Content-Type: text/html

                                       Effects of Communication on Economic and Political Development    










Effects of Communication on Economic and Political Development:
A Time Series Analysis

By

Dr. Kim A. Smith

Greenlee School of Journalism and Communication
122 Hamilton Hall
Iowa State University
Ames, IA.  50011

<a href="/cgi-bin/wa?LOGON=A3%3Dind0101B%26L%3DAEJMC%26E%3Dquoted-printable%26P%3D4849309%26B%3D--%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D%253D_792627%253D%253D_%26T%3Dtext%252Fplain%3B%2520charset%3Diso-8859-1" >[log in to unmask]</a>
 515-294-0482



Paper submitted to the Theory and Methodology Division of the Association for Education in Journalism and Mass Communication

April 1, 2000





















Abstract

Using time series data from 1965-1996 in 107 nations, this paper reexamines the effects the social, mass media and telecommunication indicators on economic and political de-velopment.  Among the indicators of social development, urbanization showed a much stronger relationship with economic and political development than percent illiterate or school enrollment.  Newspapers, radios, and televisions per 1,000 people all substantially predicted economic and political development. Only telephone mainlines per capita among the telecommunication indicators significantly covaried over the 32-year period of the study with both types of development.      




























Effects of Communication on Economic and Political Development: A Time Series Analysis

