1. Introduction

In this article we shall depict the rise and expected fall of the party press in Norway. Our description will have two components, one chronological and one systematic. The systematic approach rests on one basic postulate, namely that the emergence of a party press and its vitality in Norwegian, society can be explained from factors of social and political development. More specifically we shall focus on mechanisms in the newspaper market which are interrelated with these factors of change, and ask such questions as: how did economic growth combine with expansions in the newspaper industry and the rise of parliamentarism to change many of the prevailing conditions in publishing? What stabilized the alliance between the press and the parties, and what are the forces that cause strain in the transactions between parties and the press today? Will they eventually lead to a break-down of the old tradition of party-committed newspapers? The chronological description becomes a necessary supplement to this approach, because the timing of past events must be understood before a theory can be applied for broader generalizations. This does not imply, however, that we intend to present a coherent line of interrelated events from 1763 - when the Norwegian press was founded - up to the present. Rather, we shall emphasize two phases in the historical trend: the period of alignment of partisan politics in the press, which coincides with the formation of the party system, and a period of depolitization which covers the post-war era. The sixty years that separate these periods will only be touched on briefly when we turn our attention to the next problem: why have stable party commitments in the press not succumbed with the transition from a two-party to a multi-party system?

---

* The research on which this article is based is part of a broader project on the Norwegian press being conducted by the Institute for Social Research in Oslo. The project is financed by the Norwegian Research Council and the University of Oslo. Grants have also been provided by the Michelsen Fund in Bergen and by the Aschehoug Fund for social research in Oslo. I am indebted to my colleague for many years, Per Torsvik, for his critical reading of an earlier draft, to Einar Ostgaard, who shares my long-standing academic interest in the press, and to Ulf Torgersen for his many valuable comments during our numerous discussions on Norwegian history and related subjects. I also thank my assistants Bjornar Ottesen and Dag Nicolaisen for their spirit and patience shown in an often tedious task. My informants in the press are too many to be named in this connection, but their openness and willingness to give of their knowledge have been a great encouragement to me.

* The research on which this article is based is part of a broader project on the Norwegian press being conducted by the Institute for Social Research in Oslo. The project is financed by the Norwegian Research Council and the University of Oslo. Grants have also been provided by the Michelsen Fund in Bergen and by the Aschehoug Fund for social research in Oslo. I am indebted to my colleague for many years, Per Torsvik, for his critical reading of an earlier draft, to Einar Ostgaard, who shares my long-standing academic interest in the press, and to Ulf Torgersen for his many valuable comments during our numerous discussions on Norwegian history and related subjects. I also thank my assistants Bjornar Ottesen and Dag Nicolaisen for their spirit and patience shown in an often tedious task. My informants in the press are too many to be named in this connection, but their openness and willingness to give of their knowledge have been a great encouragement to me.

2. The Initial Stages of Growth in the Press

2.1. The Norwegian press had a modest start in 1763 when the first newspaper was permanently established in Oslo. Even before this time, however, foreign newspapers were circulated among the urban establishment. The late arrival of a national press may be ascribed to the fact that until 1814 Norway was the submissive partner in the union with Denmark. The king of Denmark-Norway controlled the trades and it was clearly his intention not to inititate new and independent institutions. In fact, the first Norwegian newspaper was started by evading the royal prerogative of publishing, for which permission was given solely to Danish citizens. Neither were the economic conditions present for an independent Norwegian press throughout most of the eighteen century. Norway was a country on the periphery of Europe, with no natural resources, a limited population and few urban centres.

Eventually geography would become a factor of growth in the Norwegian press, but before the technical means for cheap and effective publishing were available, the topography of the country was a barrier to expansion. Conditions for an integrated economy came with the advance of a coherent network of communications, but they were not present before the middle of the nineteenth century. Independence in 1814 brought a change in the institutional framework. For example, freedom of the press was made one of the main principles in the constitution. However, independence had hardly any impact on the cultural life of the country until some decades had elapsed and more efficient socio-economic infrastructures had been erected to give the vital core of institutions an autonomous basis in Norwegian society.

That part of a modern culture which springs from the press was even more dependent on this development - first to become distinctly Norwegian in content and then to expand throughout the country. As long as administrative regulations and the general lack of resources were a barrier to the permanence of enterprises, most of the attempts were bound to fail. Expansions came only when and where the ground was prepared.

For nearly one century of the Norwegian press, newspapers remained a phenomenon confined to the larger cities. In part this was due to their price, which was prohibitive except for the upper class. It was also a result of the restricted finances of publishers, which did not allow for much independent journalism even when the laws permitted it. Most of the local news was given in advertisements or through articles written by readers. Ordinary news was mainly translated from foreign newspapers. Political journalism or reviews of domestic politics were even more of a rarity in the ordinary press before the 1830s, though they appeared in journals. Economic conditions for a press that could awaken the consciousness of the middle class - the founding of journalism on a social basis - did not develop until the late 1860s, and not until the last two decades of the nineteenth century did the economy of newspapers improve enough to make journalism a profession by which a man might earn a living.

As Figure 1 reveals, expansion in the industry was slow and modest for most of the
first hundred years. From 1763 to 1850 the net increase was only 42 enterprises,

which means that most of the cities had one local newspaper. But outside the capital competition was nearly non-existent. What is surprising about the development of the Norwegian press, however, is not its scaricity until the 1830s, but the amazing speed by which it caught up with and even advanced the economic prosperity that came from the 1840s onwards. This expansion continued unbroken on all levels well into the twentieth century, and by its advance the newspapers became the possession of every household throughout the country even in the remote periphery.

2.2. It took fifty years to establish the first ten newspapers on a more permanent basis, 18 years to add the next ten and then 25 years to redouble the amount to 40 newspapers by 1847. This is net increase basedon number of newspapers operating at various points of time. Activity in the industry was much greater than these figures indicate. Net increase between 1830 and 1840 was 16 newspapers, while about 50 newspapers were started in this decade.¹ In the history of the industry the 1860s may be designated the point of "take-off": from then on the number of newspapers rose more rapidly than the population at large and even the population in urban centres. The rate of growth was higher than any comparable trends and was seemingly unaffected by the economic crises which periodically occurred.

The sudden rise in activity which set in with the sixties seems to indicate that the market had for some time been ripe for newspapers. It was not a lack of initiatives, however, that had so far prevented the industry from expanding. The entrepeneurial spirit of publishers was already awake, as the scattered data on turnover clearly show. Rather, much of the initial growth may be ascribed to a series of technical innovations between the fourties and the seventies.

Until about 1840 newspapers had been produced on a wooden letterpress that was slow and cumbersome to operate. In 1840 the high-speed press was introduced in Norway, and more than tripled production capacity. About the same time improved technology brought prices of newsprint down to a quarter of the previous level. During the 1860s paper made of pulp reduced prices even more, and the 1870s saw the development of a Norwegian paper and pulp industry.

These more favourable conditions were clearly reflected by the rising number of printing houses. By 1830 thirteen were in operation, five of which were situated in Oslo. Within ten years fourteen new printing houses were in operation outside the capital, and during the 1840s another fourteen cities got their first printing house. Roughly thirty printing houses were active outside the four largest cities in the fifties, as against two in 1830. (Fiskaa 1935).

This proliferation, not only in the number of printing houses, but also in their capacity, was not met by a concurrent rise in the demand for printed matter. Technology went ahead of business and prompted many printers to change their terms of trade and become publishers. Traditional book-production was not enough to sustain the ordinary work shop. To utilize the idle hours many printers started journals or newspapers of their own. This became the pattern for the establishment of newspapers well into the 1890s.

3. The Structure of Growth

3.1. We shall now present a statistical review of the Norwegian press in the last century, basedon a list of names of journals and newspapers compiled by the postal service since 1862.² The available data give us two main criteria by which to measure growth in the press: a) the number of newspaper units and their frequency of publication, and b) the spread of places of publication according to their population or 'size'.³ Some meagre substitutes for circulation figures will be discussed at the end of this section.

3.2. The first stage of growth from 1830 was mainly an outward thrust to capture the new markets of the districts with above average population beyond the urban centres. By 1850 six of the cities had two or more newspapers compared to 23 towns with only one.⁴ The period 1830-50 may be termed an initial stage of growth marked by geographical expansion rather than local competition, except for the central areas. The growth between the 1860s and World War II covers all aspects of expansion: number of newspapers, places and competition, editions per week and weight of annual output, and finally circulation. By the review given in Table 1 we may follow some of the rates of growth more closely.

Most salient is the continuity of growth on all three levels that remained almost unbroken until 1939. With regard to number of newspapers the peak was reached in 1930, while editions and places continued to grow until 1939. A reversed pattern can be observed for periods of rapid growth: number of newspapers expands mostly from 1880 to 1911 while editions and places both have a lag of about a decade concerning rapid growth. This seems to indicate another pattern in the stages of growth: rising competition and activity in central areas overflows into a geographical expansion

which is followed by increased competition in the geographically intermediate
area, and this in turn triggers off publishing activity in even more remote places
until an optimal level is reached between 1920 and 1939.

From Table 2 we may follow a parallel growth, both in monopoly towns and in competitive towns. As the aggregate data shows, competition clearly leads the trend both in absolute numbers and in intensity as measured by number of competing newspapers in each place. If we consider only the ratio of competitive versus monopoly towns, the most dramatic increase in competition occurs during the three decades 1870-1900. This holds true also for the intensity of competition: in 1870 there were 3 towns with more than two local papers, compared to 7 in 1880 and 15 in 1890. This period coincides with some of the most far-reaching changes in Norwegian politics.

