# How I can have the equal width of columns?

I am trying to have the equal width of two columns under 'Predictors' in Panel B.

I've checked other online question. I tried to use tabularx but it didn't work.

The following is my code.

\documentclass[12pt]{article}  %It can be article / report / and book
\usepackage{amssymb} %One of the math packages
\usepackage[top=1.3in, bottom=1.3in, left=1in, right=1in]{geometry}
\usepackage{setspace}
\usepackage{tabu}
\usepackage{pgflibrarysnakes}
\usepackage{graphics,graphicx,pgfplots,rotating,multirow}
\usepackage{xcolor}
\usepackage{amssymb,amsfonts,amsmath}
\usepackage{color,colortbl}
\usetikzlibrary{snakes}
%\usepackage{ltablex}
\usepackage{caption}
\usepackage{subcaption}
% Footnote space
\usepackage{lipsum}% dummy text

\begin{table}
\caption{Predicting Ex Post Covariances between between Stock Returns and Consumption Growth} \label{Tab6}
\vspace{0.3in}
\begin{spacing}{0.8}
{\footnotesize Table 6 reports }
\end{spacing}
\vspace{0.3in}
\begin{center}
\resizebox{5.5in}{!}{
\begin{tabu}{l|c|c|cc}

\multicolumn{5}{l}{\textbf{Panel A: 1st stage regression}} \\ \cline{1-5}

\T  Dependent variable & \multicolumn{2}{|l|}{Predictors} & $F$-statistic & $P$-Value \\ \cline{1-5}

$r^e_t$ & \multicolumn{2}{|l|}{$D/P_{t-1}$, $SVAR_{t-1}$, $BM_{t-1}$, $NTIS_{t-1}$, $LTY_{t-1}$}  & 4.45 & 0.001  \\
$\Delta c_{CE,t}$ & \multicolumn{2}{|l|}{$\Delta c_{CE,t-1}$, $\Delta c_{CE,t-2}$, $\Delta c_{CE,t-3}$, $r_{f,t-1}$}    & 43.67 & 0.000  \\
$\Delta c_{NIPA,t}$ & \multicolumn{2}{|l|}{$\Delta c_{NIPA,t-1}$, $\Delta c_{NIPA,t-2}$, $\Delta c_{NIPA,t-3}$, $r_{f,t-1}$}  & 65.14 & 0.000  \\
$\Delta c_{H,t}$ & \multicolumn{2}{|l|}{$\Delta c_{H,t-1}$, $\Delta c_{H,t-2}$, $\Delta c_{H,t-3}$, $r_{f,t-1}$}  & 19.11 & 0.000  \\ \hline \\ \\

\multicolumn{5}{l}{\textbf{Panel B: 2nd stage regression}} \\ \cline{1-5}
\T &  \multicolumn{2}{|c|}{Predictors} & & \\ \cline{2-3}
Dependent variable & $\hat \epsilon_{r^e,t-1} \hat \epsilon_{\Delta c_{i},t-1}$  & $M/C_{i,t-1}$ & $F$-statistic & $P$-Value \\ \cline{1-5}
\T  $\hat \epsilon_{r^e,t} \hat \epsilon_{\Delta c_{CE},t}$ & -0.143** & $1.06\times10^{-5}$ & 2.93 & 0.055  \\
\rowfont{\small}    & (-2.26) & (0.85) &  &  \\
\T  $\hat \epsilon_{r^e,t} \hat \epsilon_{\Delta c_{NIPA},t}$ & 0.108  & $-9.27\times10^{-6}$  & 1.34 & 0.264 \\
\rowfont{\small}    & (1.64) & (-0.23) &  &  \\
\T  $\hat \epsilon_{r^e,t} \hat \epsilon_{\Delta c_{H},t}$ & -0.133** & $-2.65\times10^{-5}$ & 2.28 & 0.105 \\
\rowfont{\small}    & (-2.12) & (-0.51) &  &  \\  \hline

\end{tabu}%
}
\end{center}
\end{table}
\clearpage

• Welcome to TeX SX: What is \T? – Bernard Jun 29 '17 at 19:56
• please make your code sniped compilable! missing are definition of \T, \begin{document} ... some packages are loaded twice ... – Zarko Jun 29 '17 at 20:00

1. I added some shortcuts, so the code is a bit easier to read.

• \mc for \multicolumn
• \dc for \multicolumn{2}{l|}{#1}
2. replaced \usepackage{tabu} by \usepackage{tabularx}