The geopolitical forces that shaped the world after World War II also signifi-cantly influenced the research agendas in the political science and communication fields.  As Wiarda (l989-90) has described, the emerging Cold War and the newly formed nations freed from colonial dominance in the l950s resulted in research in both fields being sponsored and used by policymakers.  The foreign policies of developed nations were concerned with maintaining the political development of these newly independent nations.  The United Nations, at the same time, began to make huge investments in the mass media infrastructures to enhance their economic development.[1] This paper reviews this research from the political science and communication fields and then offers a new analysis of factors influencing economic and political development from the perspective of the l990s.
Early Studies of the Political Development Paradigm
 Pye (1956) provided one of the first analyses of the importance of commun-ication in the political development process.  From his perspective, the general com-munication process in developing nations performed a key function in structuring the political process.  Based on case studies, he argued that the influence of communication on political structure differed in urban and rural areas.  The Western-style mass media in urban areas tended to cover the political process from a national perspective, thus enhanc-ing a sense of nationalism and support for developing political institutions.  In more rural areas, where word of mouth was still the predominant form of communication, villagers remained oriented to local events, their opinions on which were influenced by informal opinion leaders.  Pye noted that the urban frames of reference of newspapers had little meaning to them, when they were read to them in their villages.
Seymour Martin Lipset (l959) focused on the relationship between economic development and the legitimacy of political systems.  Lipset hypothesized that demo-cratic systems of government were more likely to flourish in wealthier nations, because only in them could citizens be expected to participate intelligently in the political process and avoid the appeals of demagogic authoritarian leaders.  A country divided between a large impoverished mass and a small wealthy elite would likely result in political systems based on oligarchy or tyranny, according to Lipset.
        Lipset (l959) provided correlational evidence that supported this hypothesis for Latin American, European and other English-speaking nations. Stable democracies in these regions were higher than unstable democracies or dictatorships on all the main "indicators of wealth," as Lipset referred to them: per capita income, doctors, motor vehicles, telephones, radios and newspapers per 1,000 people.[2]  Urbanization, literacy, school enrollment, per capita energy consumption, and the percentage of nonagricultural workers were also higher in the countries with stable democratic regimes. 
        Similarly, Phillips Cutright ( l963) tested the proposition that social, economic, communication and political development were interrelated.  To create an index of political development, Cutright assigned scores to 76 nations from l940 to l961, based on the degree of competition present among political parties and in elections of chief executives.  Communication development, which was measured by an index of daily newspapers per 1,000 people, newsprint consumption, telephones and pieces of domestic mail per capita, was the strongest predictor of political development, followed by degree of urbanization, school enrollment and the percentage of the labor force engaged in agricultural work. 
        Karl Deutsch (l956) conducted among the first studies to measure communication flow among countries.  His analysis of first-class mail flow among 48 countries prior to WW II showed that more than 75 percent received more letters from abroad than they sent.  In the seven most industrialized countries, however, this ratio was reversed. Ratios of local to non-local mail also tended to decrease in larger US cities, as did those for  non-local to foreign mail. Such national measures of "intake-output ratios," Deutsch argued, served as indicators of integration among countries.
Criticisms of the Political Development Paradigm
        These and other studies generated considerable research from which emerged the political development paradigm, which posited that an array of socioeconomic variables, once they reached certain levels, would lead to more democratic and stable political regimes (see also Shannon, l958; Schnore, l961).  The practical implications of this research for guiding the development aid programs of world powers and the UN in developing countries generated considerable debate in the l960s. Wiarda (1989-90) summarized the issues this way:
ù     The political development paradigm was derived from Western experience and therefore was of limited usefulness in non-Western areas.  As a result, the optimum outcome of political development was too often narrowly defined in terms of Western-style voting participation rather than a broader concern for political and civil freedoms..  
ù     The timing and sequences of stages of political development may differ between developed and less developed nations or from region to region.
ù     The complex webs of interdependencies among nations, which may affect their economic and political development, were not sufficiently investigated.
ù     The role indigenous cultural or social institutions could play in promoting change and helping bridge the gap during the transition from traditional to modern societies were not studied.  Some formulations of the political para-digm suggested that indigenous cultural or social institutions should be replaced by those from Western countries.
Some political scientists, in fact, argued that developing democratic political institutions too swiftly could hinder economic development.  A proponent of this view was Huntington (l968), who stated that political democracy may be a luxury that developing nations could not afford.  He argued that the pressures of participatory democracy can quickly overwhelm the political institutions of developing countries, which tend to be fragile to begin with,, thereby magnifying the forces of instability.  Therefore, more authoritarian regimes were better able to cope with the stresses of economic development.        
Taken as a whole, the evidence concerning the political development paradigm was decidedly inconsistent.  Sirowy and Inkeles (l990) reviewed studies that examined the economic consequences of different types of political regimes.  Of the 13 studies they examined, three supported the conclusion that democracy may hinder economic develop-ment, six reported no relationship between democracy and economic growth, and four indicated a conditional relationship.  Furthermore, half of another 12 studies indicated that democratic regimes actually reduced economic inequality or had no effect, while the other half suggested an inverse relationship between the two indicators. 
Early Studies of the Development Communication Paradigm 
An important focus of research in the communication field in the l950s and l960s was how mass media technology could be used to stimulate and modernize the econo-mies and political systems of developing countries.  The development paradigm that evolved from this research posited that indigenous people needed to abandon their trad-itional ways of thinking and life on subsistence farms to participate in the cash sector of the economy (see, for example, Melkote, l991; Moemeka, l994).    The quickest and most effective way of bringing this about was argued to be information campaigns dissem-inated via the mass media (See Schramm, l964; Rogers, 1976).
        By the far the most influential early study shaping the development commun-ication paradigm was undertaken in the Middle East by Daniel Lerner (1950).  Lerner theorized that economic and political development resulted from a process that began when a country's population was about 10 percent urbanized.  At that point, mass media and literacy growth interacted in a "multiplier effect" to produce a critical mass of citizens with "mobile personalities;" that is, citizens who desire to participate in the economic and political life of their country.  When a critical mass of such citizens is reached, which he predicted to be at about 25 percent urbanization, then economic development takes off, ultimately resulting in a democratic political regime, which Lerner termed the "crowning glory" of the process.
His secondary analysis of census data from Turkey, Lebanon, Egypt, Syria, Jordan and Iran indicated that the percent voting in these countries was highly correlated with degree of urbanization, literacy, and media consumption.  Survey results indicated that "transitionals," those who were in the process of moving from a traditional to more modern lifestyles, tended to most have recently moved from a rural to urban home, had the highest exposure to media, and the most understanding and empathy for the role of the media.  Furthermore, transitionals in the six countries had the highest recall of international news items, the greatest variety of opinions, and were most likely to believe that they could influence national problems. 