The geographical conditions of the country are among the most important factors behind the expanding Norwegian press. The community spirit is clearly one of the most obvious entities on which to base a product differentiation, another being politics. From Table 3 the picture becomes more clear: geographical expansion reaches far beyond urban centres. The number of publishing places becomes stable towards 1920, and has since reduced only slightly compared to the number of newspapers (79) that have closed down between 1939 and 1965. Turnover is very low (Table 3). Compared with the changing urban structure, the geographical structure of the press expanded along with urbanization up to about 1911. From then on urbanization alone seems to be the most important factor. (Table 4).

Along with this process we shall expect first a differentiation and then a growing centralization of the communication network which goes far beyond the developing urbanization, and to which the press has in large measure contributed. While activity in the centre has triggered off activity in the periphery, the size of towns eventually set a natural limit to what was possible in the way of additions to the geographical structure of the press. Instead improved technology filled the remaining

gaps. Better transportation facilities and printing methods allowed for larger circulations. This development required larger and more effective units in the press and eventually contributed to the formation of larger units of interaction within the districts.

3.3. Until now we have tried to depict how the Norwegian newspaper industry grew in the period of a century, leaving out one of the most important effects of growth: the media explosion within the population. This will only be indicated here. We may nevertheless give a fair impression of the rate at which this explosion occurred, even when circulation data are lacking.

The annual statistics of the postal service from 1876 onwards give as standard information the number of copies mailed. The major part of these probably consisted of journals, which generally go through the mail, while newspapers are distributed to a substantial degree, by the publishers themselves. The volume of mailed copies gives a good estimate of consumption in less urbanized districts. In relation to total amount consumed, the curve in Figure 2 should start and end at a higher level, while its shape — the rates of growth and decline - may correlate better to the assumed total picture.5

Through newspaper statistics are in a bad shape, one could safely disregard the possibility of downward errors. The total consumption never went below the number of copies accounted for. Furthermore, if we say that this minimum level correlates with the maximum level of consumption assumed at various intervals, the rate of growth can be compared to other indicies. This is been done in Table 5. Consumption of the printed media started at a considerably lower level than figures for private consumption and the size of the working-age population in the 1870s, but

grew faster, except for twenty years of economic depressions and stagnation between 1918 and 1938. Subsequently it went ahead of both private consumption and the expansion in working-age population. During the 1950s the rate of growth in consumption of printed media began to slow down compared to private consumption. This coincides with the introduction of television.

This development gives us an important clue to the understanding of the social

process behind expansions in the newspaiper industry. In its early stage the mass media are clearly restricted by socio-economic factors, but as the general infrastructure makes for a point of take-off, the media explosion becomes a self-propelled process until its optimal level is found. In the meantime the press has become part of the economic system and a part of the social infrastructure itself. How the press became part of the political system will be considered in the following section.

4. Two Stages in the Formation of a Party Press

4.1. We may now return to our general postulate that the rise of a party press can
be explained from socio-economic developments and more particularly form mechanisms
in the market of newspapers.

In the present structure we may distinguish four party presses: the conservative, the liberal, the labour and the agrarian. We have ranked them according to the time of their introduction, which also corresponds roughly to their size as measured by present circulation. While the conservative and liberal press sprang from the same historical events during the 1870s and 1880s, and thus can be explained by the same configuration of factors, an explanation of the later offshoots from this original core of the party press must include the foregoing stage.

In this section we shall focus on the initial stage of political mobilization in the press and the formation of the Labour press. The two stages will be illustrations of an economic theory of political involvement in the press, which will be further discussed in the two succeeding sections.

4.2. When a more efficient socio-economic infrastructure was erected towards the 1870s, and the technical means of the newspaper industry were considerably improved to utilize the potentialities of a widely literate population, a series of conflicts arose, which in less than twenty years changed the fundamentals of Norwegian politics.

When Norway became independent in 1814 the bureaucracy remained in power until 1883 when the combined forces of the farmers and the intelligentsia successfully broke the system. The opposition that stemmed from the farmers originally had its roots in traditional mistrust of central authorities (Rokkan 1967). From passivity this mistrust was transformed into an active opposition during the 1860s. This was in no small measure a result of economic growth. Farming gradually became part of the money market as its prosperity became dependent on other parts of the national economy, i.e. with the extension of communications and the credit system. This was largely the domain of government. The farmers then set out, first to seize a firm control over the Storting (parliament), and next to control the government (Kaartvedt 1956). Their first goal was easily reached through the constitutional clause by which two-thirds of the representatives were reserved for the rural districts. The intellectuals made this clash of interests a constitutional issue, clearly intending, in the last analysis, to change the political order. In this effort they were aided by the growing urban middle class.

A systematic opposition in the Storting came early in the 1870s (Kaartvedt 1956).

Resistance from the government became stubborn, however, and it did not give in before the opposition had won two mobilizing elections where the principle of parliamentarism was made the dominant issue. From their victory the Liberals won complete control in the Storting, and were able to impeach the government before a constitutional court (Riksrett) in 1883. The government was accused of using its veto for the third time on a constitutional reform which gave their members access to the proceedings of the Storting. The verdict was predictable and it set down the principle of governmental responsibility to the Storting. Finally, the King took his government from the Liberal majority in 1884.

From the battles two parties had emerged on the causus level. The Liberals immediately began to organize the country politically and were closely followed by the Conservatives. Right from the beginning of party politics electoral support showed a distinct pattern which is still discernible today (Rokkan and Valen 1964). The Liberals had their strongholds in the countryside, especially in the Western and Northern regions, while the conservative support was mainly found in cities, concentrated especially in the Eastern and Central regions. This may help to explain why the liberals took an early lead in number of local organizations (Torgersen 1967), while the conservatives won support from the majority of newspapers.

4.3. Even before the establishment of formal parties, newspapers had sided with the opposing factions of the Storting. "Ten years ago", wrote a Liberal commentator in 1883, "the amount of politically colourless newspapers was considerable, but today their numbers are rapidly diminishing".⁶ The press emerged from the political battleground divided along party lines, and this has remained the strongest characteristic of the Norwegian press up to the present time. In fact, as we shall scon discover, by 1885, at the first election with formally organized parties, only 7 out of 99 newspapers were uncommitted to the parties, while another 4 were moderates, somewhere between the Liberals and Conservatives.

Besides the astounding extent to which party politics invaded the press and stratified local markets, the most surprising fact at this stage is the speed at which this occurred. It is obviously that political conflicts precede a politically conscious press in its early stage of growth, and that parties must exist before a party press can emerge. But it is by no means as obvious that involvement in politics automatically produces party commitments in the press. Nevertheless, in the case of Norway, it seems that the evolution of factions in politics and the development of stable party commitments in the press became one and the same process. We shall explain this process only in part as the result of rising tension within the electorate, which made it hard for the newspapers to ignore what was going on. The other part of the explanation must be derived from developments in the newspaper industry itself, which tempted newspapers to become participants in the conflict.

Many other lines of explanation may contribute to clarify this problem, of course, and we shall only mention briefly some alternative factors that may have been working. At the level of elite structure we may return to the factor of small size of the Norwegian society. The rapidly developing Norwegian economy at this

stage of history led to a scarcity of competent personnel to man elite positions outside the administration and politics. As was mentioned, journalism did not become a profession until the last two decades of the nineteenth century. Consequently, the role of editor was often interchanged with other positions of the elites, which eventually made it easier for the feuding factions of the political elite to drive their points of view directly through to the columns of newspapers. On its own side the press was not sufficiently well-founded in society to become an independent institution and thus to pretend to be a "fourth branch of government".

This direction of influence, however, is only likely as long as it does not run counter to the interests of publishers, which is equivalent to saying that market conditions must be present that support a strategy of political involvement. As may have been guessed from the long-term trends of the Norwegian press, these conditions were clearly present by the 1880s. An essential trait of the market was competition. Competition increased both in extent and intensity from the 1860s, when about a quarter of publishing towns had competing newspapers, to cover 71 per cent of places of publication by 1890. As a result of both competition and improved technology, the printed media became available to new segments of the population. The newspaper-reading public became socially differentiated as prices of newspapers declined sharply, which is brought out by Table 6. Perhaps the most interesting detail to come out of this Table is the sharp decline in subscription rates that occurred precisely in this decade of electoral mobilization. It coincided with a more general pricefall during the 1880s, but it clearly overtook it.

This situation must in turn have brought the nessecity of market strategies to the fore. Competition by itself calls for a certain extent of product differentiation, and now this differentiation was given the qualities of social stratification through the potentialities of an expanding public. Newspapers had to decide whether to hold on to an established audience or to expand, and, if the latter, in which direction to expand with what consequences for their present status. The mounting political tension combined with the prospect of increased sales to point out some courses of action. Competition made the active choice of a target group a nessecity for publishers, while electoral mobilization suggested that it could be found within the voter-market.

Newspapers which were established before the age of party politics, and which had survived the obstacles of an early period in the Norwegian press, may have found it hazardous to gamble on an unexplored market among the middle class. This strategy could boomerang by causing defections from the ranks of present readers, especially when political tension was high and tended to split the interests of the more well-to-do and the middle class. It is clear from the electoral history of Norway that during the phase of party formation the farmers and the lowermiddle to working classes were predominantly liberal, while the upper-middle to upper classes in the cities were mainly conservative (Rokkan and Valen 1964). If successful or able to survive, the early establishments had settled for the richer part of the community market, and thus relied on a basically conservative audience by the 1880s.