## The code (fixed my error messages)

\documentclass[12pt]{article}  %It can be article / report / and book
\usepackage{amssymb,amsfonts,amsmath} % math packages
\usepackage[top=1.3in, bottom=1.3in, left=1in, right=1in]{geometry}
\usepackage{setspace}
\usepackage{booktabs}
\usepackage{tabularx}
\usepackage{array}
\usepackage{graphics,graphicx,rotating,multirow}
\usepackage{xcolor}
\usepackage{color,colortbl}
%\usepackage{ltablex}
\usepackage{caption}
\usepackage{subcaption}
\begin{document}
\thispagestyle{empty}

\def\T{\texttt{T}}
\let\mc=\multicolumn
\newcommand\dc[1]{\mc{2}{l|}{#1}}
\newcommand\heps{\hat\epsilon}
\newlength\myLength
\setlength\myLength{3.7cm}

\begin{tabularx}{\linewidth}{p{2.7cm}|>{\centering}p{\myLength}|>{\centering}p{\myLength}|>{\centering}Xc}

\mc{5}{l}{\textbf{Panel A: 1st stage regression}}
\\ \cline{1-5}
\T-Dependent variable  & \mc{2}{c|}{Predictors}                                                          & $F$-statistic & $P$-Value \\
\cline{1-5}

$r^e_t$                & \dc{$D/P_{t-1}, SVAR_{t-1}, BM_{t-1}, NTIS_{t-1}, LTY_{t-1}$}                   & 4.45          & 0.001     \\
$\Delta c_{CE,t}$      & \dc{$\Delta c_{CE,t-1}, \Delta c_{CE,t-2}, \Delta c_{CE,t-3}, r_{f,t-1}$}       & 43.67         & 0.000     \\
$\Delta c_{NIPA,t}$    & \dc{$\Delta c_{NIPA,t-1}, \Delta c_{NIPA,t-2}, \Delta c_{NIPA,t-3}, r_{f,t-1}$} & 65.14         & 0.000     \\
$\Delta c_{H,t}$       & \dc{$\Delta c_{H,t-1}, \Delta c_{H,t-2}, \Delta c_{H,t-3}, r_{f,t-1}$}          & 19.11         & 0.000     \\
\hline

\mc{5}{l}{\rule{0pt}{1cm}\textbf{Panel B: 2nd stage regression}}
\\ \cline{1-5}
& \mc{2}{c|}{Predictors}                     &                      &               \\
\cline{2-3}
\T-Dependent variable                         & $\heps_{r^e,t-1} \heps_{\Delta c_{i},t-1}$ & $M/C_{i,t-1}$        & $F$-statistic  & $P$-Value \\
\cline{1-5}
\T  $\heps_{r^e,t} \heps_{\Delta c_{CE},t}$   & -0.143**                                   & $1.06\times10^{-5}$  & 2.93           & 0.055     \\
& (-2.26)                                    & (0.85)               &                &           \\
\T  $\heps_{r^e,t} \heps_{\Delta c_{NIPA},t}$ & 0.108                                      & $-9.27\times10^{-6}$ & 1.34           & 0.264     \\
& (1.64)                                     & (-0.23)              &                &           \\
\T  $\heps_{r^e,t} \heps_{\Delta c_{H},t}$    & -0.133**                                   & $-2.65\times10^{-5}$ & 2.28           & 0.105     \\
& (-2.12)                                    & (-0.51)              &                &           \\
\hline

\end{tabularx}%

\end{document}

• interesting idea, but MWE doesn9t work. Error: illegal pream-token (\myLenght) ... . Also you should also clean up OP preamble ... – Zarko Jun 29 '17 at 20:48
• @Zarko Thanks ! Didn't even notice and ran right through the errors... As to the preamble, that shall be for another day ! ;) Feel free to edit, if you know better than I do how to simplify all that... – marsupilam Jun 29 '17 at 21:14
• @chanik Hope I didn't messed up your stat'l data ! BTW, the SVAR spacing looks really terrible ! – marsupilam Jun 29 '17 at 21:26