Tests of the Lerner Model
Correlational evidence appeared to support the Lerner model of economic and political modernization.  Per capita income, proportion of adults literate and daily news-paper circulation, as gathered from l959 UNESCO data on 85 countries by Nixon (l960), were strongly correlated with degree of press freedom as rated by the International Press Institute.  A regression analysis by Farace and Donohew (l965) showed that press control, newspaper circulation and radio receivers per 100 people were significantly predicted by illiteracy rates, per capita income, ideological orientation of the country, population, life expectancy, and the legislative-executive structure in 115 countries in l962.  Economic and media development indices dominated in a factor analysis by Farace (l966) of 54 political, economic and cultural indicators in l09 countries in l964.  A later analysis by Farace (l966) demonstrated that six distinct regional patterns existed among this group of indicators of national development. 
However, causal analyses testing the Lerner model produced mixed results.  Alker's (l966 ) path analysis of aggregate national data from the early l960s showed that urbanization predicted literacy and media development, while literacy tended to predict economic and political participation indirectly through media development.  Another path analysis by McCrone and Cnudde (l967) tested six alternative models of the basic vari-ables contained in the Lerner model.  The best causal fit to their data was one in which urbanization leads to education, which directly predicted communication development, which in turn led to higher voting participation.
Time series analyses provided stronger evidence on the Lerner model.  Based on cross-lagged correlations from UN data from l951 through l961, Schramm and Ruggels (l967) concluded that per capita income and urbanization affected literacy in early stages of development more than the reverse case.  Yet, the causal relationships among per capita income, urbanization, literacy and indicators of newspaper and radio development were not clear in their analysis.  
Winham (l970) tested the Lerner model by utilizing time-series data on 10-year intervals in the United States from 1790. An index of political development was created for each 10-year period from a sum of scores for minority parties in the US Congress, a directly elected chief executive with a vote participation multiplier.  Communications development was operationalized as an index of post office expenditures, total number of newspapers and periodicals, as well as telephones, radios and television per capita.  Time-lagged correlations indicated that communications were a stronger predictor of political development than vice versa.  The same was found for the relationship between political development and education, but only after a 20-year lag.  On the other hand, the US data indicated that communications development may precede urbanization, contrary to the Lerner model.
Criticism of the Development Communication Paradigm  
As with the political development paradigm, a reassessment of the commun-ication developmental paradigm among communication scholars began in the late l960s (see Schramm & Lerner, 1976; Jayaweera & Amunugama,1987).  Also in part motivated by the policy implications of the research, the debate on the development communication paradigm produced criticism similar to that of the political development paradigm.  Ac-cording to Jayaweera (l987), these criticisms included: 
ù     Development was far more than economic growth in a country, but requires equal distribution of wealth also.
ù     Promotion of political participation at the national and local levels, educational opportunities, civil liberties, as well as factors pertaining to quality of life, should also be important goals of development communication.
ù     The development communication paradigm tended to replicate the individualism and materialism of the Western society. Instead, mass communication campaigns should be conducted within the context of the culture, customs and values of the people.
In reality, after a decade of heavy investment in the mass media infrastructures of developing nations, the results were mixed.  In 1975, Sathre (1976) provided evidence that in the past 10 years the GNP in developing countries had increased more than 50 percent, ex-ports to them were up 124 percent, nonagricultural employment had increased 52 percent and food production was up 30 percent, among other indicators.  During the same period of time, mass media development had also significantly increased in these countries, including a 100 million more radio receivers (Schramm, l976).  
However, while overall economic development did increase in real terms, the gap between rich and poor grew, population rates continued to skyrocket, unemployment rose and the generally quality of life indicators improved only for those of the highest socio-economic status (Melkote, 1991; Jayaweera & Amunugama, l987). Moreover, economic development had not necessarily lead to political stability and democracy, as many of the newly created democracies during the l960s were replaced by military dictatorships or one-party regimes (Wiarda, 1989-90).  Lerner's (1967) explanation for these develop-ments was that the "want/get" ratios among the populations of developing countries had become so disproportionate that the predicted "revolution of rising expectations" pre-dicted by the development communication paradigm had apparently resulted instead in a "revolution of rising frustrations." 
Some Methodological Considerations
Early tests of both the political development and develop communication para-digms were generally not adequate to test the hypothesized causal relationships among indicators (see Frey, l973).  Most had to utilize either cross-sectional data or time series of short duration, because organizations such as the World Bank and UNESCO had just begun compiling yearly statistics in the l960s.  Furthermore, statistical techniques for assessing causality among indicators were not yet highly developed, as well as the computer software and hardware necessary to perform them (see Schramm & Ruggels, l967).  Evidence from early studies of both paradigms, therefore, should be replicated, utilizing the greater variety of time series of longer duration now available.
        Another consideration is the extent to which the evidence regarding the political development and communication development paradigms is historically bound.  The studies of both paradigms conducted in the l960-l970 era, when the development in most countries had been underway for only about a decade (Schramm & Lerner, 1976; Wiarda, 1989-90). Increases in economic and political development might well have lagged behind increases in these socioeconomic indicators by long periods of time.  Moreover, both economic and political stability in developing countries might also take varying lengths of time achieve, preceded by irregularly occurring periods of disruption.  Only time-series data of long duration can be used to assess the lag between stimulus and response in this way.
More Recent Analyses
Surprisingly little macro-level research, however, was published regarding the political development and development communication paradigms after the l970s. [3] Recently, however, the new struggling democracies in Eastern Europe and the intro-duction of new communication technologies with their economic and political impli-cations have resulted in renewed interest in this area of research (see Jayaweera & Amunugama, 1987).  Many of these studies have used considerably broadened con-ceptualizations of both economic and political development.
Political Development Studies
Fedderke and Klitgaard (l998) found that GDP per capita was associated with greater civil and political rights, when measured by a number of socioeconomic time series up to 25 years in length for as many as 118 countries.   GDP per capita was also significantly associated with higher levels of education, greater likelihood of a market economy, a higher percentage of the population exposed to modernizing influences, and greater consumption of the news media.  In a multivariate analysis, political rights proved to be the strongest predictor of per capita GDP.  (News media consumption was not included in the multivariate analyses.)
Borner, Brunetti and Weder (l995) noted that most previous operationalizations of democracy in studies of economic development have focused on the existence or fairness of elections, competition among political parties, or the freedom of opposition groups to operate.  But such dimensions of democracy, they argued, are of low relevance to inves-tors in developing countries' economies, who are more concerned with the uncertainties posed by political instability in both democratic and authoritarian regimes. 
