Economics Review

Transparency of Government (Part 3) - What Roles Should the Media Play?

1. Introduction

Following the preceding two articles (Part1 and Part 2) on government transparency,I would like to examine, in this last article in the series,what roles the media can and should play to enhance governmenttransparency. As discussed in the first article, effectivegovernance can be administered to a government and politicsby disclosing government information to an indiscriminatenumber of citizens. However, individuals have little incentiveto seek information on their own because expected benefitsare too modest compared to the costs involved, creating asituation prone to the so-called "free rider" problem. Meanwhile,it is possible for members of an incumbent regime (governmentand ruling party/parties) to win public trust and reinforcetheir political power by enhancing the transparency of theiractivities. But government members are apt to pursue theirpersonal interests by restricting information disclosure ortrying to hide policy failures. Here, it is important fora free and independent media to provide and disseminate informationabout a government and government activities, thereby enhancingpeople's power as a "principal" to provide governance on onehand and giving incentives for the government to improve itspolicy responsiveness on the other. In this regard, the presenceof a free and independent media is indispensable for ensuringa sound democratic society.

On the roles and functions of the media, which have beenattracting much attention among economists, a series of theoreticaland empirical analyses have been published. In this article,I would like to introduce some of these analytical works andshow how the media can alleviate moral hazard (ref "Transparencyof Government [Part 1]") and other problems surroundingthe "principal-agent" relationships between the general public/votersas a "principal" and the government/politicians as an "agent"by presenting specific routes and concrete empirical examples.Next, I would like to examine what kinds of organizationalform, market structure and restrictions must be put in placeto enable the media to function properly as a "political monitor"to supervise the government and politics. Restrictions overthe media have been implemented, supposedly, for the benefitsof the general public as "customers" who "consume" the "output"(information) of the media. However, it is important to seethe situation from a different angle, focusing on the needto improve the welfare of the general public as "voters,"who provide governance to the incumbent regime (governmentand politicians), rather than on the welfare of the generalpublic as "consumers" (Besley, Burgess and Prat, 2002). Inthis article, I would like to take up the issue of governmenttransparency from this new viewpoint.

2. Effectiveness of Mediaas Monitor of Government

For the general public, the mass media are supposed to bea powerful ally through which they can access informationat a low cost. What are the conditions that enable a freeand independent media to provide the general public with sufficientinformation for ensuring that the government serves as a good"agent" for the people?

First, there must be efficient "sorting" of information.Voters must be adequately furnished with the proper information(past achievements of each candidate) necessary to selecta specific politician in elections. Second, there must be"discipline." This, for instance, concerns whether the mediaare capable of properly detecting and informing the generalpublic of the self-serving behavior of bureaucrats or politicians,such as accepting bribes. Third, the media must be capableof influencing policy decisions by taking up "salient issues"for voters. When the mass media focus on a certain issue,the government responses to that particular issue become widelyknown to the general public, thereby providing voters withgrounds upon which they can judge and cast a ballot. Thismeans that, by proactively responding to problems highlightedby the mass media, the government would be able to build upa good reputation, win people's trust and reinforce its politicalpower. In other words, by providing relevant information on"salient issues" for the general public, the mass media wouldbe able to enhance the policy responsiveness of the government.Moreover, the provision of information in such a manner wouldhelp alleviate a series of problems that exist in the "principal-agent"relationship between the government and the general public,and eventually lead to greater accountability and transparencyof the government.

Media and policy response:Empirical analysis of Indian case
India provides an interesting case example for showing howthe media can influence government policy responses. The freedomand independence of the mass media generally rests upon thedegree of a country's economic development. India, however,is known for having had a free and independent press fromlong ago. Amartya Sen, Nobel Prize winner and professor atthe University of Cambridge, says that India has never beenafflicted with a major famine since its independence becausethe press keeps the general public well informed of any foodproblem if it occurs, creating a situation in which the governmentfinds it imperative to act quickly (Sen, 1984). In contrast,China, where democratization has been slow to proceed andthere has been little freedom of information, suffered anextremely severe famine from 1958 to 60. Meanwhile, amongAfrican countries, those with a representative democracy andan effective media have been successful in preventing famines(Dreze and Sen, 1989).