Newcomers in the market, by the logic of competition, had to differ in their appeal. When the economic and technical means for expansion of the audiencemarket were ready, newcomers were also given the middle class as their public, not yet aligned to newspapers. This was probably the more inviting alternative. Hence a strategy of expansion and one of political involvement easily combined. The historian of the largest provincial newspaper today qualifies this assertion when he remarks that the awakening political consciousness of the middle class towards the 1880s became a journalistic asset used to expand circulation (Fasting 1964).

4.4. It seems, though, that the succession of establishments in local markets and the constraints and incentives present during the early period of operation for each newspaper are important clues in understanding their performance when political conflicts arose. Therefore one sound hypothesis is that competition produces commitments which are concomitant to predominant qualities in that part of the community market which is open to expansion at the period of establishment.

Newspapers will judge the possibilities of expansion within certain segments of the population by the probability of their delivering readers. The choice will fall first on those where sufficient support for continuous operation is most readily found. By the law of competition newcomers will tend to chose a different target group from those already covered by preceding establishments, and commitments will then be selected according to the modal attitude of the prospective audience.

Chronologically we shall expect to find two distinct steps in the mobilization of newspapers in local markets, first one from neutrality to commitment, then one of successive differentiations on the basis of the first commitment which appears. Stable commitments are likely to develop first on the larger and more competitive markets. Given the political situation at the turn of the 1870s, however, partisanship in the press may have advanced competition.

At the first step, interactions between competitors will be decided on the basis of their audience or target groups given by succeeding establishments and the expanding universe of newspaper readers. First establishments will probably stay out of the conflict as long as possible, and the first challenger to neutralism will be found among the second or later establishments. When the first step is passed, newcomers at the second step will tend to take on their commitments primarily on the basis of given positions. By this process commitments will be varied until all positions are occupied, and provided the market allows for the sufficient number of newspapers, the cycle will be repeated. Modifications to this pattern may be expected in cases when a majority of voters provides sufficient support for two newspapers with the same commitment.

Consequently, our hypothesis for the first stage of political mobilization in the
press will be that the order of establishment and the number of competitors at local
markets will produce commitments in newspapers along the following lines:

H l First establishments - or newspapers started without local competition - will
have conservative inclinations.

That is, first establishments may be uncommitted at any given time, bu

when they pledge one of the parties their loyality it will be the Conservative.
H2 Consecutive establishments will tend to differ in their commitments from thos

that preceded them

That is, second establishments will primarily develop Liberal commitments

provided that the first establishments have not remained neutral for a sustainet period. In such cases they will act as first establishments and conform to H l By the same token, third establishments will mainly come out Conservative fourth Liberal and so on.

H 3 Deviations from these rules will mainly be found where product differentiatioi

— j. ---S -- I f —-IS--
does not provide newspapers with a target group or an audience of sufficien
size to support continuous operation.

That is, at any point in the succession of establishments beyond the first one

product differentiation may be biased in the direction of a current majority ir the local market. Likewise, a bias beyond the second establishment can go ir the direction of a current minority, provided that two or more of the earliei establishments have adopted the same commitment.

Deviations from H l will all be cases of first establishments which have Libera]
commitments, while first establishments remaining neutral will also be neutral ir.

this context. Strictly speaking, deviations from H 2 are confined to cases of succeeding establishments which have adopted the same kind of commitment, but as our commentaries indicate, H 1 must be seen as prior to H 2. Together they will not allow for a second establishment Liberal newspaper when the first establishment remains neutral. With these reservations we can exemplify conforming and deviating cases of H 1 and H 2 in Figure 3.

4.5. To test these hypotheses we decided to classify the commitments of newspapers in 1885 by means of content analysis. Few and very insufficient sources exist on the political press in this period, but fortunately the collection of newspapers in the University Library of Oslo proved complete for 1885.⁷

In 1885 the first election with regular parties was arranged. The campaign was unusually strong with regard to the vocabulary used and compared to more recent campaigns. This made it more easy to classify the newspapers on the basis of a few selected issues. The procedure was as follows: from the date of the election the coder went back five separate issues and read all articles on the forthcoming election. All relevant articles thus found were coded by their bias, neutral or leaning in the direction of one of the parties. If no slanted articles were found in one of the five issues, the period of coding was prolonged up to a limit of ten issues.

If no statements were given in favour of the parties for ten consecutive issues, the newspaper was classified neutral or uncommitted. In all but seven cases there was no doubt as to classification of commitments, since the bias pointed only one way. Of the seven residual cases, four had unequivocal statements in support of one party only, but the number of issues containing such statements did not add up to five. Only one of the four was classified uncommitted, because the single statement found appeared in a letter to the editor. The three remaining cases (out of the seven) were classified as uncommitted. We found clear statements in support of the parties in all of them, but the slanted articles were mainly identified as having been written by readers. Moreover, as the articles were evenly distributed in support of both parties, editors were evidently sitting on the fence.

Although useful from the point of view of coding, our choice of the 1885 election ruled out the possibility of finding traces of the first step in mobilization in local markets. As is revealed by Table 7, only seven out of 99 newspapers remained uncommitted in this election. Even the majority of the newspapers operating as monopolies became committed, altogether 13 out of 15 cases. The political events prior to this election made it difficult for editors to keep silent.

Ranking of establishments were mainly made on the basis of year of establishment, the oldest newspaper in a local market in 1885 being the first establishment and so on. A basic idea for our purposes is that first establishments shall either have been started or have been operating for a sustained period without competitors. In two cases the younger of two competing newspapers was ranked as first establishment, the reason being that the paper which was chronologically speaking, the first establishment, had merged with the younger of the two operating in 1885, while the paper which was older in 1885 had been started in competition.

We have calculated support among the electorate for the conservatives on the basis of valid votes in constituencies in which newspapers were published, and on the basis of valid votes in the surrounding county, including the cities. The two alternative figures are given because we have no means of finding out what were the circulation districts for newspapers in 1885. The cities probably made up the most significant markets, while the surrounding county was an alternative market for expansion.₈

4.6. We shall now look at the results and see how much of the structure of the party press in 1885 is accounted for by our three hypotheses. In Table 7 we find H l confirmed in 29 out of 44 possible cases and disconfirmed in 10 cases, while 5 cases are neutral. That is, out of the 44 places of publication which we have analysed, the first establishments have committed themselves at 39 places. At 29 or 74 % of these 39 places first establishments have committed themselves as Conservative, while 10 or 26 % went Liberal in or before 1885. Of the ten cases deviating from H l, H 3 will account for five when the electoral majority at the place of publication is reckoned with, while H 3 accounts for altogether nine of the ten deviating cases when we use electoral support in the county as our guideline. This adds up to a level of confirmation varying between 87 % and 97 % when both H 1 and H 3 are applied.

Supporting evidence for H 2 is found in Table 7. When first and second establishments are compared in the 29 competitive markets where H 2 is applicable, the totals are predicted rather well by the factor of product differentiation alone. If there are 18 Conservative first establishments, this should be matched by 18 Liberal second establishments, and we find 16, or 19 when moderates are included. By the same token, when 3 uncommitted and 8 Liberal first establishments are found, this should produce 11 Conservative second establishments, and we find 10. Provided that all newspapers have committed themselves, the distribution in the party press by 1885 should be as follows when both H 1 and H 2 are applied: 61 Conservative and 38 Liberal newspapers. As will be observed in the Table, the agreement between expected and actual distribution is rather close if we reckon uncommitted newspapers as Conservatively inclined and moderates as Liberally inclined.

When product differentiation works properly it will be independent of current support in the electorate. This is not the case, as will be discovered, when first and second establishments are compared along the x-axis. Altogether seven deviations from H 2 are found, out of a possible 29. Anyway, this rather high level of confirmation might have been the result of systematic distortions in the configuration of commitments at local markets as predicted by both H 1 and H 2.

To test the agreement between the expected and the actual configurations, we have listed all cases deviating from H 1 arid H 2 in Table 8. As appears, there are altogether 16 towns with more or less deviating configurations out of a possible 29. However, when we apply H 3, between 12 and 14 deviations may be accounted for. That is, if we say that a conservative minority in the electorate at the place of publication will explain a liberal bias in product differentiation, and the corollary of this, that a conservative majority or strong support in the electorate explains a conservative bias, then H l, H 2 and H 3 together explain between 86 % and 93 % of configurations at competitive markets.

It may be objected that a bias in product differentiation should be compensated for by more than a mere majority in the local electorate. If we set the requirement for strong support as 60 % of valid votes for towns, the overall level of confirmation will be lowered somewhat to 75 % of observed cases. If we say that strong support

in the county is equivalent to 60 % for Liberals and 50 % for Conservatives, the level of confirmation will be 79 %. Another objection is of course that we use the election data of 1885 to postulate the behaviour of newspapers prior to that event. Publishers may, however, have sensed the direction of local opinion, not only on the basis of election results from 1882 when the franchise was more restricted, but also on the basis of prospective gains when the voter-market expanded in 1885. In any case, when 1882 data were used, the conclusions were not considerably altered. From this computation the level of confirmation varies between 83 % when election data for towns are applied and 86 % with election data for counties.

