University of Oslo

1. The Problem

The introduction of television has given rise to a number of challenging questions
concerning its relevance to political and cultural life.

First, there is a great deal of interest in the fast growth of television, which in most European countries has taken place well within the life experience of one generation. At present a majority of Europeans have grown up in societies with radio and the press as the dominant mass media, and yet they have shared the experience of television during their mature years. Along with the novelty of the experience came the vital decisions concerning television: they made it fit into family life and corned the nomenclature, which is still an integrative part of its culture.

The search for relevant dimensions in this diffusion process is itself intriguing. How can the remarkably fast spread of television through all segments of the population be explained? Does it have properties vital to existing social and cultural needs that make it fit well into the way of life, or does it change prevalent life patterns and make people adapt to television? Compared to the older mass media, television is more expensive, more difficult to manipulate, and is tied in with household activities in a way which evades the viewers' control, yet its attraction remains undisputed.

In this article we shall be concerned with some of these problems. Within this field, however, we can only begin to probe the dimensions of political information-seeking. What difference does political television make to people of unequal predispositions in their more or less deliberate efforts to learn about politics? What is the relation between the individual's choice of information and his politically relevant acts? In relation to the plethora of information presented to the voters, what is forgotten and what is retained to make up the rational basis of their political choice?

Considering the attraction of television, we might expect to find an increase in
the level of information, but this may take one or more configurations:

1. It may even reduce the contrasts between the informed and the uninformed citizens and make for a more equal distribution of information in the electorate. This function of equalization takes place when a greater proportion of the ignorants become aware of and familiar with politics; it closes the information gap repeatedly observed in the study of information and knowledge in political contexts.

2. It could possibly provide; a better opportunity for the alert voter already well at home in politics to extend his knowledge. Better intellectual resources could prove vital for the use of this highly attractive medium, compared to people in less favorable positions. This may lead to unexpected consequences for the relation between the informed and ths uninformed: information will increase at a faster rate for the privileged extremes and extend the information gap.

The literature on opinion and opinion formation has repeatedly pointed to a correlation between social status and differences in exposure to the mass media, in learning capacities, and in opinionation. In this process the role of the mass media has been specified and in no small degree reduced: their role appears to change in the same degree to which intervening variables have found a permanent place in the relation between the mass media and the individuals. Actually, the potent theorem of the two-step flow of communication frequently blurs the vision of any direct influence of the media. Klapper,¹ in his work on mass communication effects, has not entirely precluded the possibility, but he does not outline any of the conditions under which these residual effects may take place.

Implicit in the demonstration of correlations between life situation and exposure is frequently an assumption of some measure of learning as a result of exposure. This is the case in Key's² study of the relation between public opinion and mass communication, in which it is demonstrated that social class and education determine the chances for exposure to political fare in the mass media. Wade and Schramm³ in a recent article on the role of mass media for the acquisition of knowledge have underlined the importance of education as a very relevant prerequisite for the acceptance of public information.

While assumptions about learning are kept implich in many reports of this kind, some reports do present conclusions about differential rates of retention. Most relevant for this problem is an article by Tichenor, Donohue and Olien⁴ in which the authors test very specific hypotheses about differential growth of knowledge. They found most of their data consistent with a hypothesis stating that in cases where the infusion of mass media in a society increases, the segments of the population with high socioeconomic status tend to acquire knowledge at a faster rate than do the lower status segments. In their concluding remarks they observe that 'the mass media seem to have a function very similar to that of other social institutions: that of reinforcing or increasing existing social inequities'.⁵

Another point concerns the qualities inherent in the different media. There is
a tendency to treat television as simply another member of the family of mass
media without due consideration of the aspects that seem to be unique to television.

McLuhan's philosophical treatment of this difference may be exaggerated, but some of the points deserve serious attention. Pool,⁶ writing during an early phase of the introduction of television, called attention to the 'new dimension in polities' by isolating perceived personality factors of the political candidates conveyed by television.

