Friday, August 04, 2006

Seats of power.

Media ownership is typically by either government or by private families. Limited degrees of freedom, indeed.

[The Journal of Law and Economics, vol. XLVI (October 2003)]
© 2003 by The University of Chicago. All rights reserved. 0022-2186/2003/4602-0014$01.50



World Bank

World Bank

World Bank

Harvard University

We examine the patterns of media ownership in 97 countries around the world. We find that almost universally the largest media firms are owned by the government or by private families. Government ownership is more pervasive in broadcasting than in the printed media. We then examine two theories of government ownership of the media: the public interest (Pigouvian) theory, according to which government ownership cures market failures, and the public choice theory, according to which government ownership undermines political and economic freedom. The data support the second theory.

* We thank Mei-Ling Lavecchia, Stefka Slavova, and especially Lihong Wang for excellent research assistance; Tim Besley, Edward Glaeser, Roumeen Islam, Simon Johnson, Lawrence Katz, Philip Keefer, Aart Kraay, Rafael La Porta, Mark Nelson, Russell Pittman, Andrew Weiss, and Luigi Zingales for comments; and the referee and two editors of this journal for helpful suggestions.


IN modern economies and societies, the availability of information is central to better decision making by voters, consumers, and investors.1 Much of that information is provided by the media, including newspapers, television, and radio, which collect information and make it available to the public. A crucial question, then, is how the media should be optimally organized. Should newspapers or television channels be state or privately owned? Should the media industry be organized as a monopoly or competitively? In this paper, we consider two broad theories of organization of the media and evaluate them using a new database of media ownership in 97 countries.

The first theory of the mediaand of institutions more generallyis the public interest (Pigouvian) theory, in which governments maximize the welfare of consumers. Government ownership of the media, perhaps even as a monopoly, is then desirable for three reasons. First, information is a public goodonce it is supplied to some consumers, it is costly to keep it away from others, even if they have not paid for it. Second, the provision, as well as dissemination, of information is subject to strong increasing returns: there are significant fixed costs of organizing information gathering and distribution facilities, but once these costs are incurred, the marginal costs of making the information available are relatively low. Third, if consumers are ignorant, and especially if private media outlets serve the governing classes, then state media ownership can expose the public to less biased, more complete, and more accurate information than it could obtain with private ownership.2 All these arguments were adduced by the management of the newly formed British Broadcasting Corporation (BBC) in support of maintaining a publicly subsidized monopoly on radio and television in Britain,3 and subsequently repeated in many developing countries.

In contrast, the public choice theory holds that a government-owned media outlet would distort and manipulate information to entrench the incumbent politicians, preclude voters and consumers from making informed decisions, and ultimately undermine both democracy and markets. Because private and independent media supply alternative views to the public, they enable individuals to choose among political candidates, goods, and securitieswith less fear of abuse by unscrupulous politicians, producers, and promoters.4 Moreover, competition among media firms assures that voters, consumers, and investors obtain, on average, unbiased and accurate information. The role of such private and competitive media is held to be so important for the checks-and-balances system of modern democracy that they have come to be called "the fourth estate," along with the executive, the legislature, and the courts.

Interestingly, even the Pigouvian economists, who advocate regulation or even nationalization by a benevolent government when considering other industries, support the free and private media.5 Ronald Coase points to this hypocrisy of Pigouvian economists: in the very industry where the case for state ownership is theoretically attractive, they shy away from taking it seriously. "It is hard to believe that the general public is in a better position to evaluate competing views on economic and social policy than to choose between different kinds of food."6 Nonetheless, the assumption of benevolent government often stops at the doorstep of the media, perhaps because economists want to protect their own right to supply information without being subject to regulation.

The two theories have distinct implications for both the determinants and the consequences of who owns the media.7 The public interest theory predicts that the more "benign" or "public-spirited" governments should have higher levels of media ownership and that the consequence of such ownership is greater freedom of the press, more economic and political freedom, and better social outcomes. The public choice theory predicts the opposite.

To understand the basic facts about media ownership, and to evaluate these predictions, we collect data on ownership patterns of media firmsnewspapers, television, and radioin 97 countries. Our paper provides a first systematic look at the extent of state and private ownership of media firms around the world, of the different kinds of private ownership, and of the prevalence of monopoly across countries and segments of the media industry. Our basic finding is that the two dominant forms of ownership of media firms around the world are ownership by the state and ownership by concentrated private owners, namely, controlling families.

Many hypothesize that the "amenity potential," also known as "the private benefits of control,"8 arising from owning media outlets is extremely high. In other words, the nonfinancial benefits, such as fame and influence, that are obtained by controlling a newspaper or a television station must be considerably higher than those that come from controlling a firm of comparable size in, say, the bottling industry. Economic theory then predicts that private control of media firms should be highly concentrated: the control of widely held firms with a high amenity potential is up for grabs and cannot be sustained in equilibrium.9 Our findings are broadly consistent with this prediction.

Having described the basic patterns of media ownership, we evaluate the data in light of the public interest and the public choice theories. We find that government ownership of the media is greater in countries that are poorer, have greater overall state ownership in the economy, lower levels of school enrollments, and more autocratic regimes. The last finding in particular casts doubt on the proposition that state ownership of the media serves benevolent ends.

We then consider the consequences of state ownership of the media, as measured by freedom of the press, political and economic freedom, and health outcomes. To this end, we run regressions of a variety of outcomes across countries on state ownership of the media, holding constant various country characteristics. We find pervasive evidence of "worse" outcomes associated with greater state ownership of the media (especially the press). The evidence is inconsistent with the Pigouvian view of state ownership of the media. Still, since we have only a cross-section of countries, we cannot decisively interpret this evidence as causal. Other, unmeasured, factors may account for the observed relationships.