A more detailed analysis could probably have accounted for even more than our three hypotheses. The close interaction between competing newspapers in the early days of mobilization and the timing of evolving commitments may provide us with an answer to why the inverse pattern of liberal first establishments and conservative second establishments sometimes developed.⁹ But the same impression might well come out of such an analysis, namely that at the first stage in the formation of a party press both politics and business were considered by the publishers, but that politics seldom got priority over business. Generally both business and politics supported each other to expand the audience-market.

In a somewhat broader perspective the party press and its internal structure produced by political mobilization may be seen as a result of an early stage of expansion in the newspaper industry. In the 1880s it gave the party press a Conservative bias. Had the long-term developments in politics and the press come more out of phase, however, this bias might have become Liberal or perhaps more independent of parties. During a period of strong expansion and hot competition, when the conditions for journalists were more primitive, the press became more vulnerable to outside control, and this control seemed mainly to flow from that part of the audience-market which was most developed when mobilization came, namely the conservatively inclined upper-middle to upper class.

4.7. The Labour Party9_* was established in 1887, only three years later than the Liberal and Conservative parties. Its strategic position was, however, widely different from that of the bourgeois parties, on two important accounts. Whereas the Liberals and Conservatives started from a caucus in the Storting and immediately found popular support, the Labour movement had to build up from the bottom to achieve the same basis of political power. And whereas the Liberal and Conservative parties were mainly 'given' a party press by the private publishers, the Labour Party was eventually forced to establish its own press. Provided the conditions in the newspaper market combined with conditions in the voter-market to form a bourgeois party press, the priority of factors must be reversed to explain the rise of a Labour press. Here it was mainly the reluctance of established newspapers to side with Labour that more or less compelled the party to seize the initiative, and this initiative sprang from mechanisms in the voter market.

Until 1900, when universal manhood suffrage came into effect, the Labour
party was nearly barred from the voter-market. The first elections in which the

party presented independent candidates were, moreover, disappointing. The workers at first seemed politically restive.¹⁰ When Labour gradually became the largest party in votes, the bourgeois parties joined together to manipulate the election system and keep Labour from its due share of representatives. Equally decisive was the fact that not until the first two decades of the twentieth century did large-scale industrialization gain momentum. Many of the new industries were dispersed throughout the countryside, creating pockets of worker communities in predominantly rural areas. This apparently set: the tone and the programme for party action, with its heavy emphasis on continuous agitation to infuse workers with classconsciousness.

4.8. When the Labour press appeared at the turn of the century, Liberal and Conservative papers had already taken their fair share of the audience-market. Only a few privately owned papers sided with Labour in the 1890s and the first decade of this century. An old mechanism seems to have worked again. Established papers had their audience. To shift target groups to a more "unprofitable" fraction of a community must have been too risky for generally small business enterprises. Neither Labour nor the trade unions had the means of bying themselves into the market. The initial stage, therefore, was marked by the private efforts of idealists. In 1886 Social-Demo craten - the first modern working class paper - was startedon these lines. Ownership of the paper was transferred to the national Labour Party in 1892.

The need for a party press was realized early. A party organization without a press had gloomy prospects in sourounding conditions that opposed its efforts. Organizational work and the building up of a party press had to be linked together from the beginning. Party workers served both as agitators and salesmen for their press. Annual reports from the national party sometimes made note of the fact that local agitation was weakened because much of the funds were spent on newspapers, mainly Social-Democraten.

The most urgent goal to be attained was that of securing Social-Democrat en economically. The paper ran with a net balance for the first time in 1904. Until this task was accomplished the party refused any new responsibilities. But soon the situation was reversed. Social—Democraten grew steadily in circulation from about 4,000 in 1898, to 10,000 in 1907, to 38,000 in 1918 and to 58,553 by 1938. After World War II circulation levelled out at 67,000. Profits from Social-Democrat en were mostly used as subsidies for local newspapers. This practice was followed until 1938, when allocation of subsidies was centralized in The Press Fund of the Labour Movement, founded by a decision of the National Congress of Trade Unions. After World War II responsibilities have been shared among all party newspapers on a mutual basis together with the party and the Federation of Trade Unions.

The national Labour Party convention early developed a procedure for authorizing new papers. The most significant criterion was that of ownership: a 'party' paper could only be owned by local party organizations or trade unions. When the party had taken responsibility in the building up of a labour press, dues from

members could not be channelled out of the organization to private persons. In the beginning, however, the convention was reluctant to accept newspapers stated by local organizations. The reason was partly economic, and the underlying intent was obviously to save money by making Social-Demo_craten a national daily. However, as local party organizations gained in strength they often acted on their own, and the issue of one big daily versus many local papers arose in the convention itself. In 1904 the local press spokesmen had a slight majority and some projects for new papers were authorized in advance. Contributing to this shift in policy was the fact that Social—Demo craten had gained substantially in circulation within the Oslo area but was not a success as a national daily. From 1904 onwards the policy was adopted of strengthening the local party press.

A national committee was formed in 1904 and took over responsibility for authorization. In addition to ownership-policy it tended to ensure that new papers did not compete locally with established party papers. Considering the limited resources, the number of newspapers grew at an amazing speed. The succession of establishments seems to have followed a logical trend, from the big cities to smaller communities. Even if it had wanted to, the central party apparatus was unable to stop this trend. Newspapers were under strictly local control as long as they ran with a profit.

This decentralized system had its weak and its strong points. It had strength because local organizations often took a great pride in their newspapers and spent considerable time in raising money and in extending their circulation among supporters. This was also beneficial for the party, because the many committees needed to run a party paper, and the editorial board itself, provided an exellent training ground for a political career. As will be seen from Table 10, journalists are heavily over-represented in the Storting. Even more significant is the fact that the journalists who became representatives were mainly labourites. What the Labour press may sometimes have lost in journalistic expertise, the party has gained in having able representatives in decision-making bodies.

The weakness of the system was revealed in the 1920s when two major splits racked the party. In 1921 a moderate faction broke with Labour and formed the Social Democratic Party. In addition to some social-democratic papers already established, four of the Labour papers went with the new party. In 1923 the Communist Party was formed and 11 papers - including some of the larger ones - were lost to the communists. Out of 29 newspaper-owning organizations that stayed loyal to Labour, 13 had their editors replaced (Kirkvaag 1935). Most of the communist papers soon died, but the reestablishment of many Labour papers was a heavy drain on the party funds. In 1927 the Social Democrats joined Labour again and six papers were added by the merger. In number of newspapers the Labour press gained in strength by the splits.

This experience provided a lesson, and when the Labour press was reorganized after World War II one of the features added was a stronger control by the central party apparatus. Nowadays no paper can be disloyal to the party, and neither can local organizations use the plant and office in their newspapers if they want to side

with another party. In this case all property will be taken over by the national party
as legal owners of the installations.

4.9. What ever the Labour press has accomplished has been done through organizational efficiency and discipline. By planning and cooperation much of the fierce competition which has characterized the bourgeois press has been avoided. In 1912 the Labour party pioneered the organization of a press bureau to aid local papers in political journalism, furnishing them with commentaries and news from Oslo, maintaining a network of correspondents in the Scandinavian capitals and including every Labour newspaper as a correspondent for the others. The practice was later on adopted by the other party presses. On the other hand, the Labour press was often denied the advantage of an early start. Today it has gained 23 % of national circulation, comparable to that of the Liberal press, which was established under more favourable conditions. Compared to the great success of the Labour Party, however, the Labour press has been lagging behind. Mechanisms of the audiencemarket proved decisive in preserving the essential features that formed the pattern of the early party press, and this made the historical trend of the party press less vulnerable to changes that occurred in the voter-market.

5. From Mobilization to Stability off Commitments

5.1. Once the main characteristic of newspapers throughout the country had become one of political involvement, the system of party commitments tended to perpetuate itself. Newcomers, by the logic of entering a more settled market, had to commit themselves by taking a position different from the established newspapers so as to capture the remainder of a local market. This is not to say that markets in the 1880s to 1890s were yet politically stable. On the contrary, both the voter-market and the audience-market were expanding, and transitions of loyalities and new alignments made the market for newspapers politically unstable.

It seems natural that fluctuations in commitments should have followed in the wake of an initial mobilization. The popularity of politics from the end of the 1870s probably had a bandwagon effect on publishers, who often overestimated the economic potentialities of their party choices. A Liberal commentator of 1883 provides us with a clue in this matter. "As the pendulum swung towards the left [during the 1870s]", he remarks, "the press advanced even further in this direction than did the electorate." Trondheim was a prominent case, with only one Conservative newspaper compared to three Liberal, while 58 per cent of the electorate voted Conservative in 1882. The only explanation offered by the commentator was that the newspaper-reading public was more radical than the limited electorate at that time.11

When more than one Conservative or Liberal newspaper competed on the same local market, adjustments had to be made. Party "loyalities", of many shades and with many reservations, developed. Even though no systematic investigations have been made, these impressions are readily obtained from many local historiographies. In some places where one Liberal newspaper was more radical than another, the

latter drifted towards the right, becoming spokesman for "Liberals among Conservatives" or openly Conservative. In other cases the shifts started from an unspecified brand of liberalism or conservatism and moved towards a specified one, or the editorial policies vacillated between cautious party support and neutralism.