Besley and Burgess (2002) have conducted quantitative empiricalanalysis on the relationships between the development of themedia and government policy responses. Attentive to the factthat the degree of media penetration and the roles playedby the media differ from one state to another within India,they created a panel data set by collecting state-level datawith a particular focus on two policy responses, namely, thedegree of public food supply in a year of drought and cropfailure, and the scale of government expenditures on disasterrelief for farmers. They have shown that the amount of governmentexpenditure on providing relief for drought or flood damagestends to be large in a state with a high newspaper circulation.Their analysis has also shown that the circulation of localnewspapers in the local language/vernacular has a particularlystrong impact. This is coherent with the general perceptionthat when there is a local shock, such as a natural disaster,vernacular press is the most important source of informationfor both the general public and politicians, and those directlyaffected by the incident. Focusing on the relationships betweenthe mass media and politics, they have also demonstrated thatpolicy responses tend to be more generous in a state wherechanges in political power occur more frequently. This isalso consistent with the "principal-agent" model describedearlier showing that politicians, when faced with greaterpolitical competition, have greater incentive to respond properlyto the needs of the general public and earn a good reputation.

Media and policy response:Empirical analysis of U.S. case
Stromberg (2001, 2002) conducted empirical analysis on therelationships between the mass media and policy responses,focusing on the United States of the 1930. He examined therelationships between the spread of radio and the allocationof funds under the Federal Emergency Relief Act (FERA), anincome support program for low-income earners that was implementedin an early phase of the New Deal Plan. The FERA program,which ran from 1933 to 1935, at its peak covered 16% of theU.S. population. There were substantial gaps in the degreeof radio penetration, because implementation of the programcoincided with the rapid increase in use of radios in U.S.households. This provided favorable conditions for investigatingthe relationships between the spread of radios and policyresponses. Using county-level data, he controlled unemploymentrates and other variables indicating the degree of necessityof income support. Still, a statistically and economicallysignificant correlation was found between the ratio of householdsthat possessed a radio and the amount of expenditures on incomesupport, with a 1% rise in the ratio of radio use pushingup the amount of expenditures on income support by 0.52%.Likewise, there was a statistically significant correlationindicating that a county with a higher voter turnout ratiotends to receive a greater allocation of relief funds, thus,demonstrating that a government is more responsive where thereis greater participation by voters in politics.

Media captured by government
The media, however, are not always free and independent. Thereare many cases in which they are taken in by the governmentand/or politicians (media capture). Why does this happen?First of all, there is an incentive for the government, whichwishes to be treated favorably by the "monitor" that is themedia. Thus, a government tries to bring the media onto theirside by cajoling and coaxing them. Meanwhile, the media, fortheir part, may be willing to be taken in so as to receivemore favorable treatment from the government, which imposesnot only explicit restrictions but also discretionary controlover media activities for the sake of improving the welfareof consumers as receivers of information provided by the media.

A typical example of give-and-take relationships, in whicha government influences the media and the media are willingto be influenced, can be seen in a government practice ofleaking important information to a particular media organization.Given the exclusivity of the information, the media organizationwould then run an article on it or broadcast it, and it maytry to be taken in further by the government so as to receivemore scoops over other media organizations. Here, the "informationrent" derived from the rarity of that particular piece ofinformation is being split between the government that hasleaked the information and the media organization that obtainedit. As such, the relationship remains stable as long as itbrings benefits to parties on both sides. The fact that agovernment is able to attract the media, however, means thatit is not proactive in information disclosure and lacks transparency.In other words, by failing to fulfill its duty of informationdisclosure, a government is able to bring the media on toits side in spite of their original mission as a promoterfor the improvement of transparency.

