Knows Little, Learns Less? An Experimental Study of the Impact of the Media on Learning

One central claim for the value of democratic elections has been their potential for civic education. This paper is designed to understand how and under what conditions people learn about public affairs from different channels of communication provided by the news media and parties. In particular, much of the previous American literature assumes that newspapers are more effective channels of political information than television news. This poses three questions: (i) Is this also the case in other countries where the press and television differ systematically from the American model? (ii) What are the effects of new media that blend print and visual components like party websites? And (iii) do any learning effects vary by type of voter? The Campaign Learning Study uses large-scale experiments involving over 900 participants in the June 2001 British general election campaign. Learning about the government's record and party policies is monitored following exposure to campaign coverage in five outlets: tabloid and broadsheet newspapers, television news, election broadcasts by the major parties, and party websites. Using a typology based on voter's prior cognitive mobilization and strength of partisanship, voters are classified into four major types: apoliticals, ritual partisans, deliberators and cognitive partisans. The results confirm that: (i) in the midst of the campaign, prior to the experiments, British voters displayed widespread ignorance about party policies on some of the key issues in the campaign, such as Europe, asylum seekers, and taxation; (ii) after brief exposure to the media, there were substantial gains in knowledge about prospective party policies and, to a lesser extent, about the government's retrospective record; (iii) the knowledge gains about party policy were similar across TV news, broadsheet and tabloid papers, and party websites (although not party election broadcasts) even after applying the standard battery of social controls; (iv) the learning effects were fairly evenly divided among different types of voters, and; (v) the knowledge gains persisted well after the experiments ended. The conclusion considers the implications of the results for theories of the function of the news media as a civic forum in democratic elections.