        Accordingly, Borner et al. designed a measure of political credibility to tap the degree of uncertainty arising from state actions.  Entrepreneurs operating in developing countries were sent questionnaires asking them how regularly they had to cope with unexpected laws and policy changes and the extent to which governments stick to an-nounced major policies in the countries in which they operated.  Respondents indicated that these factors were more important in their investment decisions than high transaction costs or corruption.  In 10 developing countries, political credibility proved to be more highly correlated with economic growth then other frequently used measures of demo-cracy.  
Development Communication Studies
A number of studies have explored the relevance of telecommunication develop-ment to political and economic development.  Using data from 45 countries, Hardy (l980) found that GDP and the number of telephones per capita mutually influenced each other between l960 and l973, based on time-lagged offsets of one year.  Similarly, Cronin, Parker, Colleran and Gold's (l991) analysis of 31 years of US data from l958 through l988 showed a two-way effect between telecommunications investment and overall economic growth, based on two-year lag intervals. Using data from 23 countries from l975 to l991 from several sources, Zhu's (1996) pooled time-series analysis showed that the previous year's telephone develop-ment but not radio or television was a significant predictor of GNP.  But with a three-year lag, television and radio penetration also become significant predictors of GNP.
                The Japanese Johoka Shakai approach to the study of development commun-ication takes into account the role of new media technology (see Dorick & Wang, 1993).   As Ito (l980) described it,  johoka shakai translated means "informationalized society," as opposed to kogyoka shakai or "industrialized society." The Johoka Index was composed of a simple, summed index of indicators in four categories: 
ù     amount of information (telephone calls per person per year, newspaper circulation and books read per 1000 people, and population density); 
ù     distribution of communication media (telephone receivers per 100 people, radios and television sets per 100 households); 
ù     quality of information activities (proportion of service workers in labor population and proportion of students in total appropriate age group; 
ù     the information ratio of each country (a ratio of household expenditures for information-related activities to total expenditures) 
The first analyses in the l970s showed that each of the five major industrialized societies-US, Japan, UK, West Germany, and France-had significantly increased on all the johoka measures between l953 and l975.  While the United States was the most informationalized society throughout this period, Japan's growth, according to the  johoka measures,  was steepest, resulting in the country being ranked No. 2 ahead of the United Kingdom, West Germany and France, respectively, by the late l960s.  Also, the information ratio index was highly correlated with per capita income in a broader sample of developed and developing nations.
Dorick and Wang (l993) argued that development communications has been replaced in this era by development informationization.  Using measures from the measures the Johoka Index, the results showed that telephone and computer/data terminal penetration grew significantly between l970 and l980, but the rate of penetration was greater in richer than poorer countries.  Differences in mass media penetration rates among nations were less dramatic compare to telecommunications, computers and data terminals. While television and radio diffusion slowed in more developed countries in the l970s, as they reached saturation levels, their penetration rates continued to grow in the less developed countries, with the exception of newspaper circulation per 1000 people.
 Among the eight most highly industrialized nations, Dorick and Wang's analysis showed that by l980 the information sectors constituted at least 40 percent of their work forces, while wide differences existed in this proportion among less wealthier nations.  The average percentage contribution of the information sectors to the GDP per capita of wealthier nations rose from 43 percent in l980 to 65 percent in l988.  While a less impres-sive contribution by the information sector to GDP per capita was evidenced for middle-level countries, none was apparent among those with the lowest incomes. 
Factors for Further Study of Economic and Political Development
Taking into consideration both early and later studies of the political and develop-ment communication paradigms, a number of factors emerge that should be taken into account in future studies:
ù     Extensive time-series data, which now are available, are necessary to study adequately the influence of socioeconomic and communication indicators on economic and political development.
ù     Both economic growth and distribution of wealth should be included in de-velopment studies.
ù     Political development is a more complex variable than is indicated by voting participation rates.  Not only the type of regime should be considered, but also whether the political system has regulations in place to maintain political stability. 
ù     Both mass media and telecommunication development, such as access to computer technology, telephone lines, and international calls should be compared in studies of political and economic development.
  Hypotheses
        Based on the analysis of evidence above, these hypotheses were tested:
H1: Over a 32-year period, economic development will be significantly related to social, mass media and telecommunication development indicators.
        Based on past studies, it is expected that economic development in general will be related to the social, mass media and telecommunication development of a country (Sirowy & Inkeles, 1990; Fedderke & Klitgaard, l998).  However, comparisons will also be made among indicators of level of economic development, growth and income inequality.  Growth and income inequality indicators may well be related to social, mass media and telecommunication development than overall level of economic development (Sirowy & Inkeles, l990).     
H2: Over a 32-year time period, political development will be significantly related to social, mass media, and telecommunication development indicators.
This hypothesis again predicts that relationships among these variables will replicate those found in earlier studies of political development (Lipset, l959; Cutright, 1963).  However, comparisons will be made between general regime type (degree of democratization) and the degree to which recruitment of chief executives and partici-pation in the political process are regulated, as suggested by Borner, Brunetti & Weder (1995).
H3: After controlling for the social development indicators, both mass media and telecommunication time series will significantly predict economic development indicators.
Past studies have indicated that social development indicators-urbanization, school enrollment and literacy-are strongly correlated with mass media and telecommunication development (Lerner, 1958; Frey, 1973; Fedderke & Klitgaard, l998; Hardy, 1980; Cronin, Parker, Colleran & Gold, 1991).  Accordingly, to assess the independent contribution of mass media and telecommunication development, it is necessary to control for the social development.
H4: After controlling for the social development indicators, both mass media and telecommunication time series will significantly predict political development indicators.
Here again, to assess the independent contributions of mass media and telecommun-ication indicators, it is necessary to remove the correlated influences of social development indicators.
H5:  The political and economic time series time series indicators will be significantly related to one another.
        As described above, the relationship between economic and political development is not clear, based existing evidence.  However, using a broader range of economic and political development indicators, the nature of the relationship between them is expected to be clearer (Frey, l973; Wiarda, 1989-90).
H6: Absent more definitive evidence, causal relationships among the political, economic, social, mass media, and telecommunication time series are expected to follow a pattern similar to predicted by the Lerner (1958) model of development:
        H6a:    Urbanization will precede all other development time series.
H6b:    Mutual influence will be apparent among the mass media, literacy and school enrollment time series. 
H6c:    Increased literacy and school enrollment is expected to lead to greater telecommunication development.