It is important to recognize the presence of a strong complementaritybetween the role of the media and government transparency.Pulling out of the "trap" of small media role and low governmenttransparency is difficult, however, once the current changesits course, positive forces begin to work. Government transparencyimproves in tandem with the increase in the role of the media,which in turn further strengthens the media and turns it intoa powerful "monitor." Actually, an analysis of country-by-countrycross-sectional data finds that a country with a higher degreeof media freedom (measured by a comprehensive index incorporatingfactors such as legal restrictions over content, politicaland economic influence, and suppression of the media, Source:Freedom House) tends to have a lower corruption level (corruptionindicates a lack of government transparency) (Ahrend, 2002;Brunetti and Weder, 1999).

3. Features and Roles ofMass Media as Industrial Organization

What conditions are necessary for the mass media to fulfilltheir role as a "political monitor" that is free and independentfrom the government? Let us consider this question by focusingon the organizational characteristics of the media, marketstructure and regulations.

Increasing returns, monopoly,and government ownership
The prime role of the media is to produce and distribute "information."While fixed costs for the production and distribution of "information"amount to a large sum, the marginal cost - the cost requiredto produce and distribute an additional unit of "information"- is extremely low. That is to say, mass media is a type ofindustry that follows the law of increasing returns. In thecase of a newspaper, for instance, the costs of gatheringinformation, writing and editing articles for its initialpublication are large. But once such fixed costs are borne,the marginal cost of additional copy is limited to the costsof printing and distribution. Given such technical factors,the degree of market concentration in the mass media industrybecomes inevitably high, necessitating the government to implementsome sort of economic restrictions (such as price restrictions).

People holding the view that only a government is capableof maximizing the welfare of consumers often contend thatgovernment-owned media are preferable to privately-owned media.For one thing, they say, government ownership of media wouldat least prevent the occurrence of a natural monopoly. Theyalso argue that government-owned media would be more accurateand less biased than commercial media which, they say, tendsto target certain interest groups or specific layers of consumers.Indeed, with regard to providing a certain type of information- such as arts and education - which may not be commerciallysuccessful but is nevertheless valuable to the general public,government-owned media may be more desirable. On the otherhand, however, there is a fundamental question as to whethergovernment-owned media are capable of playing their role of"monitoring" the government. There must be at least some formof guarantee (for instance, in terms of funds) to secure theirindependence from the government.

Product differentiation,competition and private ownership
The model of information production described above definesinformation very much as homogeneous goods. In reality, however,the mass media are required to convey diverse and differentiatedinformation to serve people who greatly differ in tastes andpreference. Given this, it is important to create a situationwhere a number of relatively small companies compete witheach other in differentiating information and find a niche,instead of letting a small number of mega-media corporationsmonopolize the market. Local newspapers' emphasis on localnews is an example of information differentiation.

Meanwhile, some people believe that a government does notnecessarily maximize public welfare, pointing to the tendencyof ruling party politicians to pursue their personal interestsby restriction of information disclosure to the general public.From this standpoint, government ownership of media is detrimentalto the enhancement of government transparency. Those holdingthis view rather believe that people (on average) would beable to access less biased and more accurate information whenthere is competition among privately-owned media corporations.Many countries, based on this view, are implementing deconcentrationmeasures to ensure the presence of a variety of media organsthereby preventing monopolistic control of the industry.

