\begin{table}[htbp]
\centering\small
\caption{Distribution of reliability grades across cell-months (percent of cell-months)}\label{tab:grades}
\begin{threeparttable}

\begin{tabular}{lrrrrrr}
\toprule
Level / class & A & B & C & D & E & Cell-months \\
\midrule
mrc / condo & 10 & 13 & 23 & 10 & 45 & 5{,}600 \\
mrc / house & 10 & 47 & 33 & 8 & 2 & 7{,}140 \\
mrc / plex & 0 & 7 & 34 & 37 & 22 & 7{,}070 \\
municipality / condo & 3 & 12 & 14 & 12 & 59 & 17{,}150 \\
municipality / house & 0 & 2 & 5 & 10 & 83 & 77{,}630 \\
municipality / plex & 0 & 0 & 1 & 4 & 94 & 67{,}690 \\
province / condo & 100 & 0 & 0 & 0 & 0 & 70 \\
province / house & 100 & 0 & 0 & 0 & 0 & 70 \\
province / plex & 1 & 99 & 0 & 0 & 0 & 70 \\
region / condo & 32 & 24 & 30 & 1 & 12 & 1{,}120 \\
region / house & 57 & 33 & 7 & 3 & 0 & 1{,}190 \\
region / plex & 0 & 43 & 41 & 10 & 6 & 1{,}190 \\
\bottomrule
\end{tabular}
\begin{tablenotes}\footnotesize\item A: effective sample size $\ge 150$ and posterior s.e. of the log level $< 1.5$\%; B: $\ge 75$ and $< 2.5$\%; C: $\ge 30$ and $< 3.5$\%; D: $\ge 15$ and $< 5$\%; E otherwise (published for transparency, flagged).\end{tablenotes}
\end{threeparttable}
\end{table}