Manipulations like these may have been part of a planned market strategy, but instability of this kind is also likely to ha.ve been caused by fluctuations in the market itself, through mergers and by the buying and selling of newspapers. Another factor was the scarcity of competent personnel. Editors were not readily available and the politics of the newspaper often changed with the editor. Clearly enough, this was due to the situation in which the journalist found himself: Without professional prestige, and in the absence of a permanent party organization, the recruitment to political journalism seems to have been somewhat haphazard.

The basis of recruitment improved with both the advance of the printed media and the organizational drives of parties. As the barrier to entry into the newspaper market was erected by expensive technology, differentiation in the party press became more dependent on direct cooperation with the parties, which often provided newspapers with capital for investments. Through alternative means of direct and indirect support, parties have always tried to maintain the number of loyal papers at a high level and to keep present supporters in the press within the party ranks. The Labour, the Agrarian and the Communists press have all had regulations on the loyality of editors included in the party by laws. This undoubtedly reflects the conditions prevalent when entering a market already dominated by the Liberal and Conservative press, not formally bound by similar regulations.

5.2. How closely the changing political structure interacted with changes in strategies by publishers is hard to determine. Neither can we assess in more exact terms fluctuations in commitments and for what period they lasted. We can maintain with certainty, however, that the post-war period differs from the period between 1890 and 1930 with respect to political stability in the press, and that the era of fluctuations coincided with important shifts in the party system. Significant fluctuations probably lasted into the 19205: while the socialist parties have always been able to list their newspapers, and the Conservatives reportedly gave a complete list in 1924,¹² the Liberals did not make a similar classification at this time. It seems that stability of commitments first appeared at the extremes of the political spectrum.

5.21. We are able to present only limited evidence on the changes in commitments within the Liberal press by 1920. Comparing the lists of members in the Association of Liberal Journalists from 1894 to 1919 (Venstres Presselag 1919) with an almost complete list of journals and newspapers in 1920, which also contained classifications of party commitments (Diesen 1920), defections among original members may be observed and non-affiliated Liberal newspapers can be identified. The results are specified in Table 11. As can be seen, defections of members publishing until 1920 were exactly counterbalanced by the addition of non-members, but it is impossible to know whether this balance was maintained at all times. The amount of defections may in fact be reduced if we take Diesen's list to be incomplete. 'Commitments unspecified' in the Table means that commitments are not given by Diesen, and not that newspapers thus classified are neutral or diffuse in their commitments. Another interesting detail in the Table is the inverse relation of defections to non-member additions, chronologically speaking. This may indicate that the momentous battles of the early mobilization brought newspapers closer to the party and that newspapers

established after this period became less stable or more diffuse in their commitments. In Table 12 we see that competition among liberal newspapers located in the same town may account for 12 of 18 cases of defection. This points to an important trait of the market after the 1880s: for the bourgeois press at least, politics became an important factor of product differentiation.

5.3. It may seem mysterious that political journalism in general did not drift towards the middle ground of unspecified commitments. From this position editors could obtain a free hand in party politics. We shall return to this problem throughout the rest of this article. Let it suffice at this point merely to mention three important factors:

1. The success of mobilization in the press from the beginning of party politics made the newspaper industry one of the major fields of activity for parties and sometimes an integrated part of their organizations as became the case with Labour. When important districts and towns lacked a local party paper, parties and often interest groups started a new one. Newspapers in economic difficulties were often subsidized, tacitly agreeing to remain loyal to their party.

2. Polarization of local markets came with the rise of a Labour press. Labour newspapers were part of the organizational infrastructure of the Labour movement, a channel for party activity as well as for propaganda. This may have been a challenge to more independentminded editors in the bourgeois press that brought them back to the party.

3. The aggregate effects of party politics in the press that may have worked at
several levels within the industry were:

a) To delay the formation of strong professional associations among journalists and the establishment of a common training in journalism, prompting them to seek professional esteem and cooperation within, rather than across party barriers.

b) To dominate competition. When the majority of competing newspapers are party committed, the dynamics of competition seldom brings out the demand for politically independent newspapers, and will hardly be an impetus to innovations in political journalism outside party politics.

Some of the more important factors that many eventually contribute to the explanation of the drift towards stabiliy in pary commiments will probably be found among the conditions behind a gradual closing of the newspaper market. The rates of turnover and the prospects for new entries are closely related to stability in editorial policy. Without new competitors a challenge to settled positions will disappear.

6. Market Conditions Related to Party Commitments

6.1. In broader terms we may say that while changes in the party structure spring from changes in the voter-market and mechanisms of the election system, and while the changing structures of the press are dependent primarily on developments in the audience-market and the economy of publishing, the development and stability

of a party press springs from both markets to the extent that parties and newspapers
cooperate.

Improved technology set the stage for the rise of a party press. The same rapid change in social conditions that contributed to the spreading of a new medium also gave opportunities for new parties. At some point, however, the two lines of development had to cross each other. The economy of publishing could not allow all parties membership in the press everywhere. When due representation of political alternatives is no longer possible in a committed press, some adjustments have to be worked out. Adjustments may occur mainly at the audience level or by changes in the performance of newspapers. In a rational sense, adjustment in editorial policies is the easier of the two. It will restore a democratic ideal, but probably this ideal will not come to the fore until reserves for adjustments in the audience-market have been fully utilized.

As our argument ran, market strategies of parties and papers had a conflux when both the newspaper public and the electorate expanded. Partly at least, a confluence of market strategies depends on a sustained balance of developments in the two markets. This is difficult to control when the levels of expansion are different, or become so. The dilemma is amply demonstrated by a gap between circulation and the distribution of votes today when both markets are extended close to their limits (Table 13). When a point of saturation is reached in the two markets, the margin for adjustments diminishes. Hence, as a symmetry of favours given between party and press may be kept more easily at the elite level, mechanisms of the two markets could not prevent the development of an asymmetry between preferences of the audience and party commitments of the newspapers.

This in turn brings once more into focus our basic question: why have party commitments in the press survived the more drastic changes in the original conditions that brought them about? Some answers have been suggested, yet they seem insufficient. Our answers so far emphasize the effect of party interference - or the nonmarket control - while the normal function of market control is kept in the background. To restore the balance in our explanation, we must now turn our attention to the assumed effects of market control.

6.2. Non-market controls, as they were envisaged, are mainly effective when political parties can apply subsidies or alternative methods of reducing costs of publishing. Within the limits of this concept we are not able to explain the political behaviour of market-leaders, which mainly run at a profit. Accordingly this explanation is too narrow, but it does not imply that we must use a different kind of explanation to cover the remaining cases.

If we take it for granted that loyalties - once established - are easier to maintain than to change, we have in a sense turned the main problem around. The influence of market control is now assumed to be a negative one: newspapers will not change their commitments unless market conditions make it vital for continued expansion or survival. Non-market controls can be assumed to distort or confirm the normal function of market control and not to replace it.

More specifically our problem has become how editorial boards derive information from their market and transform it into a strategy. This retrieval process may be termed market control, and it is only partly a problem of direct feedback from the audience-market.

As meaningful information is difficult to retrieve from the prefereces of a hetereogeneous newspaper public, the perceived attitudes of the audience must be "normalized" to preconceived guidelines for journalistic conduct. Hence, as the demands of the audience are often acted upon in advance, reactions to journalistic conduct are registered by newspapers to the extent that they affects the paper's market position. I suggest that this is the normal course of events, and if this holds true, then market control is primarily exerted within the community of competing newspapers. Glues on which to base editorial strategies are simply the success and failure of competitors.

With this in mind it becomes clear why we cannot conclude that the system of party commitments will break down with an influx of marginal voters into the audience. If we conceive the editorial policy as including strategies towards both the audience and the competitors, a necessary condition for changing voter reactions to be relevant to the quality of performance of newspapers will, for example, be that the presence of non-committed newspapers is secured. Only when circulation of non-committed newspapers is growing may reactions in the audience-market be assessed as bearing more directly on the principle of commitments.

There is also some reason to believe that a wide margin exists in the interpretation of audience behaviour as support of the established system while the market expands. Marginal voters in an audience can be looked at in two different ways by editors: either the marginal voters are anyway basically apolitical and tend not to consider politics, or they are politically active and use agitation in newspapers as a source of information for different purposes. Neither of these will prompt editors to reconsider their strategies. What can in fact bring to mind the necessity of a change will eventually be discovered when consumption of newspapers decreases.

Let us assume that there are some limits to the deviance which can be tolerated between the preferences of the audience and the party platforms of newspapers. At some point the prospects of newspapers will be impaired by their stubborn adherence to traditional commitments. How far will the public adapt to a more permanent situation and when must the system give in?

The public will adapt to the extent that they give their attention and become readers. The level of consumption, then is a measure of the degree to which the public adapt. But as consumption is governed by other factors as well, the nature of adaptation to political criteria may differ with different levels of consumption. As we have already argued, newspapers are not prone to adapt to the public before the relevant clues are evident, which will occur primarily when the public stops consuming or threatens to withdraw its attention.

6.3. We have set forth two vital premisses for our reasoning: a) the public will
adapt to given alternatives as far as consumption goes, and b) newspapers adapt

to the public at junctures where consumption decreases, given some additional market conditions. The willingness or the capacity to adapt on both sides of existing channels will explain changes in the party press when the underlying determinants are found.