State-owned media versusprivately-owned media
The question of who should own the media is most empiricalin character. If the media are to secure freedom and independencefrom the government, serve as a "monitor" for government activities,and adequately provide information to the general public,what kind of ownership is desirable? Should they be ownedby a government or privately-owned? In examining this question,Djankov et al (2002) analyzed sample data taken from mediaorganizations (newspaper publishers and broadcasters) in 97countries. Their collected data show that only 4% of the mediaorganizations surveyed had dispersed ownership. Family-controllednewspapers and TV broadcasters respectively account for anaverage 57% and 37% of the totals, while those under statecontrol represent 29% of newspapers and 60% of TV broadcasterson average. The ratio of state-owned companies is higher inTV broadcasting than in the printed media. This is attributableto the inherent nature of TV broadcasts, that is, their beingat least in part nonexcludable and noncompetitive. Anotherreason is the higher fixed costs for TV broadcasters - asis the case with facility costs - than those of the printedmedia, which result in TV broadcasters' greater dependenceon the economies of scale. Meanwhile, in small markets suchas those serving remote areas and ethnic minorities or thosefor education programs, broadcasting services may be under-providedif the task is entirely left to privately-owned TV stations.On the other hand, from the viewpoint of restrictions of content,a government may find it more rational to hold shares in TVbroadcasters, whose contents can be distributed on a real-timebasis and therefore are more difficult to censor.

In their analysis of cross-sectional data on the 97 countries,Djankov et al (2002) demonstrated that the ratios of state-ownedmedia are higher in a country with "autocratic" characteristics,a tendency observed regardless of the type of media. Thiscorrelation is statistically significant even when variables- the stage of economic development, education level, theratio of state ownership in other industries - are controlled.Likewise, there is a statistically significant correlationbetween the degree of media oppression (as measured by thenumber of journalists arrested and media organizations forcedto close down) and the extent of state ownership of mediaeven when several variables are controlled. Government corruptionand political freedom (measured in a comprehensive index forthe freedom and fairness of elections as well as for inter-partycompetition) respectively show positive and negative correlationswith the shares of state-owned media on a statistically significantlevel. All these analytical results illustrate that the moreprevalent state-ownership of the media is, the more the freedomof the media is hampered in general, and consequently showsa strong tendency to negatively influence political freedomand development as well as government transparency.

Role of foreign media
Meanwhile, foreign media are antithetical to state-owned media.Foreign media, because they are less likely to be capturedby the government of a host country, are perceived to havegreater freedom and independence. Like Djankov et al (2002),Besley and Prat (2001) used country-specific cross-sectionaldata to show that there is less corruption (shown in threeindexes) and greater freedom of press in countries which havinga greater presence of foreign media. All these findings pointto the importance of competition among privately-owned media(in particular, those owned by foreign capital) to providefair and accurate information, rather than having state-ownedmedia play the major role, in order for the media to fulfilltheir "monitoring" role.

4. Restrictions over MassMedia

Finally, I would like to examine media restrictions, focusingon the viewpoint of the role of "free and independent" media,as a "monitor" over the government, to provide adequate informationto the general public and improve government transparency.

First of all, based on the reasoning that the economies ofscale in information production and distribution, as wellas the content of information provided, need to be regulated,excessive government involvement is detrimental to healthycompetition among the media through which the diversificationand differentiation of information would occur, thus, preventingthe improvement of information quality and heightening thepossibility of the media being captured by the government.For instance, it is quite common that TV broadcasters andradio stations are required to get a license to operate. Sucha licensing system, however, would easily lead to media captureand might hamper competition among the media. If a governmentis to make a discretionary judgment as to whether or not torenew a license, it would be quite possible for the governmentto take not only explicit conditions but also implicit conditions(for example, providing contents in favor of the government)into consideration. In 1987, South Korea abolished the licensingsystem for newspapers, shifting to a registration system.As a result, the number of daily newspapers increased fromsix to 17 and the contents they covered diversified substantially(World Bank, 2002).