1 Henry Simons, Economic Policy of a Free Society (1948); and George J. Stigler, The Economics of Information, 69 J. Pol. Econ. 213 (1961).
2 Vladimir Lenin, On the Freedom of the Press, 7 Lab. Mon. 35 (1925).
3 R. H. Coase, British Broadcasting (1950).
4 See Amartya Sen, Poverty and Famines (1984); Amartya Sen, Development as Freedom (1999); Timothy Besley & Robin Burgess, The Political Economy of Government Responsiveness: Theory and Evidence from India, 117 Q. J. Econ. 1415 (2002); and Timothy Besley & Andrea Prat, Handcuffs for the Grabbing Hand? Media Capture and Government Accountability (CEPR Discussion Paper No. 3132, London 2002).
5 See Simons, supra note 1; W. Arthur Lewis, The Theory of Economic Growth (1955); and Gunnar Myrdal, The Political Element in the Development of Economic Theory (1953).
6 R. H. Coase, The Market for Goods and the Market for Ideas, 64 Am. Econ. Rev. Papers & Proc. 389 (1974).
7 See also Simeon Djankov et al., The Regulation of Entry, 117 Q. J. Econ. 1 (2002).
8 See Harold Demsetz, The Amenity Potential of Newspapers and the Reporting of Presidential Campaigns, in Efficiency, Competition and Policy (H. Demsetz ed. 1989); Harold Demsetz & Kenneth Lehn, The Structure of Corporate Ownership: Causes and Consequences, 93 J. Pol. Econ. 1155 (1985); and Sanford J. Grossman & Oliver D. Hart, One ShareOne Vote and the Market for Corporate Control, 20 J. Fin. Econ. 175 (1988).
9 Lucian Arye Bebchuk, A Rent-Protection Theory of Corporate Ownership and Control (Working Paper No. 7203, Nat'l Bur. Econ. Res. 1999).


This section focuses on patterns of ownership in the media industry. Because ownership bestows control,10 it shapes the information provided to voters and consumers. Ownership, of course, is not the only determinant of media content. In many countries, even with private ownership, government regulates the media industry, provides direct subsidies and advertising revenues to media outlets, restricts access to newsprint and information collection, and harasses journalists. We discuss these modes of control as well.

10 Sanford J. Grossman & Oliver D. Hart, The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration, 94 J. Pol. Econ. 691 (1986).

A. Construction of the Database

We gathered new data on media ownership in 97 countries. We focused on newspapers and television since these are the primary sources of news on political, economic, and social issues.

Data on radio ownership are limited. Radio reaches a high proportion of the population, even in countries with the lowest levels of income and literacy, but it mainly delivers entertainment. The radio market is also highly regional, which precludes any single station from achieving a large market share. As a crude index, we gather ownership data on the top radio station as measured by peak adult audience and on the "all-news" radio station when one exists in a country.

Our selection of sample countries is driven by data availability. First, we identify the countries for which we have information on control variables. Since we are interested in the consequences of state ownership of the media, we need to make sure that our results are not driven by differences in economic development, education, political competition, or state intervention in the economy. To this end, we control for general levels of state ownership in the economy, primary school enrollment, autocracy, and gross national product per capita. We exclude five countries because (1) the country is in civil war (Democratic Republic of Congo, Sierra Leone), (2) the entity cannot be classified as a country (Hong Kong), or (3) no daily newspapers exist (Belize, Tajikistan). We also exclude 31 countries that lack sufficient data on media ownership. The final sample of 97 countries includes 21 in Africa, 9 in the Americas, 17 in Asia and the Pacific, 7 in Central Asia and the Caucasus, 16 in Central and Eastern Europe, 11 in Middle East and North Africa, and 16 in Western Europe.Table 1 describes all the variables used in the paper.


Within countries, we select media outlets on the basis of market share of the audience and provision of local news content for the year 1999. This approach focuses on who controls the majority of information flows on domestic issues to citizens. We exclude entertainment and sport media, as well as foreign media outlets, if they do not provide local news content.11 We include in our sample the five largest daily newspapers, as measured by share in the total circulation of all dailies, and the five largest television stations, as measured by share of viewing.12 We consult three primary data sources to select these outlets. First, we use Zenith Media Market and Media Fact Book 2000 publications, which are organized by region, including Western Europe, Central and Eastern Europe, Asia-Pacific, Middle East and Africa, and the Americas. Zenith Media's rankings of newspapers are checked with the World Association of Newspapers (WAN) World Press Trends 2000 report. The WAN data are also used as the source for total newspaper circulation, which is not reported by Zenith Media. Finally, we use the European Institute for the Media Media in the CIS report as a primary source for countries in the former Soviet Union. Alternative sources are sought in two cases: when there is an inconsistency in data reported by primary sources or when none of the sources covers the country in question. When this occurs, we use local media survey firms, World Bank external affairs offices, U.S. Department of State information offices, and direct contact with the media outlets.

Where possible, we rely on company annual reports and the WorldScope database for information on ownership of media firms. Many of our sample companies are not covered by WorldScope and operate in countries with limited disclosure requirements. Accordingly, we also use business news reports in LexisNexis and the Financial Times databases, country-specific company handbooks, media surveys, and internet information services. In all cases, we verify the ownership and other information externally by contacting World Bank external affairs offices, embassies in Washington, D.C., and regional or in-country media organizations.

Ownership data are for December 1999 or the closest date for which reliable data were available. For the majority of firms in the sample, ownership structures are stable over time. Timing is a significant issue only in the transition economies, where many media enterprises have been privatized or have increasing rates of foreign ownership. For these countries, we strictly enforce the December 1999 date of ownership information, even when we have more recent data.

We follow past work in identifying the ultimate controlling shareholder of each media outlet.13 We focus explicitly on voting as opposed to cash flow rights in firms. For each firm, we identify the legal entities and families who own significant voting stakes.14 This provides us with the first level of ownership. For each legal entity, then, we identify its ownership structure by determining all significant vote holdersthe second level of ownership. We continue to identify vote holders at each level of ownership until we reach an entity for which it is not possible to break down the ownership structure any further.

The entity that ultimately controls the highest number of voting rights, but no less than 20 percent at every link of the chain, is defined as the ultimate owner. Such control can be gained through direct ownership of more than 20 percent of voting rights of a media enterprise or indirectly through a chain of intermediate owners. For example, an individual X may control newspaper Z when he holds over 20 percent of the voting rights in Company Y, which in turn owns over 20 percent of the voting rights in Z. With indirect holdings, we define the percentage of ultimate ownership as the minimum holding along the chain of control.15

After identifying the ultimate owner, we classify each media outlet into one of the four main categories of owners: the state, families,16 widely held corporations, and "other." Examples of other controlling entities are employee organizations, trade unions, political parties, churches, not-for-profit foundations, and business associations. We define a corporation as widely held if there is no owner with 20 percent or more of the voting rights. We also keep track of whether the ultimate owner is a foreign family, entity, or government.