Our discussion will now focus on two problems related to the premisses just
mentioned. As it appears form the history of the Norwegian press, they are as follows:

1. A present imbalance between comparable market-shares for parties and newspapers,
and

2. The stability of party commitments in the press as compared to changes in the
voter-market.

As becomes evident, the two problems are closely interrelated: given the first statement, the second must be true. An imbalance cannot develop unless newspapers have been unwilling to adapt to existing preferences in the audience. But it does not follow that an imbalance will appear automatically in all instances of stability in commitments. Only when we take the level of consumption into consideration can the two statements, with some further qualifications, be interchanged.

Provided there is a strong relationship between the party preferences of voters and their selection of newspapers, the composition of audiences will tend to become politically homogeneous. The composition of audiences, as determined by current commitments in the press, will be a fair measure of the willingness of the newspaper to adapt to the voter-market. As consumption in the audience-market grows, however, the correlation between overt media preferences and voting behaviour will change. From the formula of 'one vote, many newspapers' we may easily conclude that as soon as the level of consumption rises above one newspaper per household, the audience of some newspapers must become politically more hetereogeneous. This will eventually be an indication of more rigidity in parts of the media system. From this follows an important rule: given different levels of expansion in the two markets and only one party preference by the participants, the level of consumption will interfere with the adaptability of the system.

Let us imagine that growth in consumption of newspapers became independent of technical innovations. Technology would nevertheless decide how a given volume of sales could be utilized by competitors,, If its capacity for production were sufficient, one newspaper would be able to meet existing demands for certain kinds of news at least. This is the equivalent of a national broadcasting system. On the other hand, when production capacity is limited, an increase of units in the newspaper industry correlates better with increased consumption. Hence, the flexibility of the media system relative to the party system is dependent on existing technology. The aforementioned rule may thus be reformulated to read: The combined effect of varying consumption and technology will interfere with adaptability of the press system relative to the party system.

One basic requirement for the overlapping of market-shares is, moreover, that
within districts of the territory, every party represented in the electorate must have

a newspaper represented at the audience level. The importance of districts depends on the demand for local news. If readers are given certain types of political information by one newspaper at a time, the same amount of, say, local news outside party politics can be given by all newspapers. When the economy of publishing restricts the number of newspapers within a district, some readers may combine local news and their preferred political information only when they read more than one newspaper. Hence, if the capacity to produce is sufficient, the press may meet another barrier, namely the capacity to cover local news in detail within the whole territory. We can therefore add another factor which restricts adaptability in the press system: the demand for local news, which may also be called the factor of local markets.

Closely related to the factor of market size for newspapers is the threshold of representation for parties under current election systems. It goes without saying that an overlap of comparable market-shares is easier to obtain when the number of parties is small. Any new party which appears at the national level must add to the amount of newspapers in the party press, to a point of near parity with competing parties. If the system is to work perfectly, consumption in general must not interfere with the viability of individual newspapers. This does not mean that a proliferation of parties and party papers also adds to consumption, which was obviously the case when the Labour press expanded along with the party. But when, for example, consumption stagnates, every new party added must be followed by a corresponding redistribution in reading habits under a strict rule of proportionality according to the distribution of votes. Proportionality is easier to maintain when both markets expand.

We may sum up this discussion by presenting the constellation of market conditions
which are deemed favourable or unfavourable to adaptation, as measured by
the composition of audiences or by the overlap in comparable market-shares.

The first three factors form a syndrome which it will now be our task to explain. This will be termed the 'economy of publishing', and will in turn serve as a point of departure when we want to deal with our second problem, that of stability of party commitments.

7. Changes in the Economy of Publishing

7.1. Economy of publishing is related to production methods by the initial investment costs of new machinery, and to the volume of sales that will return a profit when the price of 'first copy' is paid.13 We may term this the 'minimum circulation' that returns a profit. The size of the minimum circulation is furthermore determined

on the basis of the single copy price that can be obtained in the audience-market. Thus production methods influence the cost factor and more indirectly the level of consumption, while a combination of consumption and production costs determines the viability of established enterprises.

Initial investments, production costs and minimum circulation are the essential elements of the 'barrier to entry' into the industry. When production methods require much capital for investment, the barrier is raised. This is also the case when established newspapers have obtained a large circulation and lowered their single copy price accordingly, which means that the minimum circulation for newcomers is raised and the period of repayment of invested capital is prolonged. The barrier to entry is also heightened by alternative methods, such as increasing the frequency of editions and extending the news-service, all of which increase production costs. On the income side advertising mainly favours newspapers with large circulations, which also adds to the risks taken by newcomers.

7.2. Consumption of newspapers has been raised by new technology, first when the hand-driven letterpress was abandoned and then by transition from high-speed presses to the rotary press. With each transition efficiency was paid for in the first instance by more capital for investment. The first transition was linked with a general prosperity and growing demand in an underdeveloped market. The introduction of the high-speed press therefore accompanied a drastic increase in newspapers. The second transition, however, became more complicated. The audience-market was more developed and the rotary press also required more capital for investment than did the high-speed press. This gave the upper hand to established papers. The rotary press was a much stronger instrument to utilize expanding markets: marginal costs for an increased production made mass circulation economically feasible. As soon as investments were repaid, mass-production allowed newspapers to lower their single copy price or to extend their news-service. The net effect was a higher barrier to entry. From this stage onwards rates of growth in consumption and number of newspapers parted company.

Scale-economy, which came primarily with the rotary press, is dependent on larger markets. Hence, the size of towns became more important with the introduction of the rotary press. This was clearly also a factor which determined the rate of adoption of technical innovations.

7.3. While the component data which go into the definition of 'barrier to entry' are not available,¹⁴ some alternative evidence of the effects of a raised barrier is available for analysis. These effects are first a stagnation in the number of enterprises, and second an ensuing concentration among established competitors or stratification according to size. The final consequence of this process will be a decrease in the number of enterprises.

The size of enterprises may to some extent be judged from the weights of newspaper
volumes, systematically recorded by the postal service. Superficially applied,
these data have their pitfalls. We cannot say for certain that circulation alone

determine the weights of volumes produced.15 With these deficiencies in mind we shall be somewhat cautious in our interpretation of Table 14.16 The basis for interpretation will be this: when the number of large newspapers (viz. heavy volumes) is small, we shall conclude that these are newspapers with a wide circulation. When the number increases substantially we shall conclude that the correlation becomes weaker.¹⁷ Irrespective of this correlation we shall deem variances in the relative amount of large volumes a fairly good indication of variances in production costs and thus in the barrier to entry. When the proportion of large volumes has increased, the barrier to entry has also been heightened.

Table 14 may be looked at from two slightly different angles: along the temporal
dimension of changing numbers, or along the centre-periphery axis of urbanization,
ranked according to regions in the Table.

At the national level and along the temporal dimension we observe that the increase in large volumes gained momentum only when the number of newspapers levelled out and began to decrease from the 1920s onwards. This holds true also when seen with reference to regions. In all regions an increase in the average size of newspapers coincides with the levelling-out and subsequent decrease in number of newspapers. But we shall also discover a pattern when regions are compared. In the populous Eastern region the process started earlier and has gone farthest. Other regions lag behind according to their size. Furthermore, this development is advanced by comparable developments within the industry: the 1910s were marked by transitions from non-dailies to dailies, which took the form of an escalation to a new plateau - from 28 % to 41 % of the total number of newspapers (Table 15). This is probably directly connected to the slow-down in the growth of newspapers within more urbanized districts and the continued increase of average editions per week on those markets (Tables 16 and 17). When we return to Table 14 we can see the same pattern repeated: regional differences are levelled out when the total amount of newspapers shrinks. With regard also to number of editions per week, the differences between various levels of urbanization are decreasing (Tables 16 and 17). Regional differences, however, do not decrease when we control for the proportion of heavy volumes within each region. A disproportionate growth of larger newspapers in the Eastern region tilts the balance towards more central areas again (Table 14).

From these findings we can add some stages to the structure of growth in the Norwegian press. Expansion took place first in number of newspapers, next in a higher frequency of editions and finally in a growth in size of copy. The last steps towards a closed market are discernible in all regions. By 1900 newspapers were of a more even size, extremely so in the smaller markets of the periphery. Between 1900 and 1960, markets became increasingly stratified by size of newspapers in all regions. The periphery approached the developments of the more central areas, although it still lagged behind.

This pattern closely corresponds to what we may expect from a rising barrier to entry: the development brought fewer and bigger units, starting earlier in the richer markets of central urban areas in which the technology of mass-production suits markets better.

7.4. Market size was a factor determining the spread of publishing activity from the earliest years. As we have already observed, activity spread out of urban centres and in general kept pace with urbanization until about 1910 (Table 4). From this point urbanization took the lead while the number of newspapers stabilized in the larger markets and continued to grow in the smaller ones (Table 16). The reversed pattern which appears from 1910 onwards fits neatly into the aforementioned structure of growth, and one of the strongest determinants behind this development was probably the introduction of the rotary press.

Aftenposten, then as now the largest daily in Oslo and Norway, was the first newspaper to buy a rotary press in 1886. Bergens Tidende, the largest daily of the second largest city, aquired the first rotary press outside Oslo in 1898. Unfortunately our knowledge is incomplete on this account, we are not able to pinpoint the further

development, but it may be guessed without any difficulty. The rotary press spread
successively from Oslo to other cities according to their size, and at the local markets
it spread from the largest newspapers to smaller ones.