Pros and cons of restrictionsover media ownership
In order to secure diversity of the media, many countriesimplement various restrictions over media ownership. Theserestrictions surely pose obstacles to the realignment of themedia industry through bold and speedy mergers despite thedrastic ongoing changes in the environment surrounding theindustry. At the same time, however, it is also true thatthe diversification of the media has been steadily progressingthrough the proliferation of the Internet and the rapid shiftto broadband, as well as through the spread and developmentof broadcasting satellite (BS), communication satellite (CS),and cable television (CATV) network services. In this regard,the necessity for restrictions over media ownership as a meansto promote the diversification of the media and the informationprovided by them is diminishing. In other words, it is abouttime that the whole world rethought and overhauled media ownershiprestrictions. Under the current media ownership rules in theUnited States, for instance, four major TV networks are prohibitedfrom merging with one another, a single company may not ownTV stations that collectively reach more than 35% of U.S.households, and a single company may own or control no morethan two TV stations in one service area. In June this year,however, the Federal Communications Commission adopted newrules that would raise the ownership limit to allow a singlecompany to own TV stations reaching no more than 45%, insteadof 35%, of households. The move, however, invoked fierce oppositionfrom some members of Congress, stirring up political disputesand putting a stay on the FCC-proposed rule changes.

Japan also holds the principle of mass media deconcentration,under which a single company is allowed to own or controlonly one broadcasting station in one service area. Local broadcastingstations, however, now face an acute need to reinforce theirfinancial foundation so as to bear the massive cost of investmentfor the digitalization of terrestrial broadcasting servicesand to cope with the dwindling advertisement revenue amidthe growing concentration of advertising funds on a handfulof "key" stations (major terrestrial broadcasters coveringthe greater Kanto region). Given such circumstances, the regulatorsare now considering easing ownership restriction for localbroadcasters, for instance, by raising the ceiling for a localstation's shareholdings in another local station and allowingfull subsidization under certain conditions, that is, in caseswhere one local operator absorbs another serving the neighboring,closely-related area. (It has been decided to ease ownershiprestrictions for BS digital broadcasters.)

In Japan, local stations across the country form part ofa network led by a key station (and its related newspaperpublisher) and those within the same network quite often broadcastthe same programs. Also, there are cases in which a localstation receives assistance from the key station of the networkto which they belong. Even in such cases, however, the localand key stations must remain separate entities because ofthe principle of mass media deconcentration. As it appears,the point of the principle is more on the securing of "locality,"based on the principle of having one broadcasting stationin each prefecture to dispatch locally-oriented informationfor the sake of the local audience (and politicians). Giventhe reality of such virtual affiliation or interlocking amongthe media, it is difficult to promote the diversificationof the media by enforcing a principle of mass media deconcentration.Instead, the regulating authorities should ease ownershiprestrictions so as to promote the realignment of the mediaindustry, thereby enhancing competition among the media inproviding diversified views (especially in news programs).

Contradiction of restrictionsover foreign ownership
Another important pillar of media ownership restrictions isthose over foreign ownership that have been implemented inconsideration of foreign-owned media's influence on publicopinion. Like other countries, Japan does maintain restrictionsover foreign ownership in the broadcasting business. However,the restrictions - a 20% cap on foreign ownership - applyonly to terrestrial broadcasting service providers, BS broadcasters,and CS broadcasters using the new satellite at 110 degreeseast longitude (broadcasting services which are accessibleto those possessing a BS receiver). In contrast, CATV operatorsand CS broadcasters using the older satellites at 124 and128 degrees east longitude - including SKY PerfecTV groupcompanies - are no longer subjected to such restrictions.Thus, there exist two types of broadcasting service companiesin Japan - those subjected to foreign ownership restrictionsand those not - showing an apparent regulatory contradictionwith regard to the foreign media's influence on audience.With all these factors taken into consideration, it is suspectedthat the foreign ownership restrictions are maintained simplyto protect certain types of broadcasters - those closely linkedto politicians - from a possible takeover by foreign capital.As demonstrated by Besley and Prat (2001), a greater prevalenceof the foreign media can lead to greater freedom of the pressand help improve government transparency. Thus, drastic reviewand deregulations must be carried out, not only on the principleof mass media deconcentration, but also with regard to foreignownership restrictions.