11 We include satellite and cable television if they carry local news content and are one of the top five television stations as measured by share of viewing. Satellite and cable television account for three of the top five stations in China. Out of the top five stations, four in India, two in Hungary, and three in Norway transmit by satellite. We do not include CNN and other global stations because they generally do not carry local news (and typically have a small share of viewers). We cover foreign news stations that spill over to local audiences if they carry local news and are among the top five stations. For example, in Austria, three of the top five television stations are German, but have a "local news window." We do not account for illegal access to foreign and/or global satellite television because we do not have access to such data (but also, there is no local news content).
12 Following the WAN definition, newspapers are considered dailies if they are published at least four times per week. In the initial phase of the data gathering (first 12 countries), we focused on the top 10 media enterprises in the daily newspaper and television markets. We subsequently reduced the sample to five firms per media type for two reasons. First, the difference in market coverage from increasing the sample of companies from five to 10 was marginal. In the first 12 countries, the top five newspapers account for an average of 62.4 percent of total circulation, and the top 10 for 74.1 percent. The correlation between the two is 94.2 percent. For the sample as a whole, the top five newspapers account for an average of 66.7 percent of total circulation. Television markets are even more concentratedon average the top five firms cover 89.5 percent of total viewing. Second, 20 countries in our sample do not have more than five daily newspapers, and 42 countries do not have more than five television stations.
13 For a discussion of methodology, see Rafael La Porta, Florencio Lopez-de-Silanes, & Andrei Shleifer, Corporate Ownership around the World, 54 J. Fin. 471 (1999).
14 The cutoff level of voting stakes depends on the mandatory disclosure levels in the country. In no case, however, is that threshold higher than 5 percent.
15 In a few cases, the owner of voting rights in a media firm does not hold the broadcast license. In these cases, we say that firm and not license ownership determines control. We do this because control of all broadcast licenses ultimately belongs to the government and licenses can be revoked depending on the strength of property rights in a country.
16 We use families as a unit of analysis and do not look within families.

B. Examples of Media Ownership

We illustrate the construction of the ownership variables with examples of individual firms. We start with a simple case of family ownership. In Argentina, the third largest newspaper, with a daily circulation of 177,000, is La Nacion. The owner of each share in La Nacion is entitled to one vote. There are two large shareholders in La Nacion (Figure 1): the Saguier family, with 72 percent of capital and votes, and Grupo Mitre, with 28 percent of capital and votes. Grupo Mitre is in turn 100 percent owned by the Mitre family. Although the Mitre family holds an indirect control of 28 percent, we follow the chain of control of the largest shareholder at each level of ownership. Thus we record the Saguier family as the ultimate owner and classify La Nacion as family owned.

FIGURE 1.La Nacion (Argentina)

A more complex example of family ownership is the Norwegian television station TVN (Figure 2). TVN is the second largest television station with local content in Norway, as measured by share of viewing. It is 50.7 percent controlled by Scandinavian Broadcasting Systems (SBS) and 49.3 percent by the largest Norwegian television station, TV2. We follow the chain of control along SBS rather than TV2, since SBS holds the majority of votes in TVN. Although Harry Evans Sloan (the chairman and CEO of SBS) holds a 9.8 percent share of voting rights in SBS, the only voting interest above 20 percent is held by the Netherlands United Pan-Europe Communications (UPC), with 3.3 percent of the vote. The majority shareholder of UPC is UnitedGlobal Com (51 percent). UnitedGlobal Com is in turn controlled by the Schneider family, through a combination of three direct interests totaling 21.9 percent, as well as 50 percent control of a voting agreement with 69.2 percent control of votes. We classify TVN as family owned and the Schneider family as the ultimate owner.

FIGURE 2.TVN (Norway)

State ownership takes different forms. The BBC is classified as state owned. It is funded by government license fees and advertising. The board of governors is appointed by Royal Prerogative, in practice the prime minister, and is accountable to the government. The BBC charter specifies a number of safeguards to ensure its independence from government interference. By contrast, the largest television station in Myanmar is controlled directly by the Ministry of Information and Culture, and the second largest station is controlled directly by the Myanmar military. In both cases, the state retains full powers to manage content and appoint and remove staff. In Turkmenistan, the state maintains direct control over the press: President Niyazov is officially declared the founder and owner of all newspapers in the country.

In a number of cases, we need to distinguish between state and political party ownership. In Kenya, the ruling party, the Kenyan African National Union (KANU), is the ultimate owner of the daily newspaper the Kenya Times, the country's fourth largest daily. Yet we do not classify Kenya Times as state owned, because if there were a change of government the ownership would remain with KANU. In contrast, control of the Kenyan Broadcasting Corporation (KBC) would remain with the state regardless of the political party in power, so we classify KBC as state owned. Ruling party ownership also occurs in Malaysia and Cote d'Ivoire. We place these firms in the "other" category, along with more clear-cut cases of media owned by opposition political parties.

In several cases, family ownership is closely associated with the state. In Kazakhstan, President Nazarbayev's daughter and son-in-law between them control seven of the 12 media outlets in our country sample. In Saudi Arabia, members of the royal family are the ultimate owners of two of the five most popular dailies. In cases where there is a direct family relationship between the ultimate owner and the head of state, and the governing system is a single-party state, we classify the media enterprise as state owned.

Other associations between families and state are prevalent throughout our sample. In Ukraine, the deputy prime minister holds over 30 percent of the top television station, while in Malawi the owner of the Nation newspaper is the minister of agriculture and vice-president of the ruling party. Neither of these positions are head of state in single-party governments, and we therefore classify both media outlets as family owned. Other unofficial links to the state were documented in country files but did not influence our classification of ultimate ownership. In Russia, the close associations between the owner of one of the main television stations, Boris Berezovsky, and then-President Yeltsin are well documented.17 In Indonesia, the daughter of ex-President Suharto still controls one of the main television stations. In an effort to be conservative in our measures of state control, in all these cases we classified the media outlets as family owned, since a change in government would sever the link between the politician and the media owner.

