1\section{Global evidence on housing characteristics}2\label{sec:evidence}34\subsection{The evidence base}56The coefficient synthesis rests on \nStudiesExtracted{} eligible studies from \nCountriesExtracted{} countries,7\nModelsExtracted{} model columns and \nCoefExtracted{} coefficient rows, of which \nUsableEstimates{} harmonised estimates from8\nUsableStudies{} studies enter the pooled analysis (Table~\ref{tab:studychar}). Europe (\nStudiesEurope{} studies), Asia9(\nStudiesAsia{}) and North America (\nStudiesNorthAmerica{}) dominate; South America (\nStudiesSouthAmerica{}), Africa10(\nStudiesAfrica{}) and Oceania (\nStudiesOceania{}) are thinly covered (Figure~\ref{fig:map}, Table~\ref{tab:geography}). The11United States accounts for \shareUSextracted{} percent of extracted studies, against the near-totality of the Sirmans programme.12Condominiums and apartments (\nStudiesCondo{}) are as numerous as single-family houses (\nStudiesSingleFamily{}), a direct13consequence of the growth of the Asian and European listing-data literature; \nStudiesTransaction{} studies use transaction14prices, \nStudiesListing{} listing prices, \nStudiesRent{} rents and \nStudiesAppraisal{} appraised values. Ordinary least squares15provides \shareOlsEst{} percent of usable estimates, spatial estimators \shareSpatialEst{} percent and quantile regressions16\shareQuantileEst{} percent. The median study has \medianNobs{} observations. Figure~\ref{fig:heatmap} and17Table~\ref{tab:most_studied} show which characteristics are studied where: living area, bedrooms, bathrooms, age and distance to18the centre are reported almost everywhere, while most interior, market and AI-derived characteristics appear in fewer than ten19studies (Table~\ref{tab:gaps}).2021\begin{table}[H]22\centering23\caption{Characteristics of the extracted studies}24\label{tab:studychar}25\small26\input{tables/tab03_study_characteristics}27\end{table}2829\begin{figure}[H]30\centering31\includegraphics[width=\textwidth]{fig04_variable_frequency_heatmap.pdf}32\caption{Variable-frequency heatmap: number of extracted studies reporting each characteristic, by continent (thirty most33frequent characteristics).}34\label{fig:heatmap}35\end{figure}3637\begin{table}[H]38\centering39\caption{Most frequently studied characteristics (extracted studies; share of positive printed coefficients is a sign count40across all models and is not a pooled estimate)}41\label{tab:most_studied}42\footnotesize43\input{tables/tab05_most_studied}44\end{table}4546\subsection{Pooled implicit prices}4748Tables~\ref{tab:premiums} to~\ref{tab:elasticities} and Figures~\ref{fig:forest_premium} to~\ref{fig:forest_elasticity} report49the pooled estimates for the \nPooledFamilies{} characteristic--family combinations with at least five independent studies. Three50regularities organise the results. First, the structural core of the hedonic model is remarkably stable across the world: the51elasticity of price with respect to living area is \estLivingAreaElasticity{} (95 percent confidence interval52\loLivingAreaElasticity{} to \hiLivingAreaElasticity{}, $k=\kLivingAreaElasticity$ studies), each additional bathroom adds53\estBathroomsSemiElasticity{} percent and each bedroom, conditional on area, \estBedroomsSemiElasticity{} percent, a year of age54subtracts \estAgeSemiElasticity{} percent, and a kilometre of distance from the centre subtracts \estCbdDistanceSemiElasticity{}55percent, with the lot-size elasticity at \estLotSizeElasticity{} once living area is controlled. Second, the signature amenities56carry premiums of the order documented in the attribute-specific literature: a garage \estGaragePctPremium{} percent, a waterfront57location \estWaterfrontPctPremium{} percent ($k=\kWaterfrontPctPremium$), air conditioning, a view, a basement, a terrace or balcony58and a swimming pool between four and eleven percent, and proximity to hazardous or industrial sites a discount of about six59percent (Table~\ref{tab:premiums}). Third, every pooled family is heterogeneous beyond sampling error: the median $I^2$ is60\medianIsq{} percent and the 95 percent prediction intervals, which describe the implicit price a new study would be expected to61find, span zero for most characteristics even where the confidence interval of the mean excludes it. The mean implicit price of a62characteristic is therefore estimated with precision, but it is the centre of a wide distribution, not a parameter