The data presented here will hopefully throw some light on both the above points: the uniqueness of television in the way in which it attracts voters of every category and its central role in the political information process and in the consequences we may expect for the distribution of political information as a result of these properties.

2. The Data

The period after which television has reached national coverage usually presents difficulties for the study of these and related phenomena. The hunt for uncontaminated' areas may not be successful and even when such experimental fields are found, they are already atypical and often influenced by the ecological effect of high television saturation in the neighboring communities. The phase of introduction probably provides the optimal opportunity for such studies, when large areas can be observed in the pre-television period and where television could be predicted to reach a high coverage within a limited number of years. Simon and Stern⁷ seized the opportunity in their ecological analysis of turnout and television coverage in 98 precincts in the 1952 presidential election in the USA.

We have kept track of the diffusion of television in Norway during the period of introduction and maintain files of media ecology, and we have also studied individual reactions in the period when Norway, as one of the last countries in Europe, entered the television age in the early sixties.

We shall inquire into the problem by using a set of variables drawn from a simple model of political communication, and we shall examine the strategies applied by the voters in their information-seeking activities. We do not have sufficient documentation on the modes of presenting politics and of the extent to which the different media deal with political subject matter. However, it will probably remain undisputed that there is a great deal of difference between the media on both scores. The voters, whether tied to the political parties in loyalties and identifications or indifferent to politics, seem to perceive the options for political information-seeking surprisingly consistently: they know about the political leaning of their newspaper, and at least in the new television owners there is an almost unanimous praise of their medium as a dependable, unbiased source of certain types of political information.

We shall study the information-seeking preferences in a panel of voters according
to their life situation and from different angles: what they report on their
everyday relation to political information and how they behaved during two consecutive

national election campaigns in their efforts to get in touch with politics. We shall proceed to investigate the difference in information-seeking behavior between television owners and non-owners. Then in a preliminary way we shall study some consequences for the distribution of information in the electorate: Which of the two configurations seems to gam in the early phase after the introduction of television?

Data for the study have been collected in a survey of the electorate in the three northern provinces of Norway, first immediately after the national election in 1965 and before television was available in this part of the country, and then again when the respondents were visited after the national election four years later. By that time television coverage in North Norway had reached the national average.

Names were drawn from ths election registers in order to obtain a regional, crosssectional sample representative of the electorate in i;hese provinces. Nine hundred names were drawn to constitute the sample and 719 persons qualified as firstwave respondents. Immediately after the following national election the respondents were visited again with a questionnaire partly identical with the foregoing, but somewhat more extensive. Out of the original 719 respondent, 535 qualified as second-wave respondents. They make up the panel used in this analysis.

The panel technique makss it possible to study interesting phenomena in a simple way. Here we have applied two crude divisions of the sample: one according to the life situation of the respondents as measured with one indicator, the number of years of formal schooling beyond primary school, and the second according to their television status in 1969. The introduction of television in the years between the interviews makes it possible to register the voters' reactions to a very similar set of stimuli, in 1965 without access to television and four years later those who remained non-owners of television and the group of respondents who at that time had acquired sets.

The purpose of the study is multiple. It has been coordinated with the comprehensive studies of the Norwegian population under the election research program in Norway. Naturally we wanted to study the voters' reactions in situations in which the options for politics! information-seeking could be systematically varied. But we also wanted to study the introduction of television and to fit some of the consequences into a broader frame of reference: for the media structure in the shift of preferences of media as sources for information, for the distribution of information and knowledge in :he electorate, and for the effect of different attitudes toward the media.

In these aims we have been only partly successful. We collected detailed information on the rates of exposure during the election campaigns in both years, and we have data on people in their committed roles as citizens: their ties with parties and organizations, as information-seekers in their everyday activities, and their reactions to the mass media themselves, how they are perceived, and on the dimensions along which they are rated in the voters' perception. On the other hand, we have scarcely any data on their personal interactions. We do not know enough about the exchange of political subject matter between the voters. In spite of the

repeated observation that the mass media constitute the main source of up-to-date political information, the reservoir of political currency that rests with the voters themselves, to be re-invented and regenerated in personal relationships, can hardly be overestimated.