17 Boris Berezovsky wrote that "we helped Yeltsin defeat the Communists at the polls, using privately owned television stations" (Our Reverse Revolution, Wash. Post, October 26, 2000, at A27).

C. Media Regulations and Ownership

Throughout the world, governments regulate media using measures ranging from content restrictions in broadcasting licenses to constitutional freedom of expression provisions. The types of regulations and their enforcement vary significantly within our sample countries.

In some cases, ownership is influenced directly by regulation. In Norway, for example, regulations restrict owners from holding more than one-third of shares in media enterprises. Similar restrictions on ownership apply in Israel. Regulations of foreign ownership and cross-media ownership are also prevalent. Of the 49 countries surveyed by the World Association of Newspapers (WAN), 14 have explicit restrictions on foreign ownership of newspapers. In Brazil, for example, foreign ownership of voting capital of media enterprises is prohibited, and foreign participation in nonvoting capital is limited to 30 percent. Not surprisingly, foreign owners are absent from the Brazilian sample. A further 21 of WAN countries regulate cross-media ownership. In Australia, proprietors of major metropolitan newspapers are not permitted to own controlling interests in free-to-air television stations in the same market. As a result, the ultimate owner of the Nine Network television station, the Packer family, is limited to a 14.99 percent ownership stake in the one of Australia's leading publishers, John Fairfax Holdings.

Our data do not account for regulations that can substitute for state ownership as a means to control content. Singapore Press Holdings (SPH) publishes all of the top five daily newspapers in Singapore (Figure 3). Shares of SPH are divided into two categories: ordinary shares, which carry one vote per share, and management shares, which carry 200 votes per share. The ownership structure of SPH is characterized by complex cross holdings, with three major groups of shareholders evident in the data. First, the Lee family controls a total of 47.23 percent of votes through four companies. Second, the state holds a total of 27.23 percent of votes through various intermediary institutions. Third, there are a number of minority shareholdings held in nominee accounts at widely held financial institutions.18 Ownership of nominee accounts is not disclosed. It is possible that they are owned by families or the state, in which case our estimate of their control is conservative. We classify the Lee family as the ultimate owner of SPH. Yet by law, the government must approve the owners of management shares of SPH and can require owners to sell shares. We say that SPH is family owned and note that this is a conservative measure of the true influence of the state over SPH.

FIGURE 3.SPH (Singapore)

We use similar approaches in other cases of structural government influence of media firms. In Saudi Arabia, the government approves the appointment of editors-in-chief of newspapers and also has the right to dismiss them. Although clearly this increases the influence of the state on press content, we do not change the ownership classification because of this rule. In Malaysia, newspapers are required to renew their licenses annually. Editors of newspapers that publish critical views of government have been pressured to resign.19 In this environment, self-censorship becomes the norm. In all these instances, we nonetheless rely on ownership in constructing our measures, thus underestimating state influence.

State subsidies and state advertisement revenues enable governments to influence media content. Such subsidies are common in transition and African countries. In Cameroon, for example, the state refused to advertise in privately owned press after critical coverage of government. Defamation laws also influence content by repressing investigative journalism.

Direct regulations of content may interact with ownership. The North Korean Constitution states that the role of the press is to "serve the aims of strengthening the dictatorship of the proletariat, bolstering the political unity and ideological conformity of the people and rallying them solidly behind the Party and the Great Leader in the cause of revolution."20 In the Netherlands, the content of public service programming must be at least 25 percent news, 20 percent culture, and 5 percent education. Italy requires that 50 percent of broadcasting be of European origin. Because of these extensive regulations, our ownership classification is a conservative estimate of the true influence of the state over content.

18 In particular, Raffles Nominees Pre Ltd. holds 7.74 percent in SPH, 10.11 percent in Overseas Union Bank Ltd., and 19.44 percent in United Overseas Bank Ltd. GSBC Nominees Pte Ltd. controls 3.98 percent of SPH, 5.88 percent of the Overseas-Chinese Banking Corporation, 3.42 percent of Overseas Union Bank, and 4.31 percent of United Overseas Bank. Finally, Citibank Nominees Ltd. controls 1.63 percent of SPH, 3.82 percent of the Overseas-Chinese Banking Corporation, 4.08 percent of Overseas Union Bank, and 2.77 percent of United Overseas Bank.
19 U.N. ESCOR, 55th Sess. Report on the Mission to Malaysia, U.N. Doc. E/CN.4/1999/64.Add.1 (December 23, 1998).
20 N. Korea Const., art. 53, ch. 4 (1975).

D. Variable Construction

We construct two ownership variables from these data. First, we compute the percentage of firms in each categorystate or private. For example, two out of the top five newspaper enterprises in the Philippines are classified as state owned, as are three out of the top five television stations. We record Philippine newspaper market ownership as 40 percent state owned when measured by count, and television market ownership as 60 percent when measured by count. Second, we weight the ownership variable by market share. In the Philippines, the two state-owned newspapers account for 22.2 percent and 21.3 percent of circulation for the top five newspapers, respectively, so the newspapers are 43.5 percent state owned when measured by market share. In television, the three state-owned Philippine stations account for only 17.5 percent of the share of viewing for the top five television stations, so the television market is 17.5 percent state owned as measured by market share.

The market share variables, while more precise as a metric of state control, have the disadvantage that, in the countries with regional newspapers, such as the United States, the market share of any single firm is small. As a consequence, the variables we define are not properly compared with those in countries with national newspapers. This criticism, of course, is less compelling for television firms, which are typically national. The regressions presented below use market share variables, but our results are virtually identical using the counts.

For the radio market, we create a dummy equal to one if the top radio station is state owned, and zero otherwise.


A. Descriptive Statistics

Table 2 presents descriptive statistics on the ownership of newspaper and television markets in 97 countries. Countries are organized first by region and then sorted in alphabetical order. Several patterns emerge from the data.