• first I clean-up preamble: instead of \usepackage{graphics, graphicx, ...} is sufficient \usepackage{graphicx, ...}, instead of \usepackage{xcolor, color, colortbl} is sufficient \usepackage[table]{xcolor}, amsmath is loaded twice
• use of tabu is fragile; package is buggy and not unfortunately not maintained anymore
• for nicer numbers align in the last two columns is handy to use S column type from package siunitx; with it is also simpler to write numbers multiplied with 10^{...} (see MWE below)
• instead of (for example) $\hat \epsilon_{r^e,t-1} \hat \epsilon_{\Delta c_{i},t-1}$ is correct $\hat{\epsilon}_{r^e,t-1} \hat{\epsilon}_{\Delta c_{i},t-1}$
• it is not clear to me if SVAR, NIPA, ... are one variable or set of four; i assume the former and write them as \mathit{SVAR} etc
• for equal columns width you need to prescribe their width. You can do this with use of tabularx as suggested marsupilam in his answer, which I would change from

\begin{tabularx}{\linewidth}{p{2.7cm}|
>{\centering}p{\myLength}|
>{\centering}p{\myLength}|
>{\centering}Xc}


to

\begin{tabularx}{\linewidth}{l}|
*{2}{>{\centering}X|}
*{2}{S[table-format=2.3]}}


for use in my MWE below (and in preamble add package tabularx of course) or with use of p{...} column type as is done in MWE below:

\documentclass[12pt]{article}  %It can be article / report / and book
\usepackage[top=1.3in, bottom=1.3in, left=1in, right=1in]{geometry}
%\usepackage{setspace}
\usepackage{array, multirow, rotating}
\usepackage{graphicx}
\usepackage{amsmath, amssymb, amsfonts}
\usepackage[table]{xcolor}
\usepackage{graphicx}
\usepackage{pgfplots}
\usetikzlibrary{snakes}
\usepackage{caption}
\usepackage{subcaption}

\def\T{\texttt{T}}
\newcommand\mc[1]{ \multicolumn{1}{l}{#1}}
\newcommand\mcc[1]{\multicolumn{2}{>{\raggedright}p{82mm}|}{#1}}

\usepackage{lipsum}% dummy text

\begin{document}
\begin{table}
\caption{Predicting Ex Post Covariances between between Stock Returns and Consumption Growth}
\label{Tab6}
\centering
\renewcommand\arraystretch{1.2}
\begin{tabular}{
>{$}l<{$}|
*{2}{>{\centering}p{41mm}|}
*{2}{S[table-format=2.3]}}
\multicolumn{5}{l}{\textbf{Panel A: 1st stage regression}}                  \\
\hline
\text{Dependent variable}
&   \multicolumn{2}{l|}{Predictors} & {$F$-statistic} & {$P$-Value}    \\
\hline
r^e_t   &   \mcc{$D/P_{t-1}$, $\mathit{SVAR}_{t-1}$, $\mathit{BM}_{t-1}$,
$\mathit{NTIS}_{t-1}$, $\mathit{LTY}_{t-1}$}
&   4.45    &   0.001                                           \\
\Delta c_{CE,t}
&   \mcc{$\Delta c_{\mathit{CE},t-1}$, $\Delta c_{\mathit{CE},t-2}$,
$\Delta c_{\mathit{CE},t-3}$, $r_{f,t-1}$}
&   43.67   &   0.000                                           \\
\Delta c_{NIPA,t}
&   \mcc{$\Delta c_{\mathit{NIPA},t-1}$, $\Delta c_{\mathit{NIPA},t-2}$,
$\Delta c_{\mathit{NIPA},t-3}$, $r_{f,t-1}$}
&   65.14   &   0.000                                           \\
\Delta c_{H,t}
&   \mcc{$\Delta c_{H,t-1}$, $\Delta c_{H,t-2}$,
$\Delta c_{H,t-3}$, $r_{f,t-1}$}
&   19.11   &   0.000                                           \\
\hline
\mc{}   &   \mc{}   &   \mc{}   &   \mc{}   &   \mc{}                       \\
\multicolumn{5}{l}{\textbf{Panel B: 2nd stage regression}}                  \\
\hline
\T      &   \mcc{Predictors}     &                   &                      \\
\cline{2-3}
\text{Dependent variable}
&   $\hat{\epsilon}_{r^e,t-1} \hat{\epsilon}_{\Delta c_{i},t-1}$
&   $M/C_{i,t-1}$       & {$F$-statistic}     &   {$P$-Value}       \\
\hline
\T\  \hat{\epsilon}_{r^e,t} \hat{\epsilon}_{\Delta c_{CE},t}
& -0.143**
&   \num{1.06e-5}       &   2.93            & 0.055                 \\
%\rowfont{\small}
&   (-2.26)
&   (0.85)          &                   &                       \\
\T\  \hat{\epsilon}_{r^e,t} \hat{\epsilon}_{\Delta c_{NIPA},t}
& 0.108
& \num{-9.27e-6}        &   1.34            &   0.264               \\
& (1.64)
&   (-0.23)             &                   &                       \\
\T\  \hat{\epsilon}_{r^e,t} \hat{\epsilon}_{\Delta c_{H},t}
& -0.133**
& \num{-2.65e-5}        &   2.28            &   0.105               \\
%\rowfont{\small}
&   (-2.12)
&   (-0.51)             &                   &                       \\
\hline
\end{tabular}%
\end{table}
\end{document}