The situations in which the surveys took place have been described elsewhere.8
We shall relate, however, some of the more important trends in the diffusion of
television from a modest start in 1960.

By the end of 1959 - one year before the official start of the first television transmitter - 6500 households had acquired television sets and could watch the frequent test transmissions in the central Oslo area. By the end of the next year almost 50,000 households had sets, most of which were concentrated in the eastern centers. Two more years brought most of the densely populated areas within reach of normal transmissions and the first quarter of the households had sets very early in 1964.

By this time all of northern Norway was still untouched by the influence of television. In the beginning of January 1965 only 967 licenses were registered in the province of Nordland and only ten in the northmost provinces of Troms and Finnmark. While 'southerners' by and large had switched to television, people in the north maintained the 'old' listening pattern. The corresponding figures for radio exposure disappeared south of the Nordland border.⁹

Besides television, the options for information-seeking remained much the same for the Norwegian electorate throughout the period of implementation. During this relatively short span of time there were few extensions of the radio service, and newspaper production showed no drastic changes. Movies and weekly magazines took most of the blow, but these media never were primary sources of political information. The design adopted for the development of television was, however, of consequence. There was to be one national program with no opportunities to watch telecasts from other sources except in cases when formal cooperation with Nordic and European counterparts led to the exchange of programs.

However unsatisfactory this appeared to the public, it provided an almost ideal situation in which to study some salient questions concerning television: the consequences of imposing a new media structure on a society already saturated with devices for the exchange of information.

An almost immediate effect took place in media preferences. To new viewers television became the medium. The change in exposure habits in some cases threatened to upset the balance between the media, but to the student of political life there were effects of greater relevance. The audience for political programs during the election campaigns extended from 51 percent in 1957 - the last national election before television - to 81 percent in 1965, when television reached more than six out of ten households.¹⁰ Television also accounts for a considerable increase in the turnout level in the elections following 1959. The trend was not so clear in the eastern urban centers but was evident in the peripheral districts in which the turnout had been low in most of the previous elections. The connection between the two phenomena remains to be demonstrated with data on individual

reactions, but there is considerable evidence in the Norwegian data to make it a
likely assumption that television under certain circumstances is an effective agent
for political mobilization.11

3. The Life Situation

Almost any property or characteristic of the individual contributes to the explanation of his life situation. Position in life cycle, family context, and occupational relations determine a great deal of individual needs and aspirations. Measures of intellectual training are even more indispensable since we normally would include in a description of an individual life situation all terms of his dealings with an outer world: the range of alternatives known to him, the number of resources he can bring to bear on any problem at hand, and the basis of his preferences, i.e. why he finds one strategy more adequate than another.

No single indicator includes all of these aspects. It would put undue stress on the argument if one were to consider the entire individual position in life on the basis of occupational criteria or the amount of money available for consumption in the household. Occupation in a fair way determines the individual activity on a number of scores: leisure time, patterns of consumption, and, even more, the relevant, social contacts. However, the increasingly diffuse middle class, together with the rise of new skills and professions, blurs the once so useful distinctions between different occupational groups. Income for many reasons is becoming less important, though it still indicates a potential for the realization of a number of wants.

While both classifications seem to give a generous score of information, measures of education are more relevant. This variable reflects individual capacities particularly important for information-seeking: a high score; on an education scale indicates a wider range of relevant social contacts, increased capacity for comprehension, and greater alertness to political matters.

In the Norwegian data the indicators are interdependent. The correlation is
strongest between education and occupation, and is weaker in the case of income
and the two former variables (product-moment correlation coefficients):

The sample produces a distribution of some important information-seeking activities by these indicators of life situation as shown in Table I. In this review there are no great differences between the indicators. They seem consistently to make for a higher proportion of the voters to be alert to politics, to be in touch with

mass media, and to have a wider knowledge of politics. Education discriminates
better on most scores, but income naturally seems to be most important when it
comes to the acquisition of television.

In the following analysis we shall depend largely on education as an indicator of
life situation and study the variations in information-seeking strategies within social
groupings defined accordingly.

