentials?Income − all housing cash costs − a household-specific essentials budget. This is particularly useful for low-income households, but cannot be estimated credibly from these aggregate indices.
Economic user costWhat is the economic cost of ownership?Interest, opportunity cost of equity, upkeep, taxes, depreciation and expected capital gains. Expected appreciation and other assumptions are not observed here; we do not present interest alone as “implied rent”.

No composite affordability score. The indicators have different populations and units, and some are algebraically dependent. Combining them with arbitrary weights would create false precision. Nominal price / nominal income already compares purchasing capacity: deflating both by the same CPI leaves that ratio unchanged.

Exact accounting decomposition: prices, wages and financing

For an unchanged loan-to-value and term: Δln(payment / wage) = Δln(home price) − Δln(wage) + Δln(annuity factor). The entries below add exactly in log points, before rounding. Log points are not ordinary percentage-point contributions. This decomposition is an identity, not an estimate of causal effects.

25-year scenario: contributions in 100 × log changes
PeriodHome pricesWage growth offsetFinancingTotal
2007 → 2025+57.6-63.5-33.0-38.9
2015 → 2025+76.5-44.1+2.6+35.0
2021 → 2025+35.1-21.5+12.7+26.2

Since 2015, price growth outweighs wage growth; financing makes a small additional contribution. Since 2021, higher rates make a larger contribution. Since 2007, both wage growth and lower rates offset higher prices.

Econometric check: rate changes and affordability

We examine whether annual changes in mortgage rates are associated with annual changes in the price-to-wage ratio. The outcome excludes simulated mortgage payments, avoiding a mechanical relationship built into the payment formula.

100 × Δln(home-price index / net wage) = α + β × Δmortgage rate + ε

OLS, 18 annual changes (2008–2025). Wage-coverage changes and the pooled mortgage-product mix add measurement noise. The mortgage-rate change is in percentage points. The slope is β = +3.13 log points per +1 percentage point; the approximate 95% interval is -0.26 to +6.51, using Newey–West standard errors with one annual lag. The interval includes zero; this sample does not establish a reliable association at the 5% level.

This cannot estimate the causal effect of interest rates. Rates, wages and home prices respond jointly to economic conditions, credit supply and policy. Housing supply, population, expectations and delayed responses are omitted. A positive estimated slope does not mean raising rates causes home prices to rise.

Specification, uncertainty and sensitivity

We use log differences rather than regressing trending price and wage levels. Differencing reduces the risk of a spurious levels regression; it does not itself establish stationarity or solve endogeneity. No cointegration, structural break or causal identification claim is made. A longer consistent quarterly series and a defensible exogenous source of rate variation would be needed for stronger inference.

Newey–West covariance sensitivity: Bartlett kernel, n/(n−k) correction, t(16) reference
HAC bandwidthSlope βApprox. 95% intervalp-value
0 annual lags+3.13-0.69 to +6.950.102
1 annual lag+3.13-0.26 to +6.510.068
2 annual lags+3.13-0.18 to +6.430.062

R² = 0.21; lag-one residual correlation = 0.27. Dropping one annual observation at a time changes the slope from 0.81 to 4.86. These are sensitivity diagnostics, not multiple independent replications. Intervals are conditional on the specification and approximate in such a small sample; they do not capture input measurement errors, revisions or model uncertainty.

The directly computed affordability ratios and accounting decomposition carry the main conclusion. This exploratory regression does not explain away their changes and is not a forecast.

Evidence, reproducibility and remaining gaps

The analysis uses annual national series and preserves missing years. We check twelve monthly observations per annual rent and rate value, verify each 2015 index equals 100, align income reference years and verify the decomposition algebraically. Figures use the published point estimates; no model uncertainty is attached to the deterministic indices.

For a fuller assessment we would need regional transaction prices and new-lease rents, household incomes and liquid wealth, buyer age and tenure, mortgage contracts and a household-specific essentials budget. National averages cannot establish affordability for a young first-time buyer in Ljubljana or separate quality, location and composition effects.

Download annual indicators (CSV) · Full results and model diagnostics (JSON) · Archived inputs, calculation script and methods (ZIP)

Purchasing power and original price series

Purchasing power is the average net wage divided by the consumer price index (CPI). An index of 171.8 means the wage buys 71.8% more of the consumer basket than in 1995.

From 2015 to 2025, home prices rose 115%, net wages 55%, and consumer prices 30%. Homes became roughly 38% more expensive relative to wages. The housing trough around 2014–2015 matters: the price-to-wage ratio in 2025 was about 6% below its 2007 level.

Rates averaged 6.74% in 2008, 1.78% in 2021, 4.02% in 2023, and 2.90% in 2025. These are rates on new housing loans, combining fixed and variable products, excluding fees.

Sources and definitions