\begin{table}[htbp]
\centering\footnotesize
\caption{Frozen-first-release (one-sided filtered) policy: mean absolute gap of the monthly log change (×100) between the real-time value and the final smoothed estimate}\label{tab:revisions_frozen}
\begin{threeparttable}

\begin{tabular}{lrrrr}
\toprule
Class / level & Rel. 1 vs final & Rel. 12 vs final & 12-m change, rel. 1 & Rel. 1 vs last real-time \\
\midrule
house / province & 0.16 & 0.15 & 0.13 & 0.16 \\
house / region & 1.93 & 1.94 & 2.24 & 1.78 \\
house / municipality & 2.31 & 2.43 & 2.42 & 1.61 \\
condo / province & 0.16 & 0.16 & 0.25 & 0.16 \\
condo / region & 0.98 & 0.98 & 1.93 & 0.61 \\
condo / municipality & 1.82 & 1.87 & 3.59 & 1.07 \\
plex / province & 0.26 & 0.24 & 0.35 & 0.26 \\
plex / region & 1.37 & 1.41 & 2.05 & 0.86 \\
plex / municipality & 3.03 & 3.00 & 5.26 & 1.99 \\
\bottomrule
\end{tabular}
\begin{tablenotes}\footnotesize\item Under this policy a month's value changes only through stage-1 re-estimation; the gap to the final series is the price of never revising (filter versus smoother), and it does not decline with later releases.\end{tablenotes}
\end{threeparttable}
\end{table}