4. The Sources of Political Information

What are the sources of political information? How do the voters deal with politics in their daily routines, and what are their reactions to the escalated activity during an election campaign? The data give information on both questions, but the points need some elaboration. We have to keep in mind the difference between the citizen as an information-seeker in his everyday activity and the voter in a campaign situation: between a current scanning of available social and media contacts, and a situation in which the voter finds the number of options multiplied and candidates and issues displayed on a large scale. This is essentially a distinction between the reinforcement of generalized insight and the acquisition of specific

knowledge. A great deal of what we know about information-seeking is based on
analyses of campaign situations. But in a way the everyday activity is more significant:
it reflects the more important aspect of political socialization.

Between elections newspapers seem to be the most obvious source for political information. They deal with politics every day. To many people they represent the primary relation to politics and, aside from the: bi-annual formal voting act, their maximal participation. So far the Norwegian press provides considerable opportunity for the citizen in search of political guidance: a random copy of any large newspaper contains from 15 to 20 percent political subject matter even in periods between elections.¹² This figure refers to a wider concept of politics than strictly party activities, but this also makes the relation valid for newspapers who declare themselves to be without political ties. However modest the political information-seeking, the political commitment of the press is generally accepted. Practically all readers are familiar with this feature and, when asked, can correctly identify their newspaper by party label.¹³

Radio and television on the other hand remain detached from party politics in an election year. Allocation of trjinsmission time is done with due consideration to the balance between the parties, to the presentation of salient issues, and in some small way also to popular feedback. Campaigns are well planned and featured in a professional way with, some would argue, inordinate attention to persons in politics. Normally there is substantial pre-announcement and surprisingly much follow-up in the press.

The substance of political information in flux between the voters does not lend itself to a similar description. Utilization of this reservoir of knowledge through organizational channels, in interpersonal relations, and in family small talk is not independent of the use of other media.¹⁴ There seems to be a strong relation between the use of mass media and the rate of exchange of information between persons. However, exactly what is manipulated and the modes of exchange have so far escaped scientific attention. The field is open to investigation in many directions: toward the range and scope of the political vocabulary constituting the currency in the exchange of information, on the personal and contextual conditions under which the givers and takers operate, and on the salient themes at the individual level as compared to what is available for v/indow-shopping in the mass media. At present we are left with simple descriptors which can only disentangle the 'heavy' participants from the bashful listener, the passive contributor from those who do not care at all.

Only a small proportion oi the electorate are actively engaged in political information-seeking. A great majority of the voters have a passive relation to the politics available to them. They may be entrenched in their party loyalty and have no need for further information, or they may simply not be interested. Even so they may not touch upon political information just by mere chance: in addition to the 12 percent who claim editorials and politics to be their first or second reading preference, an additional 17 percent report that they follow 'regularly' the political reports in their newspaper. The remaining two-thirds may well choose other channels,

or their contacts may be more random, depending largely on contextually
defined conditions. Some will of course disregard politics entirely.

Politics in the mass media differs according to the level and scope to which it refers. Local issues together with national and international affairs make up the political fare. This is the essence of the political newspaper: whichever national events constitute the day's headlines, local news is always there to be recorded.

In the voters' perception, this is one of the important distinctions between the media. While newspapers are indispensable sources for information on local politics, television has taken over a great deal of the responsibility as a primary source for national and international affairs. Close to seven out of ten respondents reported newspapers to be their main source of local information and three out of four respondents made radio or television their main choice in matters concerning national or international affairs. The distributions for three questions on this topic are given hi Table 11.

This is a likely outcome since newspapers for all practical purposes remain the only source for local information of this kind. The situation is quite different when the respondents deal with other types of information. Here access to television becomes important. Comparing the media preferences of television owners and non-owners, radio is important to people without television. The overall picture is found in Table 111. Details in the tabulations stress the importance of education in the discriminate use of media.