Our first significant finding is that families and the state own the media throughout the world (Figure 4). In the sample of 97 countries, only 4 percent of media enterprises are widely held. Less than 2 percent have other ownership structures, and a mere 2 percent are employee owned. On average, family-controlled newspapers account for 57 percent of the total and family-controlled television stations for 34 percent of the total. State ownership is also vast. On average, the state controls approximately 29 percent of newspapers and 60 percent of television stations. The state owns a huge share72 percentof the top radio stations. On the basis of these findings, for the remaining analysis we classify ownership into three categories: state, private (which is the sum of the family, widely held, and employee categories), and other.

FIGURE 4.Newspaper and TV ownership

The nearly total absence of firms with dispersed ownership in the media industry is extreme, even by comparison with the finding of high levels of ownership concentration in large firms around the world.21 This result is consistent with the insight that the large amenity potential of ownership of media outlets creates competitive pressures toward ownership concentration. In a sense, both governments and controlling private shareholders get the same benefit from controlling media outlets: the ability to influence public opinion and the political process.

We say that the state has a monopoly in a media market if the share of state-controlled firms exceeds 75 percent. As Table 2 shows, a total of 21 countries have government monopolies of daily newspapers and 43 countries have state monopolies of television stations with local news. Table 2 also shows that families and the state control the media regardless of whether ownership is measured by count or weighted by market share.

Television has significantly higher levels of state ownership than newspapers.22 To explain this finding, a Pigouvian would focus on public goods and note that television broadcasts are at least in part nonexcludable and nonrivalrous. Television also has higher fixed costs than publishing and more significant economies of scale. The private sector might then underprovide broadcasting services, particularly in smaller markets serving remote areas, ethnic minorities, or students. These theories are central to many of the laws on public broadcasters in Europe. Alternatively, from the political perspective, privately owned newspapers are easier to censor than privately owned television. Because television can be broadcast live, control of content is more likely to require ownership. In this case, governments that want to censor news would own television.23

Table 3 shows that the data exhibit distinct regional patterns. State ownership of newspapers and television is significantly greater in countries in Africa and the Middle East and North Africa (MENA). On average, governments in Africa control 61 percent of the top five (in circulation) daily newspapers and reach 84 percent of the audience for the top five television stations. Seventy-one percent of African countries have state monopolies on television broadcasting. With the exception of Israel, all countries in MENA have a state monopoly over television broadcasting. State ownership of newspaperswhich averages 50 percent share of circulationis also high in countries in MENA.


By contrast, newspapers in Western Europe and the Americas are held predominately privately. In Western Europe, none of the top five daily newspapers are owned by the state. In the Americas, the majority of the newspapers have been owned and managed by single families for many decades. Levels of state ownership of television are also overwhelmingly lower in the Americas than in other regions. None of the top five stations in Brazil, Mexico, Peru, and the United States are state owned; this occurs in only one other country (Turkey) in our sample. In Western Europe, in contrast, a substantial number of public broadcasters push the regional state ownership average to 48 percent by count and 55 percent by share.

Countries in the Asia-Pacific, Central and Eastern Europe, and the former Soviet Union have ownership patterns closer to the sample mean.24

The simple statistics presented so far raise many questions. The evidence suggests that there are large private benefits of media ownership. Throughout the world, media are controlled by parties likely to value these private benefits: families and the state. In particular, the extent of state ownership of the media (particularly in television and radio) is striking, which suggests that governments extract value through control of information flows in the media. In the next few sections, we examine this evidence in light of the alternative theories.

21 Supra note 13.
22 Only five countries (Ghana, the Philippines, Uganda, Ukraine, and Uzbekistan) have more state control of the top five newspapers than of television stations.
23 A further argument is that the extent of required regulation of television is higher because of difficulties in defining property rights for broadcasting frequencies. It may be optimal from an efficiency standpoint for the state to control television stations directly as opposed to regulating the sector and spending resources in monitoring compliance. These arguments have been disputed by R. H. Coase, The Federal Communications Commission, 2 J. Law & Econ. 1 (1959), and others who do not see any need for government ownership and regulation arising from the particular technological features of broadcasting frequencies.
24 Ownership within each of these regions varies dramatically. Indonesia and Thailand have low levels of state ownership of the media compared with full state monopolies in North Korea and Myanmar. The predominantly privately owned media in Estonia and Moldova contrast with the full state control in Belarus and Turkmenistan.


In this section, we examine how ownership patterns are associated with different characteristics of countries. We examine four determinants of media ownership: the level of development, government ownership in other sectors, primary school enrollment, and autocracy.25 For all of these characteristics, it is hard to argue that causality runs from media ownership to these very basic country characteristics rather than the other way around.

Table 4 presents the results. Levels of state ownership of the press, but not of television and radio, are (statistically significantly) lower in countries that are richer. Levels of state ownership of television and radio but not the press are (statistically significantly) lower in countries that have lower levels of overall state ownership. Levels of state ownership of the radio are lower in countries with higher primary school enrollments. Perhaps most interestingly from the theoretical perspective, levels of state ownership of all forms of media are sharply and statistically significantly lower in less autocratic countries.26


This evidence challenges the public interest view of government ownership of the media. In particular, the fact that more autocratic regimes have higher levels of state ownership suggests that the unchecked and unlimited governments, rather than those constrained by the public, come to own the media. This is of course exactly what a public choice theorist would predict.

Table 5 presents data on the incidence of state media monopolies around the world (with the exception of Singapore, there are no private media monopolies in our sample). The data show that state monopoly is largely a feature of poor countriesthere is almost no incidence of state monopolies of newspapers, and relatively few of television, in the upper two quartiles of income distribution. These data themselves do not distinguish among theoriesa Pigouvian can easily explain why television and low income levels call for state monopoly.


25 We also considered ethnolinguistic fractionalization and latitude, but these variables did not enter significantly and reduced the sample size, so we do not include them in the analysis we present.
26 We also considered how state ownership varies according to the origin of commercial law in a country. Legal origins are classified into five categories: English, French, German, Socialist, and Scandinavian. Two countries (Iran and Saudi Arabia) cannot be classified in any of these groups since they practice traditional Islamic law. Legal origin has been interpreted as a proxy for the strength of property rights and inclination of the government to intervene in an economy as discussed in Rafael La Porta et al., Law and Finance, 106 J. Pol. Econ. 1113 (1998); and Rafael La Porta et al., The Quality of Government, 15 J. L. Econ. & Org. 222 (1999). It could, therefore, be argued that legal origin influences the extent to which a state chooses to control media. We find that, in television, average state ownership is remarkably similar across legal origins. Levels of state ownership of newspapers in countries of German and Scandinavian origin are significantly lower than in countries of French and socialist origin. For every other combination, state ownership of television or newspapers does not vary significantly according to legal origin.


