Suppose that you run a newspaper and one of your primary advertisers is the government.  You then get wind of a corruption scandal involving members of said government.  Do you alienate your sponsor or quash the story?  This is the real question Rafael Di Tella and Ignacio Franceschelli ask in Government Advertising and Media Coverage of Corruption Scandals.


We construct measures of the extent to which the 4 main newspapers in Argentina report government corruption in their front page during the period 1998-2007 and correlate them with the extent to which each newspaper is a recipient of government advertising. The correlation is negative. The size is considerable: a one standard deviation increase in monthly government advertising (0.26 million pesos of 2000) is associated with a reduction in the coverage of the government’s corruption scandals by 0.31 of a front page per month, or 25% of a standard deviation in our measure of coverage. The results are robust to the inclusion of newspaper, month, newspaper*president and individual-corruption scandal fixed effects as well as newspaper*president specific time trends.

The authors survey the four main newspaper in Argentina over the time period 1998-2007, and focus on corruption scandals involving government officials (2).  Their theory is that adverse coverage correlates negatively with government funding (2).  Because the government tends to finance the media to a great extent in Argentina, the results are different than in someplace like the US, where the (often left-leaning) media tend to have partisan papers, and in which party affiliation affects coverage (3-4), regardless of who is currently in power.

This difference is not simply academic.  In the Argentinian case, 200 tax inspectors were sent to investigate one newspaper the day after a report of corruption within the tax agency was published (5, footnote 6).  Aside from direct threats, there are more indirect methods:  much of the “private” advertising is actually advertising by government-affiliated firms (6-7).  As the authors note, “One of the characteristics of small developing countries is the relatively large influence of the government on business” (7).  I would re-phrase that to say that “one of the characteristics keeping countries underdeveloped is the relatively large influence of government on business.”

At any rate, the authors have a data set of 254 scandals, of which more than 150 were reported on only one paper’s front page (8).  They looked at front-page offerings, and consider a scandal “buried” if a paper does not print an article on a front page regarding that story.  They also found that 256,000 pesos (at 2000 values) led to a half a cover drop (that is, a 37% drop) in corruption reporting for one month (13).  Non-government corruption coverage, meanwhile, was not affected by government payments (15-16).  On the other hand, there is a positive correlation between corruption coverage and circulation (17).  So this leads newspaper companies to come to a financial decision:  the drop of one front-page corruption story leads to 560,000 pesos from the government.  But each front-page story leads to 1.7 million pesos from subscribers (1.48 million more subscribers, each paying roughly 1.15 pesos apiece).  It would seem as though this should be a no-brainer, but if the marginal cost of a newspaper is 0.77 pesos, the authors note that this would be a break-even point:  0.77 = (1.7 – 0.56) / 1.48.  Actually, the authors have 0.75 pesos, but they use 0.58 in their calculation rather than 0.56, which I believe to be in error.

My quick takeaway:  the best way to reduce corruption is to have a free press and a small, limited government which can neither afford nor be allowed to influence the media.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s