Repository Public Research preview
1\documentclass[12pt]{article}2\usepackage{booktabs}3\usepackage{graphicx}4\usepackage{natbib}5\title{Remote Work and Output: Evidence from a Firm Survey}6\author{A. Researcher}7\begin{document}8\maketitle910\begin{abstract}11We study how remote work relates to individual output using a cross-sectional survey of 1,250 participants in 38 firms.12Employees who work remotely at least three days a week report output that is 12.4\% higher than that of comparable office workers.13The gap is an association and we do not give it a causal reading.14\end{abstract}1516\section{Introduction}17\label{sec:intro}18Remote work expanded quickly over the last decade \citep{alpha2015, beta2020}.19Earlier studies disagree on whether it raises output \citep{gamma2021}.20We add administrative records of output to a survey of work arrangements.21Section~\ref{sec:data} describes the survey, Section~\ref{sec:method} sets out the model and Section~\ref{sec:results} reports the estimates.2223\section{Data}24\label{sec:data}25The survey was fielded in 2023 by a professional polling company (PPC).26Our sample includes 1,250 participants from 38 firms, and the response rate was 61\%.27Among them, 412 of 1,250 respondents (33.0\%) work remotely at least three days a week.28Table~\ref{tab:summary} reports summary statistics and Figure~\ref{fig:timeline} shows the timeline of the survey.29Output is measured in tasks completed per week, following \citet{gamma2021}.3031\begin{table}[t]32\caption{Summary statistics}33\label{tab:summary}34\begin{tabular}{lcc}35\toprule36Variable & Mean & SD \\37\midrule38Output (tasks per week) & 41.3 & 9.8 \\39Tenure (years) & 6.2 & 4.1 \\40Remote share & 0.33 & 0.47 \\41\bottomrule42\end{tabular}43\par\footnotesize Notes: The sample includes all 1,250 respondents.44\end{table}4546\begin{figure}[t]47\includegraphics[width=\textwidth]{timeline.pdf}48\caption{Survey timeline}49\label{fig:timeline}50\end{figure}5152\section{Empirical Strategy}53\label{sec:method}54We estimate55\begin{equation}56y_i = \alpha + \beta R_i + \gamma X_i + \varepsilon_i ,57\label{eq:main}58\end{equation}59where $y_i$ is log output, $R_i$ indicates remote work and $X_i$ collects controls.60Equation~\ref{eq:main} is estimated by ordinary least squares (OLS) with firm fixed effects.61Standard errors come from a bootstrap with 999 replications and the seed 20240117, computed in Stata 18.6263\section{Results}64\label{sec:results}65Table~\ref{tab:main} shows the main estimates.66Remote workers produce 12.4\% more than office workers ($t = 4.1$, $p < 0.001$).67The estimate changes little when controls are added, as column 2 of Table~\ref{tab:main} shows.6869\begin{table}[t]70\caption{Main estimates: log output on remote work}71\label{tab:main}72\begin{tabular}{lcc}73\toprule74 & (1) & (2) \\75\midrule76Remote & 0.124*** & 0.118** \\77 & (0.030) & (0.041) \\78Tenure & \multicolumn{2}{c}{0.019} \\79Observations & 1,250 & 1,250 \\80\bottomrule81\end{tabular}82\par\footnotesize Notes: Bootstrap standard errors in parentheses. *** $p<0.01$, ** $p<0.05$, * $p<0.10$.83\end{table}8485\section{Conclusion}86Remote work goes together with higher output in this survey.87The data and the replication code are available at a public archive described in the supplementary material.8889\bibliographystyle{plainnat}90\bibliography{refs}91\end{document}92