Independence of state-ownedmedia
Finally, let us take a look at problems concerning state-ownedmedia. As pointed out by Djankov et al (2002), a greater presenceof state-owned media may be detrimental to the improvementof government transparency. Thus, state-owned media organizations,while being required to improve their own transparency andaccountability, must be explicitly granted independence fromthe government just as a central bank is. The BBC (BritishBroadcasting Corp.) of the United Kingdom, for instance, hasits operations almost entirely funded by government-approvedsubscription fees and its governors appointed by the government,just as is the case with the NHK (Japan Broadcasting Corp.).The BBC's articles of association, however, explicitly stateits status as an independent organ from the government, stipulatingthat its internal management - including the contents of programsand the timing of broadcasting them - is free from governmentinterference (World Bank, 2002). The BBC is currently in conflictwith the British government over its controversial reportin May, in which the public broadcaster said that an officialdossier from September 2002 about Iraqi weapons had been embellished.The very fact that the BBC is in conflict with the Britishgovernment, regardless of the authenticity of the information,stands as evidence of the BBC's independence from the government.

Government subsidies to privately-owned media should be minimizedas such a practice would also easily lead to media capture.In Germany, for instance, law prohibits the government fromproviding direct subsidies to news organs, a measure meantto prevent the independence of the media from being underminedby the government (World Bank, 2002).

Considering the "principal-agent" relationship between agovernment/politicians and people/voters in light of the pointsdescribed above, a free and independent media, by providingthe general public with information on government and politicalactivities, can enhance government policy responses and helpimplement governance thereby preventing government officials(politicians) from pursuing personal benefits (corruption).Based on the analytical results of data on various countries,there is a greater likelihood for the media to secure trueindependence and play a proper monitoring function in a countrywhere state-owned media have a smaller presence and foreignmedia a greater presence. It has long been perceived thatrestrictions on the media are being implemented primarilyfor the benefit of readers and audience, who are the end usersof information provided by the media. In reality, however,there are many cases in which restrictions are kept in placesimply to maintain the existing order of the media industry,often under political considerations. Restrictions implementedin such a manner often lead to media capture, where the mediais taken in by the government, and would eventually lead tothe media losing its freedom. It is therefore necessary torethink and overhaul media restrictions, shifting focus ontothe people role as "voters" - those who use information toactively participate in politics and exert influence on thegovernment's information disclosure and policy decisions -rather than simply in their capacity as "consumers."

Augst 4, 2003

>> Original text in Japanese


Ahrend, R. (2001), "Press freedom, human capital and corruption",mimeo.

Besley, T. and R. Burgess (2002), "The political economyof government responsiveness: Theory and evidence from India",Quarterly Journal of Economics 117(4), pp 1415-1452.

Besley, T., R. Burgess, and A. Prat (2002), "Mass media andpolitical accountability", The Right to Tell: The Roleof Mass Media in Economic Development, World Bank.

Besley, T. and A. Prat (2001), "Handcuffs for the grabbinghand?: Media capture and political accountability", mimeo.

Brunetti, A. and B. Weder (1999), "A free press is bad newsfor corruption", mimeo.

Djankov, S., C. McLiesh, T. Nenova, and A. Shleifer (2002),"Who owns the media?", mimeo.

Dreze, J. and A. Sen (1989), Hunger and Public Action,Oxford University Press.

Sen, A. (1984), "Food battles: Conflicts in the access tofood", Food and Nutrition 10(1), pp 81-89.

Stromberg, D. (2001), "Radio's impact on public spending",mimeo.

Stromberg, D. (2002), "Distributing news and political influence",The right to tell: The Role of Mass Media in Economic Development,World Bank.

World Bank(2002),World Development Report 2002: BuildingInstitutions for Markets (Chapter 10, "The Media").

August 4, 2003