Senay Sokullu, "A Semi-parametric Analysis of Two-Sided Markets: An Application to the Local Daily Newspapers in the U.S.", Journal of Applied Econometrics, Vol. 31, No. 5, 2016, 843-864. This file contains information about the data and program files that can be used to replicate the results of the paper. datamatlab.txt contains all the variables used by the matlab estimation program newspapersf.m psiconf.txt provides confidence band values for the parametrically estimated functions. These values are computed in SAS. newspapersf.m contains the matlab code to estimate the nonparametric model npest_instd.m, npest_instadv.m and kernel.m are the functions that are needed to run newspapersf.m newspapersestimation.sas contains the code to replicate the parametric estimation results revisionconfint.sas contains the code to obtain bootstrap confidence bands for parametrically estimated network effect functions. full2.sas7bdat is the dataset for SAS which contains the same variables as matlabdata.txt. It also contains some other variables which are created in the process of reaching the final variables and others that have not been used in the estimations. All files except full2.sas7bdat are ASCII files in DOS format. They are zipped in the file ss-files.zip. Unix/Linux users should use "unzip -a". The file full2.sas7bdat is a binary file. It is zipped in ss-sasdata.zip. %%% The variables in the file matlabdata.txt %%% Population_2010: 2010 population of the city under_5: Percentage of population under 5 years-old, city level under_18: Percentage of population under 18 years-old, city level _65_and_over: Percentage of population over 65 years-old, city level Female: Percentage of females in the population, city level White_alone: White alone, percent, 2010, city level Black_or_African_American: Black or American Alone, percent, 2010, city level American_Indian_and_Alaska_Nativ: American Indian and Alaska Native alone, percent, 2010, city level Asian_alone: Asian alone, percent, 2010, city level Two_or_More_Races: Person who are of Two or More Races, percent, 2010, city level Hispanic_or_Latino: Hispanic or Latino, percent, 2010, city level White_alone__not_Hispanic_or_Lat: White alone, not Hispanic or Latino, percent, 2010, city level Living_in_same_house_1_year___ov: Living in same house 1 year & over, percent, 2008-2012, city level Foreign_born_persons: Foreign born persons, percent, 2008-2012, city level other_than_English: Language other than English spoken at home, pct age 5+, 2008-2012, city level High_school_or_higher: High school graduate or higher, percent of persons age 25+, 2008-2012, city level Bachelor_s_or_higher: Bachelor's degree or higher, percent of persons age 25+, 2008-2012, city level Veterans: Veterans, 2008-2012, city level Mean_travel_time: Mean travel time to work (minutes), workers age 16+, 2008-2012, city level Housing_units: Housing units, 2010, city level Homeownership_rate: Homeownership rate, 2008-2012, city level Housing_units_in_multi_unit: Housing units in multi-unit structures, percent, 2008-2012, city level Median_value_of_owner_occupied_h:Median value of owner-occupied housing units, 2008-2012, city level Households: Households, 2008-2012, city level Persons_per_household : Persons per household, 2008-2012, city level Per_capita_money_income: Per capita money income in past 12 months (2012 dollars), 2008-2012, city level Median_household_income: Median household income, 2008-2012, city level below_poverty: Persons below poverty level, percent, 2008-2012, city level Retail_sales: Retail sales, 2007 ($1000), city level Retail_sales_per_capita: Retail sales per capita, 2007, city level Land_area: Land area in square miles, 2010, city level per_square_mile: Persons per square mile, 2010, city level sincome: county level income sprint: county level wages in printing industry m_inches: average weekday column inches of advertisement in the newspaper m_dollars: average weekday advertisement revenue of the newspaper m_unit: Average weekday number of advertisements in the newspaper m_peri: Average daily ad rate per column inch m_peru: Average daily ad rate per unit weekday_circ: Average weekday circulation over the six months period covering April 1, 2011 to September 30, 2011 weekend_circ: Average weekend circulation over the six months period covering April 1, 2011 to September 30, 2011 price_yearly: Yearly price of the newspaper dailyprice: Daily price of the newspaper extra: Sum of number of special editions, special weekly selections and the magazines of the newspaper income: county level income (only the lead county) wage_in_printing: wages in printing industry, county level (only the lead county) avr_page_number_per_issue_wd: Average weekday page number per issue of the newspaper _2011: 2011 population, city level wprinting: Wages in printing industry, state level earnpaper: Earnings in printing industry, state level wpaper: Wages in paper industry, state level gdp_growth: 2011 state level gdp growth nr: weekday_circ/(_2011*100) na: m_inches/total weekly average column inches of advertising (this definition of ad share is not used in the estimations) loginc: log(sincome) logprint: log(sprint) logpop: log(_2011) totalci: weekday average column inches per issue (weekday number of pages in terms of column inches) na2: m_inches/totalci (this is the varaible used as the share of advertisers in the estimations) days: Number of days in a year that the newspaper is published logminc: log(median_household_income) %%% The variables in psiconf.txt %%% paper: Name of the newspaper phi95: 95th percentile of phi function (upper bound) psi95: 95th percentile of psi function (upper bound) phi5: 5th percentile of phi function (lower bound) psi5: 5th percentile of psi function (lower bound) nr: share of readers reading the newspaper na2: share of advertisers in the newspapers psihat: Fitted value of psi function obtained by parametric estimation phihat: Fitted value of phi function obtained by parametric estimation