In this section, we consider some of the consequences of state ownership of the media for a number of development indicators, such as freedom of the press, political and economic freedom, and health. In this analysis, we control for the factors that we identified as shaping media ownership. Even so, the results should be interpreted with caution, since it is possible that some omitted third factors determine both media ownership patterns and outcomes and we do not have instruments. In our defense, one must recognize that this omitted characteristic of a country must reflect the state's interest in controlling the information flows.

A. Freedom of the Press

Perhaps the clearest way to compare alternative theories of state ownership of the media is by focusing on freedom of the press. After all, the main implication of the Pigouvian theories is that greater government ownership should if anything lead to greater press freedom, as media avoid being captured by individuals with extreme wealth or political views.

Table 6 presents the results from the regressions of "objective" measures of media freedom on state ownership of the media. One measure of media freedom comes from Reporters sans Frontières (RSF) and reports on journalists jailed and media outlets closed by governments in 1999. Another measure was constructed from the reports by the Committee to Protect Journalists (CPJ) on the actual numbers of journalists jailed during 199799. These measures should be interpreted with caution, since presumably a truly repressive state, with full ownership and control of the media, does not need to repress journalists. We also look at a measure of internet freedom.

TABLE 6 MEDIA FREEDOM (N = 97 Countries)

Table 6 shows that greater state media ownership is associated with a greater number of journalists jailed and media outlets closed by the government, holding per capita income, primary school enrollment, state ownership, and autocracy constant. With one of the measures, the results are statistically significant. In our data, 45 democracies do not jail journalists, but seven (Benin, Malawi, Moldova, Niger, Nigeria, Russia, and South Korea) do. While 17 "near-democracies" do not jail journalists, 11 (Algeria, Angola, Cameroon, Ethiopia, Gabon, Iran, Peru, Togo, Tunisia, Turkey, and Zambia) do. Of the autocracies, 11 do not jail journalistsperhaps because there is no dissent, but six (China, Egypt, Kuwait, Myanmar, Syria, and Uzbekistan) do.27

Table 6 also establishes that countries with greater state media ownership censor the internet more heavily, as measured by a dummy that equals one if the government does not monopolize internet access and content (as measured by CPJ reports). This association is more naturally consistent with the public choice theory.

27 We have also measured freedom of the press using subjective indicators from Douglas A. van Belle, Press Freedom and the Democratic Peace, 34 J. Peace Res. (1997); and Freedom House, The Annual Survey of Press Freedom 2000 (2000). The effects of state ownership on these measures of freedom were also negative but in general insignificant.

B. Political and Economic Freedom

We examine the association between state media ownership and civil, political, and economic rights of a country's citizens. Under the public interest theory, state ownership of the media enhances these rights; under the public choice view, it curtails them by suppressing public oversight of the government and facilitating corruption. In this analysis, we again control for per capita income, government ownership of state-owned enterprises (SOEs), primary school enrollment, and autocracy.

The results are reported in Table 7. Government ownership of the press is associated with (statistically significantly) lower levels of political rights, civil liberties, security of property, and quality of regulation and higher levels of corruption and risk of confiscation. The effect of state ownership of television is also usually adverse, but only sometimes significant. These results support the public choice view that government ownership of the press restricts information flows to the public, which reduces the quality of government.28


Studies of election coverage illustrate the effect of state ownership of the media on the supply of political information. In Ukraine, election monitors from the Organization for Security and Cooperation in Europe recorded significant biases in media coverage related to ownership. Although all major television stations devoted more time to the incumbent than the opposition candidate, the state-owned television was more unbalanced in coverage and biased in content (despite legal requirements for the state-owned media to provide balanced and neutral coverage). Of its total first-round election-related coverage, the state-owned UT1 devoted 51 percent to the incumbent, and 75 percent of that coverage was positive. Each of the six opposition candidates received substantially less coverage (a maximum of 16.7 percent), and the vast majority of opposition coverage was negative. The television channel Inter displayed similar prejudice48.5 percent of coverage was allocated to the incumbent, and 73 percent of that coverage was favorable. Although Inter is classified as privately owned, it has strong informal links to the state because one of the three shareholders is the First Deputy Speaker of Parliament.29 The channel 1+1 is 51 percent privately and foreign owned, with a 49 percent nonvoting minority stake held by the State Property Company. The channel 1+1 devoted 34 percent of coverage to the incumbent, and 50 percent of that coverage was positive. Finally, STB, which is privately owned, was the least biased of the four stations. STB dedicated 23 percent of its coverage to the incumbent, with 40 percent of that coverage recorded as favorable. Similar observations have been made about Mexico,30 Ghana, and Kenya.

Our results are much stronger for the press than for television. For the latter, the effects of government ownership are insignificant. One reason might be that private press, which is more common, provides a check on state television, ensuring freer flows of information than would occur if both were in state hands. The data confirm that the outcomes are worse when the state owns both newspapers and television than when it owns only one of them.

Luigi Zingales31 argues that one benefit of private media is to provide information for stock market participants, thereby improving security pricing and revealing abuses by corporate insiders. The last row of Table 7 shows that greater state ownership of the media is associated with a lower number of companies (per capita) listed on the national stock market. These results suggest that state control of information flow is detrimental to financial development, which is consistent with the public choice theory.

Taken in its entirety, the evidence is broadly supportive of the public choice view that governments own the mediaespecially the pressnot to improve the performance of economic and political systems, but to improve their own chances to stay in power. When the two theories yield different predictions, there is no evidence consistent with public interest.