The relation between the media is the same, but education accounts for important variations. With non-owners education further stresses the importance of radio, and while most television owners beyond primary education still hold television as their preference, an increasing proportion turn to newspapers for information. The data do not permit detailed breakdowns, but the relation also seems to hold for the highest educational levels. To television owners, newspapers become increasingly important with education; to non-owners radio is the obvious alternative.

Surprisingly few people report other persons to be sources of information. Close
to 7 percent say that neighbors, friends, and colleagues are the main source of
information on politics and decisions of local importance. For political information

on other levels the local informant is ruled out. The mass media dominate entirely. This does not dispose of the opinion leader in the political process. The question measures the difference between the different channels to the extent that the respondents perceive a distinction between neutral information and opinions. The mass media may carry information; people convey opinions. Some evidence confirms this assumption: only 2 percent find friends and neighbors to be dependable' or 'unbiased' as informants on political matters.

Even if the introduction of television has not increased the public demand for political information, it has certainly brought more people into contact with it. It is yet another question whether the extension to include visual aspects of political communications has made politics more available. Again this is largely a question of education. The more formal schooling, the greater the proportion of the electorate who find it easy to keep informed. In the sample, 27 percent found it very easy to keep informed about politics, and 34 percent said it was difficult, with the remaining 39 percent in the middle category.

The differences between the television owners and the non-owners are insignificant at the lower levels of education, but television changes the situation for people with more education: twice as many television owners find it easy to keep informed than do non-owners (Figure 2).

We must stress that this does not necessarily mean that they learn more about politics, only that they perceive information to be a more easily accessible good. We shall, however, keep the difference in mind when we turn to the question of differences in knowledge about politics.

The observations are open to interpretations in other directions. They could reflect a predisposition of the press to deal with politics in a local frame of reference, or they could refer to a difference in what the readers perceive to be neutral political information about national and international affairs, and what they expect to be party politics and partisan presentations in the press. Both explanations tie in well with the structure of the Norwegian press and important features in the party system. In both cases the local aspects are significant: the regionally limited circulation of the press and the dominant principle of the party system of organization at the grassroots level.¹⁵

From our point of view the analysis clearly emphasizes the importance of television as a national infrastructure, a vehicle with a strong potential for dissemination of national information. More than radio, and certainly more than the press, it cuts across community boundaries which still frame the daily life of most people, displaying information and symbols with, as we shall see, considerable success.

In daily life, it seems that people have certain expectations of the mass media, to the ways in which they can be utilized for political information-seeking, and they are quite capable of presenting their strategies clearly. Evidently television plays an important role for the spread of certain types of political information. At least the voters seem to think so. The differential use of mass media is more manifest at higher levels of education in daily association with politics.

How do people react in a campaign situation when the communication output
is multiplied and the normative pressure to keep informed is at the highest?

5. The Election Campaigns

The election campaigns in radio and television followed much the same pattern in both years under study. By agreement between the Norwegian broadcasting company and the political parties, transmitting time was distributed among the parties to present their platforms and to answer questions. In 1965 most programs went on the air simultaneously on radio and television.¹⁶ While in 1969 party representatives were interviewed by a panel of trained journalists on television, radio programs were organized so that the party spokesmen could be interviewed according to listener feedback: the representatives could be contacted by telephone. In both cases time was allotted equally among the parties, but radio in 1969 had more time at the disposal of politicians than television did. The parties had their representatives appear on specific programs in both years, and a concluding debate between the party leaders was transmitted simultaneously in both media in both years, as was the report of election results during the night, after the polling stations had closed. Furthermore, radio and television presented panel debates on what were considered to be the most salient political issues.

There is at least one consistent tendency in our data: television seems to bring
more people in touch with politics. Drawing additional information from earlier
surveys in 1957 and 1965 the tendency is clear:

In 1957, the last national election before the introduction of television, 47 percent of the electorate did not follow any of the election programs on radio, which at that time had quite a long tradition in Norway. Five years after television had started regular transmissions, only 14 percent did not listen or watch the campaign on at least one of the media. In our present panel 28 percent did not tune in their radios to the debates in 196:5, and in the next election exposure patterns were much the same as for the national average in 1965. In this Table we cannot account for the difference in political climate during the different elections. The 1957 election was indeed different from the tense situation in 1965, and in 1969 the situation was again more relaxed.¹⁷ However, the difference in media options and the models of presentation still comprise the most persistent variable.