H6d:    The telecommunication time series are expect to have the most immediate direct effects on political and economic development..
H6e:  The economic and political development time series will mutually influence one another.
Methodology
The data used in this analysis were drawn from two archives.  
The mass media, telecommunication, socioeconomic time series used in the analysis were taken from the World Development Indicators on CD-ROM, which contains more than 500 indicators for up to 210 countries covering the years l960-l996.  When alter-native indicators were available, the time series that adjusted for population size and were of the longest duration were chosen.  In addition, the years from l960-64 were deleted from the time series to minimize missing data.      
Mass media time series: 
ù     Daily Newspapers: Number of newspapers published at least four times per week per 1,000 people.
ù     Radio Receivers: Number of radio receivers per 1,000 people in use for broadcasts to the general public.
ù     Television sets: Number of sets in use per 1,000 people.
Telecommunication time series: 
ù     International Telephone Traffic: Outgoing international telephone traffic in minutes per subscriber.
ù     Telephone Mainlines: Mainlines connecting a customer's equipment to the public telephone network per 1,000 people.
ù     Computing Technology  Imports: Communications, computer, information and other related services as a percentage of service imports.
Socioeconomic time series:
ù     Illiteracy: Proportion of adults aged 15 and above who cannot with understanding read or write a short, simple statement on their everyday life.
ù     Gross National Product per capita: Gross national product divided by midyear population.
ù     Gross National Product Growth per capita: Annual percentage growth in GNP per capita.
ù     Income  Inequality:  Gini coefficient, which indicates the degree of income distribution within the population of the country (See Melkote, 1991). The higher the Gini coefficient, the greater income inequality. 
ù     School Enrollment: Total number of pupils enrolled at the primary level in public and private schools.
ù     Urbanization: percentage of the total population living in areas defined as urban at midyear.
Political Development Time Series
The time series indicators of political development were taken from the Polity98 Project: Regime Characteristics: 1800-1998, which is housed at the Center for Inter-national Development and Conflict Management at the University of Mary-land-Coolege Park.[4]  This data set contains information from 1800-l998 on regime and authority characteristics for all independent nations, based on a coding scheme designed by Eckstein and Gurr (l975).   In developing this data set, they were concerned with the state's authority pattern, or "the set of structures and processes through which directives applicable to members of the state are made, issued and supported," rather than simple voting rates. (Gurr, l974, p. 1483). 
Several time series from this data set were selected for use in this analysis:[5]
Democracy: general openness of political institutions, based on a 10-point scale constructed by summing scores from these indicators:
a.) Competitiveness of participation: 1= factional, 2= transitional and 3=competitive.
b.) Executive recruitment competition: 1= dual executive sin which one is chosen by hereditary succession and the other in a competitive election or 2= competitive election between two or more parties or individuals.
c.) Executive recruitment openness: 1= heredity succession plus electoral selection of an effective chief minister or 2= open election.
d.) Executive constraints: 1= unlimited authority; 2= slight to moderate limitations; 3= substantial limitations; 4= executive parity or subordination.	
Two additional measures of authority characteristics were included in the analysis:
Recruitment Regulation: degree to which executive recruitment is regulated through institutionalized procedures regarding the transfer of executive power:
                1= changes in chief executive occur through forceful seizures of power.
2= chief executives are chosen by designation within the political elite without formal competition.
3= chief executives are determined by  hereditary succession or in competitive elections.
Participation Regulation: degree to which institutional structures are developed for political expression:
1= fluid political participation without enduring national political organizations or systematic regime controls in political activity.
2= relatively stable and enduring political groups which compete for political influence at the national level, although competition among them is intense, hostile and frequently violent.
3= regular oscillation between intense factionalism and restriction with one group restricting its opponents power until it is displaced in turn.
4= some organized political participation is permitted without intense factionalism, but significant groups, issues and other types of conventional participation are regularly excluded from the political process.
5= stable and enduring political groups regularly compete for political influence with little use of coercion and no types of political action excluded from the process.
The two data sets were merged, excluding countries that were included on one data set but not the other and which were not in existence throughout the l965-l996 time period. (Appendix A shows the countries which were included in the data set after this procedure by region of the world.)  As a result, a data set of 14 socioeconomic, commun-ication and political time series for 107 countries over a 32-year period was created.  Missing data for time series extending throughout the 32-year period were replaced by predicted values from a regression of the existing series on an index variable scaled 1 to n.  When the time series naturally did not exist throughout the 32-year period, such as for computer technology imports, then missing values where replaced with "0" until the series started, at which point they were replaced as in other cases.   
 Given this amount of data, a pooled cross-sectional time-series model was chosen for the following analysis, in which the selected time series across T= 32 years for the N= 107 countries were combined, resulting in a data set with a total of 3,424 cases (107 countries x 32 years).  Two important advantages of the pooled cross-sectional model are that it allows the N x T cases to be simultaneously analyzed with a single set of parameter estimates and the use of straightforward methods for handling autocorrelated distur-bances, which result in biased and inconsistent estimators in panel models (Hannan & Young, 1977).  
The Analysis Strategy
        The pooled cross-sectional time-series model allows a number of statistical tech-niques to be used to test the causal relationships stated in the hypotheses. Most basically, the covariation between two time series can be assessed either on a bivariate basis or after controlling for the influence of other time series at T-0 or lagged using multiple regres-sion analysis (see, for example, Smith, 1987; Trumbo, 1995).  However, while providing stronger evidence than correlations from cross-sectional data, only weak causal inferen-ces can be made from significant relationships among time series, in comparison to the strong causal inferences that can be made from laboratory experiments, which better control for threats to validity.  In fact, seasonal or periodic trends can result in spurious relationships emerging among time series (see, for example, McCleary and Hay, 1980).  To create stationary time series, each was "whitened" by regressing it on year of observation and then using the resulting residuals in the analysis.[6] One then lagged each of the whitened time series, two and three years.  A preliminary regression analysis indicated that a one-year lag was sufficient to capture the influence of past history on future values of most of the time series.[7]  In the multivariate analyses, the T-0 time series were regressed on T-1 values of the independent values in the equation, again allowing for weak causal interpretations of results.  
Results
        The results shown in Table 1 below provide some support for  Hypothesis 1 that economic development would be related over time to social, mass media and telecom-munication indicators. Over the 32-year period from l965-96, the T-0 newspaper, radio and television time series strongly covaried with GNP per capita and to a lesser extent with GNP Growth per capita and Income Inequality.  Among the telecommunication time series, however, only telephone mainlines per capita was consistently correlated with the three economic development indicators over time, most strongly with GNP per capita.  The social time series were only weakly related to economic develop-ment indicators, with the exception of the relationship between urbanization and GNP per capita. 
Table 1
Correlations Among T-0 Economic, Social,  Mass Media, and Telecommunication Time Series