As we deemed the rotary press a strong instrument by which to freeze given market positions, especially among the higher-ranking papers, we shall conclude from the foregoing that traces of earlier developments may still be found in the present structure of the press. The chances of holding on to a position as market-leader should be inversely related to the size of markets. In the medium-sized and smaller markets, allowing for competition, early-established newspapers could not utilize the technology of mass-production to full advantage, or they have been unable to adopt it in time to prevent newcomers from gaining a bridgehead in their markets. This pattern becomes evident from Table 18: present market-leaders, having replaced older ones, almost exclusively settled in relatively smaller markets after the turn of the century, while leadership in the richer and more populous markets was reserved for earlier establishments.

7.5. In terms of party affiliation, market-leaders of the Labour press belong exclusively to the group of newcomers, a fact that may serve as a reminder of our point of departure for this analysis. The technical revolution of the industry conserved many of the essential features of the party press left over from an initial expansion of the audience-market. Important changes in later periods have mainly occurred through a gradual decrease in alternatives. It was in the smaller markets that the established order could be upset in any measure. This will explain many of the gaps that have developed between party platforms of newspapers and voting behaviour of readers, and it will also add some perspectives to the present structure of the press. What it cannot explain, since stability of commitments in newspapers is the rule under which this pattern can appear, is why adaptability was worked out mainly at the audience level. We shall only be able to present an answer after making another detour into the cycles of newspaper consumption.

7.6. Consumption serves a manifold purpose in our analysis. Rising consumption has helped to explain the speed and pattern of the adoption of party commitments and of new technology. Declining consumption may eventually contribute to an explanation of when and where newspapers will take their strategies up for a reconsideration. By itself, the level of consumption may tell us something about audience morphology. Expansions and contractions in the audience-market are related to strategies and more widely to the structure of the press, by their implications for the prospects of individual firms. As far as strategies concern the party system, they are also related to developments in the voter-market.

When the party press was formed the relevant factors which could explain it were primarily: a) a high rate of growth or very favourable prospects for expansion, b) a low level of consumption, and c) rising political tension. An expected disintegration of stable party commitments springs essentially from the same set of factors, but with a difference of values attached to them, namely: a) a low rate of growth

or unfavourable conditions for expansion, b) a high level of consumption, and c) rising political tension. Rising political tension is the common denominator and it deserves some further elaboration to specify its changing effect at different stages of development.

If the level of consumption roughly corresponds to the composition of audiences, political tension may say something about the probability of transitions in reader alliances. With low consumption, tension may not lead to realignments of readers since the audiences primarily consist of voters loyal to the party platforms of their respective newspapers. This is equally true whether the market expands or contracts. By contrast, consumption above the level of one paper per household within local markets indicates that the more prosperous newspapers have already been able to attract readers outside the party ranks. Political tension, insofar as it leads to realignments among readers, may cause a decline for these newspapers when it coincides with stagnation in consumption. Consequently, when stagnation occurs at a low level of consumption, it will provoke a strategy which works with the factor that prevents realignments among readers, namely party commitment. When consumption is high, however, stagnation may provoke a mixed response from publishers. Political tension at this stage will make the commitments of market-leaders especially vulnerable. With a politically hetereogeneous audience, market-leaders cannot hope to prevent realignments among their readers by pursuing a consistent party platform.

As will be remembered from section 6, we maintained that market conditions which are favourable to adaptation at the audience level work against adaptation among senders and vice versa. With this in mind, let us turn to the factors assumed to enhance or to impede an adaptation of the press system at large to changes in the voter-market (p. 116). Given stable party commitments in newspapers, they were primarily: 1. consumption, 2. production capacity and 3. market size. It follows from this that the same three factors can be used to explain conditions for party commitments in the press. However, one qualification must be added to the first factor, namely the level of consumption. From this we can set down the more general market conditions which can eventually break the system of party commitments in the press.

Disintegration in the party commitments of newspapers will occur with a higher
degree of probability in periods where the following market conditions are present
in this combination:

This configuration of market conditions differs from that on page 116. 'Production
capacity' and 'market size' have been replaced by 'barrier to entry'. This is done
mainly for the sake of convenience, because this composite term suits better the

aggregate data we must rely on. From paragraph 7.5., however, we learned that the benefits of big markets to the adaptability of the press system became less distinct with the advance of the rotary press. Thus the difference between 'capacity to produce' and 'market size' was gradually evened out as the audience-market expanded towards a state of saturation. At this stage the rate of growth in consumption becomes the only key factor, while political tension can be assumed to have a trigger effect. At lower levels of consumption a low rate of growth may only be a challenge to present commitments when the barrier to entry is also low and invites newcomers either to occupy positions close to existing ones or to point out alternatives to the current culture of political journalism.

The practical problem to deal with now is whether the aforementioned configuration of market conditions occurred at an earlier point of time. If not, we can explain why party commitments have survived throughout the Norwegian press. The contemporary situation will be dealt with in the next section.

7.61. Barrier to entry. Our indicators of a current concentration in the newspaper industry have been a growing proportion of heavy volumes, and transitions from non-dailies to dailies. On both accounts the 1910s mark the beginning of concentration and of a heightened barrier to entry. The transition to new technology and larger units in the press was obviously facilitated by the rapidly rising consumption during the First World War. What is remarkable, however, is the relative stability of the press - as it appears from our tables - which prevails for twenty years from the 19205. This prolonged period of slow concentration follows a slump in newspaper consumption from 1918 onwards, and covers nearly two decades of economic depression and stagnation.

7.62. Consumption. Very rough estimates set total circulation in 1910 at about a quarter of a million copies per day. Circulation rose to about a million copies per day during the First World War, from which it sagged during the 1920s and early 1930s and rose slowly afterwards beyond 800,000 by 1940 (Christensen 1964).¹⁸ This impression is also confirmed from other sources. If we estimate the volume of journals and newspapers mailed to correlate rather closely to the total circulation, we shall learn from Figure 2 that a temporary peak was reached by the end of the First World War, after which the curve bends downwards until 1936. Changes in total amount of newsprint used by the Norwegian press shows roughly the same cycle: a peak was reached in 1919 from which there was a decrease until 1924 and a return to the 1919 level by 1934 (Grini 1935).

If Christensen is correct we may read his estimates thus: in 1910 average consumption was 105 copies per day per 1000 inhabitants, compared to about 377 in 1919, about 213 in 1930 and about 275 copies per 1000 inhabitants in 1940. More accurate data for the post-war period give 465 and 462 copies per 1000 inhabitants by 1950 and 1965 respectively.

A governmentally appointed Comission on the Press says that the level of consumption has been calculated at one newspaper per household at the end of the 1930s ("Stat og Presse", 1967). Coverage reached 1.6 copies per household a day in 1960, which is about the average for the post-war period.

In 1960 the ratio of households to inhabitants was 1 : 3.1. If we allow for improved housing conditions we can stipulate a ratio of 1 :4.0 as an average for the pre-war era. The two values for level of consumption which follow will be given alternately, the 1960 ratio in parenthesis: By 1910 average consumption reached 0.42 (0.33) papers a day per household, in 1919 it was 1.50 (1.17), in 1930 0.85 (0.66) and in 1940 about 1.10 (0.85). We may thus conclude somewhat tentatively that only for very short periods at the end of the First World War and towards the beginning of the Second did consumption go beyond the level of one newspaper per household.

7.63. As a summary, let us assign values to all of our three factors as they appear
from the historical evidence since 1910. When we apply three values, Low, Medium
and High, they will appear as in Figure 6.

Except for the recent period, it is only at the beginning of the 1920s that we find a configuration of market conditions which is favourable to adaptation among senders. We have thus in a way been able to explain why party commitments have survived in the Norwegian press. But we should still bear in mind some reservations:

- National averages are a meagre substitute for local variances. Consumption in cities may differ strongly from that in rural districts. On the other hand, improved transportation may have opened rural districts as an additional market for expansion when consumption in cities reached saturation point.

- Concentration in the newspaper industry may have made some audiences heterogeneous before the level of one newspaper per household was reached. On the other hand, concentration makes for an unbalanced growth in the industry, as market-leaders can expand while others decline. This cannot be seen from the national averages. Thus a slump in total circulation does not automatically make sense out of the additional condition - that market-leaders or newspapers with a heterogeneous audience will also stagnate.

7.7. The overall implications of technical developments for the party press can
be summed up thus:

While the barrier to entry was low and minimum circulation small, the press system was more capable of adapting to changes in the voter-market. Given levels of consumption were utilized by a relatively higher number of newspapers, which could occupy positions across the party spectrum, and sometimes forshadow future changes. Scale-economy made the system more rigid. Even important changes in the voter-market became harder to observe when the barrier to entry increased and stabilized given commitments in the press. Adaptation was therefore given over to

the newspaper public. Smaller groups within the electorate often had to adapt to
newspapers representing larger groups.

The assumed effect of new technology on the structure of the political press was therefore two-fold: first, gradually to freeze the selection mechanism dominated by party preferences of readers in the early stage of expansion, and second, to confine the basis of selection to parties which had newspapers established before the introduction of improved technology. However, as market size became a factor in the adoption of innovations, the basis of selection was open to changes for a longer period in the medium-sized and smaller markets, even though these markets did not allow for much variety.

