% Author: Simon-Pierre Boucher — contact@spboucher.ai
% ============================================================================
\section{Data}
\label{sec:data}
% ============================================================================

\subsection{Source and coverage}
\label{subsec:data_sources}

The analysis draws on a de-duplicated snapshot of the Canadian Multiple Listing Service
(MLS), \texttt{realtor\_mls\_unique.duckdb}, containing 172{,}019 unique active
``for-sale'' listings. Each record carries the list price, a structured set of building
and lot attributes, ownership and dwelling-type fields, and---crucially for this
study---a geocoded location (latitude, longitude and postal code). Geographic coverage is
essentially complete: coordinates and postal codes are present for more than 99.9\% of
records, so every listing can be mapped to its Forward Sortation Area (FSA), the first
three characters of the Canadian postal code and our unit of neighbourhood.

\subsection{Sample construction and variable parsing}
\label{subsec:variables}

The raw fields are semi-structured and require parsing. Bedroom counts reported as
``3~+~1'' (main plus lower level) are summed. Living areas, reported inconsistently in
square feet or square metres, are harmonised to square metres (preferring the explicit
floor-area measurement and converting from square feet at
$1\text{ ft}^2 = 0.0929\text{ m}^2$). Lot sizes are parsed from free-text fields into
square metres where a numeric value can be recovered, with a missingness indicator
retained. We restrict the sample to residential dwellings (houses, condominiums, plexes,
townhouses and apartments) with a strictly positive list price and non-missing core
structural fields, and we trim the extreme 1\% tails of price and living area to limit the
influence of data-entry errors and ultra-luxury outliers, a standard precaution in hedonic
work with listing data \citep{malpezzi2003hedonic}. The resulting estimation sample
contains \textbf{140{,}931 dwellings}, of which 82{,}334 are houses and 57{,}857 are
condominiums.

The dependent variable is the natural logarithm of the list price. Structural regressors
are: $\ln$ living area; counts of bedrooms, full bathrooms and half bathrooms; parking
spaces; storeys; and an indicator for the availability of lot information together with
$\ln(1+\text{lot area})$. Categorical controls comprise the dwelling type (house,
apartment/condo, row/townhouse, duplex, triplex, fourplex, manufactured home, other), the
ownership form (freehold, condominium/strata, leasehold, etc.), and the broad listing
category. The neighbourhood fixed effect is the FSA; FSAs with fewer than 25 listings are
pooled into a province-level residual category so that each absorbed effect is estimated
from a reasonable number of observations.

\subsection{Descriptive statistics}
\label{subsec:summary_stats}

Table~\ref{tab:summary_stats} reports the descriptive statistics. The median dwelling
lists for roughly \$565{,}000 and offers about 125~m$^2$ of living space, two full
bathrooms and three bedrooms. Both price and price per square metre are strongly
right-skewed---means exceed medians throughout---motivating the log transformation
illustrated in Figure~\ref{fig:price}, whose right panel is approximately symmetric and
underpins the semi-logarithmic hedonic specification.

\input{../results/tables/summary_stats}

\begin{figure}[t]\centering
\includegraphics[width=\textwidth]{fig_price_dist.png}
\caption{Distribution of list prices. Panel (a) shows the raw price (truncated at \$3M for
readability) with the sample median marked; panel (b) shows log price, the dependent
variable, which is close to symmetric.}
\label{fig:price}
\end{figure}

The provincial composition is dominated by Ontario, Quebec, British Columbia and Alberta,
which together account for the large majority of listings; the Atlantic provinces and the
Prairies are present but thinner, and Prince Edward Island and the territories contain no
listings in this snapshot. Median price per square metre ranges from roughly \$2{,}000 in
Newfoundland and Labrador to over \$6{,}100 in British Columbia
(Figure~\ref{fig:province}), foreshadowing the dominant role of location documented below.

\begin{figure}[t]\centering
\includegraphics[width=0.82\textwidth]{fig_province_ppm2.png}
\caption{Median price per square metre of living area by province. Grey extensions show
the mean--median gap, reflecting right-skew from high-value listings.}
\label{fig:province}
\end{figure}