Advertising in recession

Research on Advertising in a Recession
A Critical Review and Synthesis

GERARD J. TELLIS

Based on an extensive review of research on advertising in a recession, the authors identify over 40 related studies. Ten of these studies involve original empirical analyses of cross-sectional or time series data. The rest are theoretical discussions, reviews, cases, or opinions. The empirical studiesmay be classified into four groups based on the dependent variable analyzed: (1) sensitivity of advertising expenditures to the economy, (2) sensitivity of brand versus private-label share to economic expansions and contractions, (3) impact of advertising in a recession to sales or market share during or after a recession, (4) impact of advertising in a recession to profits during and after therecession. The authors critically review these studies and synthesize the major findings.

Marshall School of Business University of Southern California tellis@marshall.usc.edu
KETHAN TELLIS

Marshall School of Business University of Southern California ketha n .tel I i s@gma i Loom

INTRODUCTION

Periodically, recessions afflict the U.S. and world economies. At such time, firms facedeclining revenues and shortages of cash. Their natural tendency is to cut back on seemingly discretionary expenditures such as R&D, marketing, and advertising. What has perplexed managers and analysts is whether such cutbacks are wise or self-defeaüng, either in the short or long term. Over the decades, a number of studies have examined this issue in the context of advertising. This article criticallyreviews the literature on the effectiveness of advertising in a recession and synthesizes the major conclusions from this review. A narrow definition of a recession is two successive quarters of negative growth in gross domestic product (GDP). The advertising literature reviewed here often has used the term narrowly (as in the above definition) and sometimes broadly (a contraction—as opposed to anexpansion—in the economy or a decline in GDP over a whole year). An overview of a century of (narrowly defined) recessions—with specific regard to the duration of the recession as well as estimated changes in GDP and advertising expenditures—leads to three observations:

1. Recessions have been getting shorter, while expansions have been getting longer. 2. Even with full consideration of recenteconomic conditions, the frequency of recessions has declined slightly in recent decades. 3. Advertising expenditures are quite sensitive to changes in GDP.

Exactly how sensitive advertising expenditures are to GDP cannot be estimated without a formal analysis; recessions occur within and across years, while the estimated changes are measured on an annual basis. One such formal analysis(reviewed subsequently) suggests that the estimated elasticity of advertising expenditures to GDP is quite high, on the order of 1.4, In the work that follows, we will use the term recession in its broader context, signifying an economic contraction. Our search of the literature followed four steps:

1. We did a search of the words "advertising" and "recession" in three major electronic databases:Google, JSTOR, and ABI/Inform.
DOi: 10.2 501/S0021849909090400

3 0 4 JOüeUBl Of

RESEeilCil September 2009

RESEARCH ON ADVERTISING IN A RECESSION

Periodically, recessions afflict the U.S. and world economies. At such time, firms face declining revenues and shortages of cash. Their natural tendency is to cut back on seemingly discretionary expenditures such as R&D, marketing, andadvertising. What has perplexed managers and analysts is whether such cutbacks are wise or seif-defeating, either in the short or iong term.

Opinions, or reviews of the empirical studies. The 10 primary studies span large time periods (1926 to the current date) and report on a wide variety of recessions (from 1920 to 2005). They also cover a variety of countries and contexts, focus on a variety of...