28 Our results are also unsurprising in a broader historical context. Dictators from Napoleon, to Lenin, to Hitler, to Marcos nationalized the press.
29 The shareholdings are approximately equally distributed33 percent, 33 percent, and 34 percentamong three individuals, with the Deputy Speaker holding one of the 33 percent stakes.
30 For Mexico, see Joel Simon, Hot on the Money Trail, 37 Colum. J. Rev., JanuaryFebruary 1998, at 13.
31 Luigi Zingales, In Search of New Foundations, 55 J. Fin. 1623 (2000).

C. Health

Lenin asked a pointed question: whom is the free press for? Our analysis has focused on political and economic freedom, but a Pigouvian could presumably argue that the true benefits of state ownership accrue to the disadvantaged members of society. Freed from the influence of the capitalist owners, state-controlled media can serve the social needs of the poor. A public choice theorist would argue, in contrast, that the government would use its ownership of the media to muzzle the press and to prevent the disadvantaged groups from voicing their grievances. Government ownership should then be associated with inferior social outcomes.

The contrasting predictions of the two views can be evaluated empirically.Table 8 reports the relationships between state ownership of the media and health indicators, holding constant our usual controls. Countries with greater state ownership of the media exhibit lower life expectancy, greater infant mortality, and less access to sanitation and health system responsiveness. Private media ownership is associated with health as well as economic and political outcomes, which is consistent with the public choice but not with the public interest theory.32


32 Earlier studies reached a similar conclusion. Duncan Thomas, John Strauss, & Maria-Helena Henriques, How Does Mother's Education Affect Child Height? 26 J. Hum. Resources 183 (1991), finds that maternal access to the media has a strong and positive effect on child health in Brazil. Sen, Poverty and Famines, supra note 4, argues that the lack of democracy, freedom of information, and an independent press contributed to almost 30 million deaths during China's Great Leap Forward between 1958 and 1961. Sen contrasts this with India, which has not experienced a major famine since independence and has stronger democratic processes and press freedom: "The Government (of India) cannot afford to fail to take prompt action when large scale starvation threatens. Newspapers play an important part in this, in making the facts known and forcing the challenge to be faced" (at 76). Besley & Burgess, supra note 4, at 1, tests Sen's proposition empirically. Using data across Indian states, it finds that higher newspaper circulation increases government responsiveness to natural shocks. Supporting this hypothesis as well is David Stromberg, Radio's Impact on Public Spending (Working paper, Inst. Int'l Econ. Stud., Stockholm 2001). Rather than focus on media penetration, our study points to a critical deterrent to media's ability to serve social goalsgovernment ownership.

D. Robustness

We check the robustness of our results in a number of ways. First, we estimate the regressions in Tables 48 using ethnolinguistic heterogeneity and latitude as additional controls. These controls are insignificant and do not affect the results, but we do lose observations.

Second, we ask whether the rich countries, with good outcomes and low levels of state ownership of the press, drive the results, by estimating all the regressions excluding Organization for Economic Cooperation and Development (OECD) countries. In Table 9, we present the results using state ownership of the press (by share) as the independent variable and using all the dependent variables from Tables 68 that were statistically significantly influenced by this state ownership variable in the whole sample. We find that, with the exception of "number of listed firms," the influence of state ownership of the press on outcomes remains statistically significant, although in many regressions the level of significance diminishes as compared with the whole sample. The magnitude of the coefficients remains very similar as well. The results thus are not driven by the OECD countries. Likewise, the results are preserved if we remove the richest quartile of countries and hold both within the richer half and the poorer half of the sample.


Third, when in addition to all the other controls we include a separate dummy for each quartile of per capita income, the results stay at comparable significance levels.

Fourth, we check whether the results are driven by any particular region. The answer is no: the results hold controlling for continent "dummies," but also within Europe, within America, and within Africa/Middle East, although at lower levels of statistical significance.

Fifth, we check whether the results are robust to controlling for the level of media penetration, and the answer is yes.

Last, alternative definitions of dependent and independent variables yield similar results.

We also try to address the question of whether our results are driven by monopoly rather than government ownership per se (there is no private monopoly of the press in the sample, only state monopoly). To this end, we divide countries into unequal groups with 025 percent, 2650 percent, 5175 percent, and over 75 percent state ownership of the press and reestimate our regressions using the quartile state ownership dummies. In general, we do not find a strong pattern of coefficients, although the results are inconsistent with the proposition that "bad" outcomes associated with state ownership are driven solely by monopoly. There is some evidence, to the contrary, that the lowest quartile of state ownership is associated with somewhat better outcomes than the middle quartiles. Although the evidence is very weak, it does not point to a "monopoly only" story.


In this paper, we examine ownership patterns of newspapers and television (and to a lesser extent radio) in 97 countries around the world. We find that media firms nearly universally have ownership structures with large controlling shareholders and that these shareholders are either families or governments. This evidence is broadly consistent with the ideas that there is large amenity potential (control benefits) associated with owning mediabe it political influence or fame.33

We then show that countries that are poorer, more autocratic, with lower levels of primary school enrollment, and with higher levels of state intervention in the economy also have greater state ownership of the media. In addition, countries with greater state ownership of the media have less free press, fewer political rights for citizens, inferior governance, less developed capital markets, and inferior health outcomes (the last result being particularly important in light of the argument that state ownership of the media serves the needs of the poor). The negative association between government ownership and political and economic freedom is stronger for newspapers than for television. Although none of this evidence can be unambiguously interpreted as causal, it obtains with extensive controls and there is no empirical evidence of any "benefits" of state ownership.

At some broad level, these results are unsurprising, as intellectuals since John Milton in the seventeenth century have advocated free press and independent media. Still, the results do provide support for the public choice against public interest theory of media ownership in an environment where, as Coase has argued, the public interest case is especially strong.34 Yet the data are inconsistent with these Pigouvian arguments and reveal no benefits of state ownership.