GNP per capita
GNP Growth
per capita

Income Inequality
School Enrollment
-.09**
.10**
-.03
Percent Illiterate
-.06**
               -.01
-.04
Percent Urbanized
 .51**
               .02
-.02
Newspapers
 .31**
.04*
   -.05**
Radios
 .53**
 .05**
   -.10**
TVs
 .65**
 .07**
                -.14
International Phone Traffic
              -.01
               .00
                 .00
Telephone Mainlines
 .76**
 .07**
   -.13**
Computer Technology Imports
 .05**
              -.01
                -.06
 Note:  N=3424. Negative correlations indicate lower levels of income inequality, based on the Genie Index.
*p< .05; **p< .01.
	
The results shown in Table 2 below support Hypothesis 2 that political develop-ment indicators in a country and procedures would significantly vary with the social, mass media and telecommunication time series between 1965 and 1996.  As with the economic development indicators, newspapers, radios and TVs per 1,000 people in the 107 countries were substantially related to the democracy index, recruitment regulation and participation regulation.  While the international telephone traffic per subscriber and the percentage of computing technology imports were only slightly related to the three democracy time series over time, telephone mainlines per capita quite strongly covaried with them.  Among the social time series, only degree of urbanization was strongly and consistently related to the three democratic indicator time series
Table 2
Correlations Among T-0 Political, Social, Mass Media and Telecommunication Series

Democracy Index
Recruitment Regulation
Participation Regulation
School Enrollment
               -.02
                .00
            -.03
Percent Illiterate
 -.13**
 -.10**
             .02
Percent Urbanized
                .43
                 .38**
.44**
Newspapers
 .19**
 .12**
.14**
Radios
 .43**
 .31**
.28**
TVs
 .47**
 .38**
.36**
International Phone Traffic
               -.03
              -.04*
            -.03
Telephone Mainlines 
 .55**
 .38**
.42**
Computer Technology Imports
               -.11**
              -.02
.10**
Note. N= 3424.
*p< .05; **-< .01

     Table 3 below shows the outcome of a regression analysis designed to test Hypothesis 3 that both the T-1 mass media and telecommunication time series will sig-nificantly predict the T-0 economic development indicators, beyond the influence of the T-1social development indicators.  In this analysis, each of the economic development time series were regressed first on the social time series.  Then the mass media and telecommunication time series blocks were individually entered into the equation and increment to R2 tests conducted to determine if contributed significantly as a block to the variance of the economic time series.        
Table 3. 
Standardized Coefficients from Regression of T-0 Economic Development Indicators on T-1 Social, Mass Media, and Telecommunication Time Series


