\begin{table}[!ht]
\caption{Overall spillover ATTs: analytic vs multiplier-bootstrap standard errors}\label{table:referee_boot}
\centering
\begin{tabular}{lcccc}\hline\hline
Neighbour outcome (log) & ATT (\%) & IF SE & Bootstrap SE & N \\ \hline
Population &      0.45 & (     0.31) & (     0.31) &       23,585 \\
Current expenditure &      1.08 & (     0.66) & (     0.67) &       23,581 \\
Education expenditure &      3.71 & (     1.01) & (     1.02) &       23,465 \\
Public-sector employment &      2.67 & (     1.29) & (     1.38) &       23,108 \\
Private-sector employment &     -1.05 & (     1.53) & (     1.54) &       23,541 \\
Number of firms &     -1.56 & (     1.00) & (     1.03) &       23,541 \\
Firms (services) &     -1.50 & (     1.05) & (     1.11) &       23,541 \\
Federal transfers (ex-FPM) &      4.45 & (     2.98) & (     3.08) &       23,522 \\
\hline\hline
\end{tabular}\par
\begin{tablenotes}\footnotesize
\item Overall Callaway--Sant'Anna ATTs, in percent (log coefficients \(\times 100\)): the simple weighted average of all post-treatment group-time effects, each cohort-by-year cell weighted by the cohort's share of the treated population (Section~\ref{sec:empirics_cs}). IF SE: analytic standard errors from the estimated influence function, clustered at the neighbour level (the paper's baseline). Bootstrap SE: multiplier (wild) bootstrap standard errors, 999 repetitions, seed 20260820, clustered at the neighbour level, as recommended by Callaway and Sant'Anna (2021). Estimation sample, cohort structure and not-yet-treated comparison group identical to the headline results (companion code: referee-round2b.do).
\end{tablenotes}\end{table}