In between the two elections the number of television sets increased from
12,000 to 74,000 in a population of about 450,000 people. In our survey the increase
is reported in terms of household data as in the following Table.

We shall follow our respondents in a number of activities related to campaign activities in the two election campaigns. First we shall relate their exposure in 1965 to their corresponding activity four years later: what difference did television make for the non-exposed non-owner from 1965? What use did he make of the new medium, was he tempted to follow the fight at close range, or did he remain indifferent to the instant confrontations on the screen?

The data are presented in Table VII. The respondents are grouped according to their exposure to election programs in 1965. We have one group saying that they followed the campaign at some time and one group stating that they did not follow the debates. Four years later television does not seem to have made much difference to the people who were already information-seekers in 1965, but for the nonexposed people the difference is striking: 55 percent of the non-owners in 1969 followed the election programs in at least one medium as against 82 percent of those who had become television owners. To most of these additional 27 percent, television made the difference between exposure and non-exposure. They were not interested enough to watch the 1965 campaign, in an election strikingly more exciting than in 1969, but were mobilized to watch when television became available.

The extent to which they followed the campaigns also depends in some measure on their access to television. Therefore we have sorted out the group with the most frequent exposure: those who followed 'most' programs, i.e. more than half the number of programs available. The relation between the four groups is the same: television mobilized the 'light' listeners to a high exposure in 1969. The concluding debate between the party leaders was transmitted simultaneously on radio and television in both years. In this case the relation between the two years offers a direct comparison. The relation still holds: twice as many people were mobilized to the longest and most exhaustive debate in 1969.

Television did not seem to make any difference for meeting attendance and involvement in discussions by the voters. We have no data indicating the frequency of political meetings during the campaigns, but by and large the parties seem to have kept up the traditions of direct confrontations with their voters. Details in the tabulations indicate a small difference in meeting attendance in favor of television owners among the alert voters, but by and large the difference takes place in mass media exposure.

Going back to 1957, radio campaigns seemed to be an uncertain strategy. Every second voter listened to the central debates at some time, but very few gave it more than a minimum of attention. Heavy exposure was clearly a marginal activity. With television, heavy exposure became the normal activity, and about as many people watched heavily as listened at all in the heyday of radio. But even if television has mobilized a greater proportion of the voters to direct confrontation with politics, the distribution of exposure still is unequal.

In Table VIII we have divided the panel according to television ownership in
1969 and have studied the exposure activities of these groups for both years. The
proportion of the groups who reported an exposure activity has been tabulated according

to education. This give us eight groups to compare: television owners and
non-owners in 1969 and the same group of respondents by the corresponding activity
as 'potential' television owners and non-owners in 1965.

The overall picture is presented in the first section of Table VIII. All respondents who reported being in touch with the election campaign at some time have been included here. If the increase in exposure in the group of non-owners can be taken to be the result of factors other than television, the net effect could be measured as the difference in increase between the two groups. This measure is by far the highest in the 'high' education group. Exposure is consistently higher in the 'owner' group, on all levels of education for all the activities we have measured.

The next section contains an intensity dimension. Here we have included the people who reported following more than half of the available programs on radio, on television, or on both media. By the same measure, the net effect of television along this dimension is found on the lowest level of education in the group of television owners. This is also the case in the last section, which includes exposure to the program in which the party leaders presented their parties.

The trend seems to be an overall gam in exposure to the election campaign in the group of television owners at the highest level of education, while the 'low' education group of television owners gained in the extent to which they followed the election campaign. On all scores, however, the 'high' education group of the television owners had the highest proportion of exposed respondents.