that transfers63across markets. The three-level estimates, which weight studies rather than estimates, and the cluster-robust standard errors,64which are typically two to four times the naive random-effects standard errors, are reported alongside.6566\begin{table}[H]67\centering68\caption{Pooled percentage premiums of binary attributes (log-price models; Kennedy-corrected; REML random effects, three-level69estimate and cluster-robust SE)}70\label{tab:premiums}71\scriptsize72\input{tables/tab06_pooled_premiums}73\end{table}7475\begin{table}[H]76\centering77\caption{Pooled semi-elasticities of continuous attributes (percent change in price per harmonised unit: m$^2$ for areas, km for78distances, years for age, counts for rooms, dB for noise, $\mu$g/m$^3$ for pollution)}79\label{tab:semielasticities}80\scriptsize81\input{tables/tab07_pooled_semi_elasticities}82\end{table}8384\begin{table}[H]85\centering86\caption{Pooled elasticities (log-log models)}87\label{tab:elasticities}88\scriptsize89\input{tables/tab07b_pooled_elasticities}90\end{table}9192\begin{figure}[H]93\centering94\includegraphics[width=0.95\textwidth,height=0.88\textheight,keepaspectratio]{fig05_forest_pct_premium.pdf}95\caption{Forest plot of pooled percentage premiums of binary attributes. Circles: REML random-effects estimates with 95 percent96confidence intervals; squares: three-level estimates; shaded bars: 95 percent prediction intervals.}97\label{fig:forest_premium}98\end{figure}99100\begin{figure}[H]101\centering102\includegraphics[width=0.95\textwidth,height=0.88\textheight,keepaspectratio]{fig06_forest_semi_elasticity.pdf}103\caption{Forest plot of pooled semi-elasticities (percent per harmonised unit).}104\label{fig:forest_semi}105\end{figure}106107\begin{figure}[H]108\centering109\includegraphics[width=0.95\textwidth]{fig07_forest_elasticity.pdf}110\caption{Forest plot of pooled elasticities from log-log models.}111\label{fig:forest_elasticity}112\end{figure}113114\subsection{Comparison with the Sirmans programme}115116Table~\ref{tab:sirmans} places the pooled estimates for the nine characteristics of \citet{sirmans2006value} next to the117qualitative findings that could be verified from that paper's abstract and references. The signs agree: living area, lot size,118bathrooms, garage, pool, fireplace and air conditioning are positive and age is negative; bedrooms, which Sirmans et al.\ found119of mixed sign, are positive on average here but with a prediction interval that includes large negative values, which is the120expected consequence of conditioning on living area in most specifications. What the present synthesis adds is the magnitude121with its uncertainty, the distinction between semi-elasticities and elasticities, and the global prediction interval; what it122cannot do is reproduce the Sirmans meta-regressions on raw coefficients, which are not comparable across functional forms.123124\begin{table}[H]125\centering126\caption{The nine Sirmans et al.\ (2006) characteristics: pooled estimates of this synthesis and the qualitative findings of the127earlier meta-analysis (abstract-level; the full text was not accessible)}128\label{tab:sirmans}129\scriptsize130\resizebox{\textwidth}{!}{\input{tables/tab14_sirmans_comparison}}131\end{table}132133\subsection{Summary}134135Figure~\ref{fig:summary} assembles the pooled percentage premiums and semi-elasticities in one view, coloured by taxonomy group.136Structural characteristics and signature exterior amenities have precisely estimated positive means; environmental disamenities137(hazardous sites, highways, noise, pollution) have negative means that are less precisely estimated and rest on fewer studies;138locational distances have negative semi-elasticities whose magnitudes depend heavily on the market's spatial structure. The139prediction intervals are the message for applied work: an appraiser or a policy analyst transferring an implicit price from the140literature to a particular market should expect the local value to lie anywhere within these intervals unless local evidence141narrows them.142143\begin{figure}[H]144\centering145\includegraphics[width=\textwidth,height=0.88\textheight,keepaspectratio]{fig14_summary_evidence.pdf}146\caption{Summary of the pooled evidence: REML estimates with 95 percent confidence intervals for all pooled percentage premiums147and semi-elasticities, coloured by taxonomy group.}148\label{fig:summary}149\end{figure}150