GNP per capita

GNP Growth
per capita


Income Inequality
School Enrollment
             -.02
.10**
             -.04**
Percent Illiterate
.05**
             -.02
Percent Urbanized
.52**
               .03
-.04**
Increment to R2
.269**
.010**
.006**
Newspapers
               .10*
              .01
              -.03
Radios
               .11*
              .00
                .00
TVs
 .44**
              .07*
 -.18**
Increment to R2
.221**
.005*
.023**
International Phone Traffic
               .01
             -.02
                .00
Telephone Mainlines
 .66**
.08**
 -.16**
Computer Technology Imports
 .05**
              .02
 -.07**
Note. N= 3424. Increments to R2 are for the amount of variance explained individually by the mass media and telecommunication blocks alone, beyond that explained by the social development time series alone. Negative correlations indicate lower levels of income inequality, based on the Genie Index.
*p< .05;p< .01

Not surprisingly, given the size of the sample (N= 4324), even small increments to the overall R2 values in each equation by the mass media and telecommunication blocks were significant.  However, from a practical perspective, both the T-1 mass media and telecommunication blocks significantly predicted T-0 GNP per capita, beyond the influence of the social development time series as a block.  On the other hand, the T-1 mass media and telecommunication blocks only marginally contributed to the variance explained over time in variations in the T-0 GNP growth per capita and income inequality time series.  Therefore, Hypothesis 3 is supported only in regard to GNP per capita.
Using the same analysis procedure,  the results shown below in Table 4 support Hypothesis 4 that mass media and telecommunication indicators would significantly predict the political development time series, after controlling for the social development indicators.  Each of the T-1 mass media and telecommunication time series blacks significantly contributed to the variance in the T-0 democracy index, recruitment regu-lation and participation regulation over the 32 year period of this study.  But the T-1 telecommunication block explained more variance over time in each of the T-0 indicators of political development than the mass media block. 
Table 4
Standardized Coefficients from Regression of T-0 Political Development Indicators on T-1 Social, Mass Media and Telecommunication Time Series


Democracy Index

Recruitment  Regulation

Participation Regulation
School Enrollment
              .03
               .04*
-.07**
Percent Illiterate
             -.03
              -.02
.04*
Percent Urbanized
.45**
               .38
 .13**
Increment to R2
.204**
.148**
.025**
Newspapers
              .02  
 .02
.04**
Radios
.15**
-.03
.07**
TVs
.27**
                .17**
.43**
Increment to R2
.086**
.023**
.126**
International Telephone Traffic
             -.01
               -.03
              .01
Telephone Mainlines
.42**
   .25**
.48**
Computer Technology Imports
-.11**
               -.01
.11**
Increment to R2
.134**
.046**
.178**
Note. N= 3424. Increments to R2 are for the amount of variance explained individually by the mass media and telecommunication blocks alone, beyond that explained by the social development time series alone.
*p< .05;**p<.01

The correlations shown in Table 5 below provide some support for hypothesis five that the democracy and economic development time series would be positively re-lated to one another.  In fact, GNP per capita is significantly related to the democracy index, recruitment regulation, and participation regulation.  Considerably smaller correl-ations are evident between GNP growth per capita and the Genie index of income inequality and the three indicators of political development. 
Table 5
Correlations Among T-0 Political and Economic Development Time Series


Democracy Index

Recruitment
Regulation

Participation Regulation
GNP per capita
.46**
.39**
.47**
GNP Growth per capita
               .05**
               .04**
               .02
Income Inequality
              -.02
              -.01
              -.11**
Note. N= 3424.

Path Analysis
A modified form of path analysis was used to provide a comprehensive picture of the overall pattern of causal relations among the time series. Among the indicators of economic and political development, only the democracy index and GNP per capita were included in this analysis, based on the strength of relationships shown in preceding re-sults.  To calculate the standardized path coefficients, the T-0 values of each dependent variable at each stage were T-1 lagged values of the time series hypothesized to be contemporaneous or precede the dependent variable in the regression.  Among time series variables hypothesized to mutually influence each other in the same time period, two way influence can be assessed.[8]  
Figure 1 on the next page provides the outcome of these procedure

                                    Effects of Communication on Economic and Political Development    


Figure 1. Path Analysis of Causal Relationships Among Development  Indicators


.42

.24

-.20

-.13

School Enrollment

-.22

Percent
Illiterate

.14

.30

.17

.14

.58

.56

.11

.16

.65

.13

-.11

.18

.66

Percent
Urbanized

Phone
Mainlines

TV
Per 1000

GNP
Per Capita

Index of
Democracy

Computer
Technology

International 
Phone Calls

Radios
Per 1000 



Newspapers
Per 1000


















                                    Effects of Communication on Economic and Political Development    

Starting from the left-hand sign of the diagram, some support for hypothesis 6a that urbanization will precede other types of development.  Most striking, direct paths predict all of the other development indicators, with the exception of the three telecom-munication time series.  Urbanization, however, indirectly predicts telephone mainlines per capita through each of the three mass media time series, most strongly through television sets per 1000.    
        The results in Figure 1 do not provide evidence that literacy, school enrollment and three mass media time series mutually influence each other, as predicted in hypo-thesis 6b.  In fact, no direct or indirect paths are emerged between any of the mass media development time series and literacy or school enrollment.  Furthermore, also contrary to hypothesis 6c, the literacy and school enrollment time series do not directly or indirectly predict any of the telecommunication time series.
As predicted in hypothesis 6d, telephone mainlines per capita has a substantial direct influence on both GNP per capita and the democracy index time series. Computer technology imports also significantly but inversely predicts the democracy index.  international telephone calls, however, are not related to either GNP per capita or the democracy index.
Finally, the analysis indicated no causal relationships between the GNP per capita and democracy index time series, contrary to expectations, contrary to hypothesis 6d.  When both the T-0 GNP per capita and the democracy index time series were regressed on thee T-1 values of all other development time series, the substantial bivariate relationships between the two, which was shown in Table 5, dropped to near zero.  