33 See Demsetz, and Demsetz & Lehn, supra note 8, for the discussion of these ideas.
34 Supra note 6.


Bebchuk, Lucian Arye. "A Rent-Protection Theory of Corporate Ownership and Control." Working Paper No. 7203. Cambridge, Mass.: National Bureau of Economic Research, 1999.
Berezovsky, Boris. "Our Reverse Revolution: Under Yeltsin, We Oligarchs Helped Stop Russia from Reverting to Its Old, Repressive Ways." Washington Post, October 26, 2000, p. A27.
Besley, Timothy, and Burgess, Robin. "The Political Economy of Government Responsiveness: Theory and Evidence from India." Quarterly Journal of Economics 117 (2002): 141551.
Besley, Timothy, and Prat, Andrea. "Handcuffs for the Grabbing Hand? Media Capture and Government Accountability." CEPR Discussion Paper No. 3132. London: Centre for Economic Policy Research, January 2002.
Coase, R. H. British Broadcasting; A Study in Monopoly. Cambridge, Mass.: Harvard University Press, 1950.
Coase, R. H. "The Federal Communications Commission." Journal of Law and Economics 2 (1959): 140.
Coase, R. H. "The Market for Goods and the Market for Ideas." American Economic Review: Papers and Proceedings 64 (1974): 38491.
Committee to Protect Journalists. "Attacks on the Press in 1999: A Worldwide Survey by the Committee to Protect Journalists." 2000.
Demsetz, Harold. "The Amenity Potential of Newspapers and the Reporting of Presidential Campaigns." In Efficiency, Competition and Policy, edited by H. Demsetz. London: Basil Blackwell, 1989.
Demsetz, Harold, and Lehn, Kenneth. "The Structure of Corporate Ownership: Causes and Consequences." Journal of Political Economy 93 (1985): 115577.
Djankov, Simeon; La Porta, Rafael; Lopez-de-Silanes, Florencio; and Shleifer, Andrei. "The Regulation of Entry." Quarterly Journal of Economics 117 (2002): 137.
Freedom House. Freedom in the World: Annual Survey of Freedom Country Ratings 1996/1997. New York: Freedom House, 1997.
Freedom House. Freedom in the World: Annual Survey of Freedom Country Ratings 1999/2000. New York: Freedom House, 2000.
Freedom House. The Annual Survey of Press Freedom 2000. New York: Freedom House, 2000.
Grossman, Sanford J., and Hart, Oliver D. "The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration." Journal of Political Economy 94 (1986): 691719.
Grossman, Sanford J., and Hart, Oliver D. "One ShareOne Vote and the Market for Corporate Control." Journal of Financial Economics 20 (1988): 175202.
Gwartney, James; Lawson, Robert; and Samida, Dexter. Economic Freedom of the World. Vancouver: Fraser Institute, 2000.
Kaufmann, Daniel; Kraay, Aart; and Zoido-Lobaton, Pablo. "Governance Matters." Policy Research Working Paper No. 2196. Washington, D.C.: World Bank, 1999.
Lange, Yasha. Media in the CIS. Duesseldorf: European Institute for the Media, 2000.
La Porta, Rafael; Lopez-de-Silanes, Florencio; and Shleifer, Andrei. "Corporate Ownership around the World." Journal of Finance 54 (1999): 471517.
La Porta, Rafael; Lopez-de-Silanes, Florencio; Shleifer, Andrei; and Vishny, Robert W. "Legal Determinants of External Finance." Journal of Finance 52 (1997): 113155.
La Porta, Rafael; Lopez-de-Silanes, Florencio; Shleifer, Andrei; and Vishny, Robert W. "Law and Finance." Journal of Political Economy 106 (1998): 111355.
La Porta, Rafael; Lopez-de-Silanes, Florencio; Shleifer, Andrei; and Vishny, Robert W. "The Quality of Government." Journal of Law, Economics, and Organization 15 (1999): 22279.
Lenin, Vladimir. "On Freedom of the Press." Labour Monthly 7 (1925): 3537.
Lewis, W. Arthur. The Theory of Economic Growth. New York: Harper Torchbook, 1955.
Myrdal, Gunnar. The Political Element in the Development of Economic Theory. London: Routledge & Kegan, 1953.
Polity IV Project. Polity IV Dataset: Political Regime Characteristics and Transitions, 18001999. College Park: University of Maryland, Center for International Development and Conflict Management, 2000.
Reporters sans Frontières. Annual Report 2000. Paris: Reporters sans Frontières, 2000.
Sen, Amartya. Poverty and Famines. Oxford: Oxford University Press, 1984.
Sen, Amartya. Development as Freedom. New York: Alfred A. Knopf, 1999.
Simon, Joel. "Hot on the Money Trail." Columbia Journalism Review, JanuaryFebruary 1998, pp. 1322.
Simons, Henry. Economic Policy of a Free Society. Chicago: University of Chicago Press, 1948.
Stigler, George J. "The Economics of Information." Journal of Political Economy 69 (1961): 21325.
Stromberg, David. "Radio's Impact on Public Spending." Working paper. Stockholm: Stockholm University, Institute for International Economic Studies, 2001.
Thomas, Duncan; Strauss, John; and Henriques, Maria-Helena. "How Does Mother's Education Affect Child Height?" Journal of Human Resources 26 (1991): 183211.
United Nations Development Programme. Human Development Report 2000. New York: United Nations Development Programme/Oxford University Press, 2000.
UNESCO Institute for Statistics. Annual Statistical Yearbook. Paris: UNESCO, 1999.
van Belle, Douglas A. "Press Freedom and the Democratic Peace." Journal of Peace Research 34 (1997): 40514.
World Association of Newspapers. World Press Trends 2000. Paris: Zenith Media, 2000.
World Bank. Database of Enterprise Indicators on Transition Economies, Europe and Central Asia Region. Washington, D.C.: World Bank, 2000.
World Bank. World Development Indicators 2000. Washington, D.C.: World Bank, 2000.
World Health Organization. World Health Report 2000. Geneva: World Health Organization, 2000.
Zenith Media. Americas Market and MediaFact. London: ZenithOptimedia, 2000.
Zenith Media. Asia Pacific Market and MediaFact. London: ZenithOptimedia, 2000.
Zenith Media. Central and Eastern European Market and MediaFact. London: ZenithOptimedia, 2000.
Zenith Media. Middle East and Africa Market and MediaFact. London: ZenithOptimedia, 2000.
Zenith Media. Western European Market and MediaFact. London: ZenithOptimedia, 2000.
Zingales, Luigi. "In Search of New Foundations." Journal of Finance 55 (2000): 162354.


Post a Comment

<< Home