This document describes the data used for the paper "Mergers in Canadian Newspaper Markets" by Ambarish Chandra and Allan Collard-Wexler (in progress). The purpose of this project is to make available to other researchers the (non-proprietary) data used in our project so that it may be easily used in other work.
The main dataset used in out project is available in Stata format; the name of the dataset is Canada5.dta. Right-click here to download it (94 KB). The circulation data are aggregate annual figures, therefore there is one observation per newspaper per year.
Source: Various issues of Editor and Publisher Magazine's International Yearbook (1995, 1996, 1998, 1999 and 2002).
The list of variables in the dataset are described below:
groupid: The ID variable for a given newspaper.
year: The year of the observation.
newspaper: Name of the newspaper
city: City where the publishing office is located
state: Name of the province
zip: Canadian Postal Code
eve: An indicator for whether the newspaper publishes in the evening.
circdaily, circsat, circsun: Circulation figures for weekdays, Saturdays and Sundays.
dailyrate, ratesat, ratesun: Advertising rates (per col. inch) for the corresponding days of the week.
avprice: Average single-copy daily price.
avpages: Average number of pages per daily issue.
area: Area of a page in square inches.
ownership: Group ownership in the current year (year of the observation).
yearest: Year the newspaper was established.
rateper: Daily Advertising rate per 10,000 readers. That is, rateper=dailyrate/circdaily*10000.
NOTE: Missing data for Saturdays/Sundays imply that the newspaper does not publish an edition on those days.