Discussion and Conclusions
This study analyzed the influences of social, mass media and telecommunication time series on economic and political development in 107 nations from 1965-1996.  As such, it attempted to replicate the findings of previous studies of the political develop-ment and development communication paradigms and test causal relationships, using time series of longer duration and statistical methods not available when the earlier studies of these paradigms were conducted.  The results suggest a number of conclusions.
Mass media development, as measured by newspapers, radio sets and television sets per 1,000 people, exhibited stronger causal relationships with both economic and political development than social and telecommunication development.  Among the social and telecommunication development indicators, only the urbanization and tele-phone mainlines per capita time series were consistently related to economic and political development.  These results, therefore, are supportive of the central role given to mass media in the development communication paradigm (Frey, 1973).
However, the strong relationships between telephone mainlines per capita and economic and political development confirm the results of recent studies regarding the importance of telecommunication (e.g., Cronin, Parker, Colleran, & Gold, 1991; Hardy, l980).  These results suggest the importance of improving technology to increase inter-personal communication in developing countries.  That international telephone traffic per subscriber and computer technology imports were unrelated to economic and political development, at least during the time-period of this study, may be because both are costly and thus not available yet to large percentages of the populations in the countries studied. 
Neither percent of population illiterate nor school enrollment, key factors in Lerner's (l958) theory of modernization, were substantially related to any of the indi-cators of economic and political development.  The path analysis also did not show that these time series were causally related to the mass media or telecommunication indi-cators, as predicted by the Lerner model.  Earlier studies showing relation-ships among literacy, education, mass media, economic and political indicators may be limited by the historical era in which they were conducted or by a focus on a particular region (Lerner, l958; Schramm, l964; Frey, l973).[9] Possibly the wide dissemination of electronic mass media and some telecommunication technologies today, even in less developed countries, has resulted in literacy and education being less important factors now than in the past in the process of economic and political development.
Supporting criticisms described earlier of the political development and develop-ment communication paradigms, the results of this study indicated that none of the social development indicators were strongly related to the Genie Index of income inequality.  Newspaper, radio and television sets per 1,000 and telephone mainlines per capita were significantly but not strongly related to lower levels of the Genie Index of income in-equality in the countries included in this study.  On the other hand, the social develop-ment, mass media and telecommunication time series were more consistently and strong-ly related to GNP per capita.  At least during the period of this study, therefore, the "trickle down" effect predicted to result from strong general economic growth has not occurred to any great extent.
In regard to political development, however, urbanization, the three indicators of mass media development, as well as telephone mainlines per capita were consistently and quite strongly related to a general index of democracy and degree of regulation of the recruitment leaders and participation in the political process.  Moreover, all three mea-sures of democracy were substantially related to GNP per capita over the 32 years of this study.  Unlike the conditional results of previous studies, therefore, the evidence from this one indicates a strong relationship between economic and political development, as measured by these indicators (e.g., see Sirowy & Inkeles, l990).
Some Suggestions for Future Research
        The primary focus of this first analysis of this data set was exploring some of general relationships among development indicators posited in the political development and economic development paradigms, a task that involved a large number of variables.  However, a number of contingent factors affecting the development process need to be explored in future studies.  First and foremost, evidence suggests that the causal relation-ships among development indicators may vary across historical eras (Jayaweera & Amunugama, 1987).  Second, these causal relationships may also be different from one region of the world to the next (e.g.,  Farace, l967).   Third, comparisons should be made between more and less developed countries before the general patterns indicated in this study can be confirmed (e.g., Jayaweera & Amunugama, 1994; Melkote, 1991).
References 
ALKER, H.R. (1966). Causal inference and political analysis. In J. Bernd (Ed.), Mathe-matical applications in political science (pp. 7-43). Dallas, TX: Southern Methodist University Press 
BORNER, S., BRUNETTI, A., & WEDER, B. (1995). Political credibility and economic development.  NY: St. Martin's Press.
BROSIUS, H.B., & KEPPLINGER, H.M. (1990). The agenda-setting function of television news: Static and dynamic views. Communication Research, 17, 183-211. CRONIN, F.J., PARKER, E.B., COLLERAN, E.K., & GOLD, M.A. (1991). Telecommunications infrastructure and economic growth: An analysis of causality. Telecommunications Policy, 15, 529-535.
CUTRIGHT, P. (1963) National political development: Its measurement and social correlates. In N.W. Polsby, R.A. Dentler  & P. A. Smith (Eds.), Politics and social life: An introduction to political behavior (pp. 569-581). Boston, MA: Houghton Mifflin Company.
DEUTSCH, K.W. (l956). Shifts in the balance of communication flows: A problem of measurement in international relations. Public Opinion Quarterly, 20, 143-160.
DORICK, H.S., & WANG, G. (1993). The information society: A retrospective view. Newbury Park, CA: Sage Publications.
ECKSTEIN, H., & GURR, T.R. (1975). Patterns of authority: A structural basis for political inquiry. NY: Wiley-Interscience.
FARACE, R, & DONOHEW, L. (1965). Mass communication in national social systems: A study of 43 variables in 115 countries. Journalism Quarterly, 42, 253-261.
FARACE, R.V. (1966).  A study of mass communication and national development. Journalism Quarterly, 43, 305-313.
FARACE, R.V. (1966). Identifying regional "systems" in national development research. Journalism Quarterly, 43, 753-760.
FEDDERKE, J., & KLITGAARD. (1998). Economic growth and social indicators: An exploratory analysis. Economic Development and Social Change, 46, 455-490.
FREY, F. (l973). Communication and development. In W. Schramm & I.S. Pool (Eds.), Handbook of communication (pp. 337-461). Chicago. Rand McNally College Publishing Company.
GURR, T.R. Persistence and change in political systems: 1800-1971. American Political Science Review, 68, 1482-1504.
HANNAN, M.T., & YOUNG, A.A. (1977). Estimation in panel models: Results on pooling cross-sections and time series. In D.R. Heise (Ed.), Sociological Methodology: 1977 (pp. 52-83). San Francisco, CA: Jossey-Bass Publishers.
HARDY, A.P. (1980). The role of the telephone in economic development. Telecommun-ications policy, 4, 278-286.
HUNTINGTON, S.P. (1968). Political order in changing societies. New Haven, CT: Yale University Press.
ITO, Y. (1980). The "Johoka Shakai" approach to the study of communication in Japan. Keio Review, 1, 13-40.
JAYAWEERA, N., & AMUNUGAMA, S. (Eds.). (1987). Rethinking development communication. Singapore. The Asian Mass Communication Research and information Centre.
JAYAWEERA, N. (1987). Rethinking development communication: A holistic view. In N. Jayaweera & S. Amunugama (Eds.), Rethinking development communication (pp. 76-94). Singapore. The Asian Mass Communication Research and Information Centre. 
LERNER, D. (1958). The passing of traditional society: Modernizing the middle east. New York. The Free Press.
LERNER, D. (1967). " International cooperation and communication in national development. In D. Lerner & W. Schramm (Eds.). Communication and change in developing countries (103-125). Honolulu. East-West Center Press.
LIPSET, S.M. 1959. Some social requisites of democracy. American Political Science Review, 53, 69-105. 
MCCLEARY , R., & HAY, R. (1980). Applied time series analysis for the social scien-ces. Beverly Hills, CA: Sage Publications.
McCRONE, D.J., & CNUDDE, C.F. (1967). Toward a communications theory of democratic political development: A causal model. American Political Science Review, 61, 72-79.
MELKOTE, S.R. (1991). Communication for development in the third world: Theory and practice. Newbury Park, CA: Sage Publications.
MOEMEKA, A.A.(1994). Communicating for development: A new pan-disciplinary perspective. Albany, NY: State University of New York Press.
NIXON, R.B. (1960). Some factors related to freedom in national press systems. Journalism Quarterly, 37, 13-28.
PYE, L.W. (1956). Communication patterns and the problems of representative government in non-western societies. Public Opinion Quarterly, 20, 249-257.
ROGERS, E.M. (1976). Communication and development: Critical perspectives. Newbury Park, CA: Sage Publications.
SATHRE, E. (1976). Indicators of Growth in the past decade. In W. Schramm & D. Lerner (Eds.), Communication and change: The last ten years-and the next (pp.6-10). Honolulu. The University Press of Hawaii.
SCHNORE, L.F. (l961). The Statistical measurement of urbanization and economic development. Land Economics, 37, 229-245.
SCHRAMM, W. (1964). Mass media and national development. Stanford, CA: Stanford University Press.
SCHRAMM, W., & RUGGELS, L. (1967). How mass media systems grow. In D. Lerner & W. Schramm (Eds.), Communication and change in developing countries (pp. 57-75). Honolulu:  East-West Center Press.
SCHRAMM, W., & LERNER, D. (Eds.) (1976).  Communication and change: The last ten years-and the next. Honolulu:  The University Press of Hawaii.
SCHRAMM. W. (1976). Data on communication systems in three developing regions. In W. Schramm & D. Lerner, Communication and change: The last ten years-and the next (pp.6-10). Honolulu. The University Press of Hawaii.
SIROWY, L., & INKELES, (l990). The effects of democracy on economic growth and inequality: A review. Studies in Comparative International Development, 25, 126-157.
SHANNON, L.W. (1958). Is level of development related to capacity for self-govern-ment?  American Journal of Economics and Sociology, 17, 367-381.
SMITH, K.A. (1987). Newspaper coverage and public concern about community issues: A time-series analysis. Journalism Monographs, 101.
TRUMBO, C. (1995). Longitudinal modeling of public issues: An application of the agenda-setting process to the issue of global warming. Journalism Monographs, 152.
WIARDA, H.J. (1989-90). Rethinking political development: A look backward over thirty years, and a look ahead. Studies in Comparative International Development, 24, 65-82.
ZHU, J. (1996).  Comparing the effects of mass media and telecommunications on economic development: A pooled time series analysis. Gazette. 57, 17-28.
Notes         
[1]  See Schramm (l976).  UNESCO's development standards were100 newspapers, 50 radio receivers, 20 television sets and 20 cinema seats per 1,000 people.
[2]  Interestingly, Lipset (l959) did not mention the large differences on each of the mass media indicators between regime types in his description of the results.
[3]  Research on diffusion of innovations, of course, was directly concerned with economic and political development. (Rogers, 1976).  But this study is focused primarily on economic and political development at the macro level.
[4]  Jaggers, Keith, and  Ted  Robert  Gurr.  Polity III: Regime Change and Political Authority,  1800-1994 [Computer file]. 2nd  ICPSR  version. Boulder,  CO:  Keith  Jaggers/College Park, MD: Ted Robert Gurr [producers], 1995.  Ann  Arbor, MI:  Inter-university  Consortium for Political and Social Research [distributor], 1996.                         
[5]  The data set contains these variables: democracy score, autocracy score,  executive recruitment regulation, executive recruitment competition, executive recruitment openness, monocratism, executive constraints, regulation of participation, compet-itiveness of participation, and centralization of state authority.  The democracy and autocracy scores were highly correlated (.90+).  Furthermore, democracy was also highly correlated (.90+) with all other indicators, except regulation of participation (.08) and executive recruitment regulation (.12).
 
[6]  This procedure has been used by Trumbo (l995) and Brosius and Kepplinger (l990). Differencing is the most commonly used technique to whiten time series.  But then the focus of the analysis is changed to examining degree of change. 
[7]  Lag lengths in such analyses need further research.  The lag time between changes in socioeconomic development and subsequent economic and political responses is likely to vary according to a variety of factors.
[8]  Granger analysis can produce stronger causal evidence (See Smith, l987 or Trumbo, 1995) .  According to this technique, a variable X is said to cause Y, if Y is better predicted from the past histories (as indicated by lagged values) of Y and X together than Y alone.  Mutual influence is indicated when both Y and X are better predicted by their past histories and those of each other than alone. But Granger analysis only allows two variables to be compared at a time, which is not practical when testing relationships among a large set of variables.
[9]  It should be noted that Lerner (1958, p. 63) predicted that once 25% urbanization is reached, it "ceases to  play a determinant role because enough people have been relocated in cities to assure the personnel requirements of modern production."