• further improvements can be obtain with use of rules from booktabs package

\documentclass[12pt]{article}  %It can be article / report / and book
\usepackage[top=1.3in, bottom=1.3in, left=1in, right=1in]{geometry}
%\usepackage{setspace}
\usepackage{array, booktabs, multirow, rotating}
\usepackage{graphicx}
\usepackage{amsmath, amssymb, amsfonts}
\usepackage[table]{xcolor}
\usepackage{graphicx}
\usepackage{pgfplots}
\usetikzlibrary{snakes}
\usepackage{caption}
\usepackage{subcaption}

\def\T{\texttt{T}}
\newcommand\mc[1]{ \multicolumn{1}{l}{#1}}
\newcommand\mcc[1]{\multicolumn{2}{>{\raggedright}p{82mm}}{#1}}

\usepackage{lipsum}% dummy text

\begin{document}
\begin{table}
\caption{Predicting Ex Post Covariances between between Stock Returns and Consumption Growth}
\label{Tab6}
\centering
\renewcommand\arraystretch{1.2}
\begin{tabular}{
>{$}l<{$}
*{2}{>{\centering}p{41mm}}
*{2}{S[table-format=2.3]}}
\multicolumn{5}{l}{\textbf{Panel A: 1st stage regression}}                  \\
\toprule
\text{Dependent variable}
&   \multicolumn{2}{c}{Predictors} & {$F$-statistic} & {$P$-Value}    \\
\midrule
r^e_t   &   \mcc{$D/P_{t-1}$, $\mathit{SVAR}_{t-1}$, $\mathit{BM}_{t-1}$,
$\mathit{NTIS}_{t-1}$, $\mathit{LTY}_{t-1}$}
&   4.45    &   0.001                                           \\
\Delta c_{CE,t}
&   \mcc{$\Delta c_{\mathit{CE},t-1}$, $\Delta c_{\mathit{CE},t-2}$,
$\Delta c_{\mathit{CE},t-3}$, $r_{f,t-1}$}
&   43.67   &   0.000                                           \\
\Delta c_{NIPA,t}
&   \mcc{$\Delta c_{\mathit{NIPA},t-1}$, $\Delta c_{\mathit{NIPA},t-2}$,
$\Delta c_{\mathit{NIPA},t-3}$, $r_{f,t-1}$}
&   65.14   &   0.000                                           \\
\Delta c_{H,t}
&   \mcc{$\Delta c_{H,t-1}$, $\Delta c_{H,t-2}$,
$\Delta c_{H,t-3}$, $r_{f,t-1}$}
&   19.11   &   0.000                                           \\
\midrule
\multicolumn{5}{l}{\textbf{Panel B: 2nd stage regression}}                  \\
\midrule
\T      &   \multicolumn{2}{c}{Predictors}
&                   &                      \\
\cmidrule(lr){2-3}
\text{Dependent variable}
&   $\hat{\epsilon}_{r^e,t-1} \hat{\epsilon}_{\Delta c_{i},t-1}$
&   $M/C_{i,t-1}$       & {$F$-statistic}   &   {$P$-Value}       \\
\midrule
\T\  \hat{\epsilon}_{r^e,t} \hat{\epsilon}_{\Delta c_{CE},t}
& -0.143**
&   \num{1.06e-5}       &   2.93            & 0.055                 \\
&   (-2.26)
&   (0.85)          &                   &                       \\
\T\  \hat{\epsilon}_{r^e,t} \hat{\epsilon}_{\Delta c_{NIPA},t}
& 0.108
& \num{-9.27e-6}        &   1.34            &   0.264               \\
%\rowfont{\small}
& (1.64)
&   (-0.23)             &                   &                       \\
\T\  \hat{\epsilon}_{r^e,t} \hat{\epsilon}_{\Delta c_{H},t}
& -0.133**
& \num{-2.65e-5}        &   2.28            &   0.105               \\
&   (-2.12)
&   (-0.51)             &                   &                       \\
\bottomrule
\end{tabular}%
\end{table}
\end{document}