The differences exist only in relation to the mass media. The corresponding relations
for activities involving more personal commitments do not reveal any similar
variations.18

So far we have only been concerned with exposure. We have tried to disentangle
how people under different conditions seek information on politics. The predominant

role of television provided the logical point of departure for the study of differential information-seeking: In what direction and to what extent does television have an effect on the changes of getting in touch with politics? The overall tendency in our data is consistent: television seems to be an effective agent for the mobilization of citizens in election campaigns. Accounting for a more discriminate use of mass media, further education reinforces this trend. This is significant for the problem at hand: even if television increases the chances for exposure for people at lower levels of education, chances for the more educated are even better. To some extent television makes up for differences in education, but still a larger proportion of the more educated make more use of their option.

How is this relation exploited at different levels of education?

6. The Distribution of Information

The survey provides some measures of political information on constituency candidates and national party leaders. In the latter case the information reported on was typically the kind to be found on television. National party leaders repeatedly appeared on the screen, participated in the debates, and eventually had one entire evening at their disposal on television and radio. In very few cases, if at all, did constituency candidates appear on the screen, at least not in that capacity. This kind of information was almost exclusively left to the local channels of information: the press and local canvassing activities.

The respondents were asked to identify by name any number of candidates from the constituency list on which they reported to have voted in the actual election ¹⁹ Each respondent was given a score according to the number of candidates he had correctly identified. In scoring the amount of national information, the procedure was more complicated. Respondents were shown pictures of the party leaders and were asked to identify each leader by party and name. Each 'bit' of correct information was scored and the resulting 15-point scale reduced to a 4-point scale on which only one mistake was allowed in order to obtain the highest score.

We shall be concerned with the latter type of information, but the distributions of both measures are presented in Tables IX and X. On the 'local information' score there are no significant changes in the distribution of information over the years, whether in the non-owner or the owner group. There is a general increase in the information level from 1965 to 1969: more people know more names, whether television owner or not. In this case we can hardly exclude the possibility of an accumulated effect. This panel effect may to some extent explain part of the increase. In the next case, however, the changes are significant. The relation between 'high' and 'low' scorers are actually entirely reversed with the 1969 television owners. This is not surprising since visual presentation is unique to television. If an effect should take place it would be precisely on these scores.

Keeping the exposure data in mind we shall turn to the initial question and
study the changes in informal ion level in the groups at the different levels of
education: what difference does television make?

In order to test this we have calculated the mean scores of the scale for each level of education for television owners and non-owners. The result is presented in Table XL The score increases both ways: toward a higher level of education and between the years. The owners consistently know more than the non-owners, which reflects the difference in status and in interest in politics. The point here is that there is a differential increase and that there is a marked increase in the distance between the groups after the introduction of television, on both levels of education.

We cannot consider this effect to be due entirely to television. We can, however, make a rough estimate of the effect if we consider the increase in information level of the group of non-owners to be due to influences other than television and to use this as a controlling factor for the group of owners. This can hardly be more than a rough test because of the differences between the owners and the nonowners, but drawing the comparison at different levels of education may indicate the most probable group in which to look for an effect.

This has been done in Table XI, which indicates a more marked effect of television
on the information score at the lower level of education. This is the group
that gained most.

The interpretation of this result stresses the equalizing function of television.
The new medium seems to close the information gap between the groups in different
life situations.

More serious attention should be given to the educational extremes, however. At the uppermost level of education the trends for differential acquisition of knowledge seem to be different, but the details cannot be tested within this sample with too few cases at this level within the group of non-owners. We may expect to find a relation here closer to the second configuration initially referred to.

7. Summary

In this report we have been concerned with some of the consequences of the introduction of television. Actually, it has only been possible to isolate some areas in which to locate such effects: in the mobilization of the voters to election campaigns and in the resulting distribution of political information. The data have been collected in the region of Norway that includes the greatest proportion of peripheral communities, but in spite of this ecology the acquisition of television increased at a faster rate here than in the south central parts.

The points made in the analysis emphasize the importance of television as a link to the central society and to central decision-making bodies. In an earlier analysis we have found the effects of television most marked in the peripheral areas,²⁰ and in this study we have indicated an effect in the social periphery, on the less educated, the less informed.

