\documentclass[12pt]{article}
% Basic packages
\usepackage{geometry}
\usepackage{amsfonts}
\usepackage{amsmath}
\usepackage{amssymb}
\usepackage{graphicx}
\usepackage[small,bf,center]{caption}
\usepackage{subfig}
\usepackage{setspace}
\usepackage{float}
\usepackage[autostyle]{csquotes}
% Packages for automated tables
\usepackage{booktabs}
\usepackage{tabularx}
\usepackage{standalone}
\usepackage{pdflscape}
\onehalfspacing
\textwidth=6.0in
\textheight=8.5in
\begin{document}
\title{My Paper}
\author{Julian Reif\thanks{University of Illinois and NBER.}}
\maketitle
\begin{abstract}
\noindent
This paper provides an example of a document with tables and figures that were automated using Stata.
\end{abstract}
\clearpage
\section{Summary}
This example paper includes tables and figures that were created using Stata and outputted into the \textbf{/analysis/results} project folder. Copy the contents of that folder to \textbf{/paper} to update the tables and figures in this document.
Table \ref{tab:my_summary_stats} reports summary statistics for Stata's \textbf{auto.dta} dataset.%
%
\footnote{Type \textbf{sysuse auto, clear} at the Stata prompt to load this dataset.}
%
The average price of automobiles in this dataset is \$6,165. The price distribution, iillustrated in Figure \ref{fig:price_histogram}, is skewed right.
I estimate the association between automobile prices and fuel efficiency using the following linear model:
\begin{align}
PRICE_i&=\alpha + \beta X_i + \varepsilon \label{eqn:model}
\end{align}
The outcome variable, $PRICE_i$, is the price of automobile $i$. The parameter of interest is $\beta$, a vector of coefficients. In my first specification (``spec 1''), the vector $X_i$ includes miles per gallon. The second specification (``spec 2'') also includes the car's weight. I estimate this model using ordinary least squares and report standard errors that are robust to heteroskedasticity. The analysis is performed first using Stata, and then repeated using R.
Table \ref{tab:my_regressions} reports my Stata estimates, separately for domestic and foreign cars. Column (1) reports that an increase in fuel efficiency of 1 mile per gallon is associated with a \$329 reduction in the price of domestic automobiles. Column (2) shows that this association becomes positive and insignificant when I also include weight as a regressor. Columns (3) and (4) show that these associations are similar for foreign automobiles.
Table \ref{tab:my_regressions_with_r} compares these Stata estimates to estimates from R. Panel A reproduces the Stata estimates that were presented in Table \ref{tab:my_regressions}. Panel B of Table \ref{tab:my_regressions_with_r} reports estimates when I repeat this analysis in R using the \textbf{lm\_robust} command from the \textit{estimatr} package. The point estimates and the standard errors are identical across both software packages.
\clearpage
\section{Figures and Tables}
%%%
% Figure: Price histogram
%%%
\begin{figure}[ht]
\caption{Automobile prices}\label{fig:price_histogram}
\begin{center}
{\includegraphics[width=1\textwidth]{./figures/price_histogram.pdf}}
\parbox{\linewidth}{\footnotesize {Notes: Data were obtained from Stata's built-in auto dataset.}}
\end{center}
\end{figure}
\clearpage
%%%
% Tables
%%%
\input{tables/my_summary_stats.tex}
\input{tables/my_regressions.tex}
\input{tables/my_regressions_with_r.tex}
\end{document}