The increase of the information level is not limited to television owners. It can hardly be explained by a panel effect alone, and further explanations may well be looked for in individual motivations and alternative information-seeking habits as well as in altered conditions for the traditional media. One promising aspect is a between-level analysis: How is the information relayed in pre- and post-television communities or in communities with different degrees of television accessibility? What is the ecological effect of television?

This effect may be mediated in a number of ways, some of which are more
significant than others, such as occasional viewing by non-owners, in the reporting
by local newspapers of the television debates - which has been frequently done

- and in the increased exchange of information by personal communication between the two groups. In the first case we have to deal with a direct effect of communication; in the latter cases we can draw on well-established theories of the two-step flow of communication in order to explain how this effect takes place.

The main findings have been established in the introductory phase of television. It may well be that the effect is limited to this phase, but the society in which we can put this to a test will be an entirely different one; there will be a general increase in the level of education and a rapidly decreasing group of families without television.

---

NOTES

1. Joseph T. Klapper, The Effects of Mass Communication, Glencoe, 111.: The Free Press, 1960.

2. V. O. Key, Public Opinion and American Democracy, New York: Knopf, 1961.

3. S. Wade and W. Schramm, 'The Mass Media as Source of Public Affairs, Science, and Health Knowledge', Public Opinion Quarterly 33 (1969).

4. P. J. Tichenor, G. A. Donohue and C. N. Olien, 'Mass Media Flow and the Differential Growth in Knowledge', Public Opinion Quarterly 34 (1970).

5. Ibid.

6. Ithiel de Sola Pool, 'TV: A New Dimension in Polities', in Eugene Burdick and Arthur J. Brodbeck (eds.), American Voting Behavior, New York: The Free Press, 1959.

7. H. A. Simon and F. Stern, 'The Effect of Television upon Voting Behavior in lowa in the 1952 Presidential Election', American Political Science Review 49 (1955).

8. Stein Rokkan and Henry Valen, 'The Election to the Norwegian Storting in September 1969', Scandinavian Political Studies, Vol. 5, Oslo: Universitetsforlaget, 1970; Henry Valen and Willy Martinussen, Vdgerne og politiske frontlinjer (Voters and Political Front Lines), Oslo: Gyldendal Norsk Forlag, 1972.

9. Gallup ratings, Fall 1964.

10. An annual license fee is necessary to keep a television set. Data from license records, Norwegian Broadcasting Comr. any.

11. Cf. Henry Valen and Per Torsvik, 'Økningen i valgdeltakelsen ved kommunevalget i 1963 og stortingsvalget i 1965' (The increase in the vote mobilization at the 1963 communal and the 1965 national elections hi Norway, with English summary), Tidsskrift for samfunnsforskning 8, 2/3 (1967). The ecological analysis indicated that in the 1963 election television and politization contributed strongest in the periphery to a mobilization of the voters to the polls. In the 1965 election politization had no effect, nor had television in the central parts of the countiy. In the peripherial parts the effect of this variable was substantial. On individual reactions in 1965, see Per Torsvik, 'Valgkampen i fjernsyn og radio i 1965', Tidsskrift for samfunnsforskning 8, 2/3 (1967).

12. Data from unpublished reporl on the distribution of the content in Norwegian newspapers, part of which has been used in N. Ramsøy (ed.), Det norske samfunn (Norwegian Society), Oslo: Gyldendal Norsk Forlag, 1968.

13. Questions according to this point were asked in the surveys.

14. E. Katz and P. J. Lazarsfeld, Personal Influence, Glencoe, 111.: The Free Press, 1955.

15. Henry Valen and Daniel Katz, Political Parties in Norway, Oslo: Universitetsforlaget, 1964, Chap. 3.

16. Itinerary for this election service 1965 is found in Torsvik, op. cit.

17. The point is made by Valen and Martinussen, op. cit.

18. Tables accordingly are not included here.

19. The Norwegian election system prescribes proportional representation from each of the 20 provinces (fylker); party list; from the three northern provinces included 10—20 names.

20. Cf. note 11.