The specific conditions deemed favourable for adaptation among senders were seldom arranged in the "ideal configuration" by historical circumstances for a prolonged period. The two main exceptions will be found in the period which preceeded the formation of a party system and in the very recent period. This will be described in our next section. Beyond the rapid mobilization of a party press the steadily increased consumption gradually made mass-production possible and in turn closed the markets to new entries. The geographical market structure, however, remained stable and tended to keep the average market size small. By the same socioeconomic developments the electorate expanded and provided a basis for new parties. As long as the newspaper industry expanded by number rather than by size of units, the new parties were able to gain bridgeheads in the press. With a reversing trend the adaptation of the system became increasingly harder when the stable commitments were upheld.

Adaptation among newspapers was not especially enhanced by this development except for very short periods, mainly because consumption was obviously kept at levels where the composition of audiences were generally politically homogeneous. Fluctuations in the rate of growth of consumption, then could not have the expected effect on market strategies of newspapers.

8. At the Crossroads: The Last Two Decades in Perspective

8.1. The post-war period stands at the junction between a past which gave many openings to the initiatives of entrepeneurs and a future that makes publishing increasingly dependent on organizational efficiency. In many respects the beginning of this period was abnormal, but it gave the industry a respite from which to gather strength after five years of enemy occupation and disrupted operation. For the rest of this period a rapidly modernizing economy and return to free enterprise strengthened the structure of the industry. This development will eventually bring strains to bear upon the old tradition of a party press and weaken the foundation of its popular support.

8.2. The German occupation from 1940 to 45 was hard on the press. About fourfifths of the newspapers were forced to close down by nazi decree and the rest were subject to strict cenconship (Johansen 1945). The radio was completely controlled, and only the relatively few members of the nazi party were allowed to keep their sets.

A clandestine press of the resistance movement circulated mimeographed newssheets through secret channels which in some measure replaced an ordinary newsservice, but the public at large became undernourished as far as reliable news was concerned.

The return of a free press was met by a phenomenal consumption of all kinds of newspapers during 1945 and most of 1946. It was probably at this moment in the historical progression that the level of consumption definitely went beyond one paper per household. A temporary setback occurred in 1947-48, but this was followed by a more steady-growing total circulation (Werner Erichsen 1960).

As is brought out in Table 19, growth in consumption during the post-war era shows two distinct movements as compared to the increase of the adult population - one of expansion until 1955 and from then on a decline until recently. From subperiod to sub-period, however, the rate of growth is almost halved.

8.3. Altogether the post-war reconstruction forced a more direct control on free enterprise. Three somewhat overlapping periods can be distinguished: 1) the period of reconstruction in 1945-49, followed by 2) a period of relaxed control from government in 1949-52 and then 3) a period of normal expansion from 1952 onwards (Central Bureau of Statistics, Oslo 1965).

A roughly similar cycle can be found in the economy of the press. Newsprint was rationed between 1940 and 1952. After 1945 compensation was given to newspapers through subsidies including further reduction in rates for postal distribution. Part of this deal, however, was a price control on advertising and subscription rates. Government interventions consequently set some limits on expansion, but one sideeffect was a more relaxed competition among newspapers. Subsidies were given on a limited quantum of newsprint for each paper, working mainly to the benefit of the smaller newspapers. Reduction in postal rates worked in the same direction to favour the non-dailies, and price control on advertising in some measure reduced the privileged position of market-leaders. In all, the basis of income was on the average

decided more by circulation than by advertising. For the total press 53 % of incomes
originated from readers in 1949 compared to 46 % in 1939.19

From 1954 onwards all restrictions on copy size were abandoned along with subsidies on newsprint. Included in this arrangement was a virtual freedom of newspapers to decide on their own prices. Originally a part of economic planning, government control gave some precedents to the economic organizations of the industry. Newspapers have since generally worked, in agreement to bargain a joint price for newsprint from the pulp factories without any discount on larger volumes bought, which have been to the benefit of smaller newspapers. Although forbidden to make an agreement on joint prices for subscription, newspapers have usually refrained from local competition on prices.

8.4. Stability of voting behaviour and a lack of ideological conflicts may have helped concentration in sales of newspapers, or at least not prevented it. This defies explanation by political factors alone. Market-leaders are the winners, their gains in circulation are generally out of proportion as measured by their share of the audience-market.2o As a group, market-leaders have expanded more than the entire Norwegian press, by 112.2 % og the national increase in circulation between 1949 and 1965. The prevailing concentration in sales becomes increasingly amplified by the unbalanced distribution of revenues from advertising, which favours the market-leaders in undue proportion to their circulation figures.²¹ From 1951 business advertising accelerated, the major part of it going to the newspapers.²² By the amount of capital spent, advertising has increased more than tenfold by now.²³ In the same event, the basis of income was tilted so that by 1965 only about 40 % of revenues originate from readers (Stat og Presse, 1967).

In per cent of total income for the Norwegian press, the share taken by marketleaders increased from 48 % in 1950 to 56 % in 1965. During the same 15 years the number of newspapers running a profit has slightly decreased, while both profits and deficits have increased. Out of 97 newspapers publishing continuously throughout the period, 49 had an accumulated profit of 2.8 million kr. in 1950, compared to an accumulated deficit of 1.0 million kr. for 48 newspapers. This situation did not change very much up to 1965, when 46 newspapers showed a profit amounting to 8.0 million kr. compared to 51 newspapers running a deficit of 3.9 million kr. (Stat og Presse 1967). About half the Norwegian press is not a prosperous business at all and has to be supported by various means.

8.5. This development has brought the organizations of the press to ask for government protection from the unfortunate effects of a free enterprise system. Current market mechanisms progressively endanger competition and a fair local representation of the major party alternatives. Of course this development is not only the concern of the press. Even the most wealthy parties now see a gradual erosion of their chances of staunch support from the local press in major districts of the country. By initiative from the press, the Government appointed a committee in 1966 to give advice on how government should support the press, if at all, to avoid a

further decline in numbers and at least keep the present structure of the press intact.
A report was submitted to the Prime Minister in September 1967.

The committee consisted almost exclusively of representatives from the press. From its investigations this committee - somewhat predictably - concluded that unless the government provides an annual amount of about 22 million kr. (about 3 million dollars) in subsidies the current status of the press would be in danger. The various measures proposed can be classified into three categories:

1. Measures intended to counteract effects of prevailing mechanisms in the
market, especially related to concentration in sales and advertising.

2. Provisions which can contribute to more efficient production, better means of
news gathering and cheaper methods of distribution.

3. Recomendations to upgrade the quality of journalism through better educational
facilities and research.

The committee was divided on a proposal for a graded subsidy on newsprint. The majority recomended a 53.8 % reduction of current prices for newsprint up to 400 tons annually. Categories of newspapers intended to receive such subsidies are: newspapers with a circulation of less than 10,000, irrespective of their position on local markets (market-leaders included), and newspapers with a circulation of 10-40,000, market-leaders excluded. Newspapers excluded from this arrangement were: market-leaders with circulations above 10,000 and all newspapers with a circulation of more than 40,000 copies per day. Two dissenting factions in the committee either opposed a graded subsidy and wanted all newspapers to receive support, or rejected the idea of state subsidies to the press.

Another measure intended to counteract effects from concentration in private advertising is supposed to be a drastically increased advertising by the central and local administrations of new regulations, laws, vacant positions etc. This kind of advertising would be given in equal amounts to all newspapers when the issues of advertising so require.

To provide necessary capital for investment in improved technology etc., it was proposed that the Government should establish a financial institute which can allocate cheap loans to the press. About l/$ of an estimated demand for annual investments would be covered through this channel. In addition the committee recommended that the Government should further reduce the rates for postal distribution and telephone and telegraph rates for the press.²⁴ At the same time it should ease restrictions on which categories of newspapers are allowed to use the mail for distribution within their own postal districts, i.e., including the dailies.

Proposals on how to improve the standard of journalism include more scholarships for journalists and compensations for newspapers that want to re-educate their staff, as well as a strengthening of resources for the Norwegian School of Journalism and for press research in general.

The committee suggested a pooling of machinery within districts and further technical cooperation between newspapers on new methods of production. To be most effective, the committee recommended that cooperation should take place across local political barriers while the present cooperatives are run exclusively on

the basis of harmonious party commitments. The committee also suggested that
newspapers should exchange editorial personnel more often, to broaden the professional
experience of journalists.

Nowhere in the report were the political dividends calculated. Such considerations were clearly outside the committee's task. To make sure that the weeklies of the Christian Peoples Party, the Communists and the Socialistic People's Party were included in its recomendations, however, the committee intentionally deviated from its own definition of a newspaper.²⁵ The prime objective given, no alternatives being mentioned or discussed in the report, was to preserve the essential features of the present structure of the Norwegian press.

The only concern expressed about the effects of its recommendations are statements on whether newspapers will be more inclined to support the administration in return for subsidies received. Such a possibility was refuted. The lack of a wider concern, however, leads the committee astray from the underlying principles of why and how support it needed for the press outside the ordinary market. The committee confines itself to confirming the old tradition that they see threatened, by declarations like this:

"The press is the most important instrument of influence given to political parties in this country. In the various walks of life the press provides an effective means of free expression and an important forum where ideas can meet and be challenged. Television and radio cannot be a substitute for newspapers in these matters." (My italics).

That this statement may contain some inconsistencies does not appear from the report. The addressee, however, is clear, namely the Storting. By emphasizing their alliance with the political elite the committee becomes entangled in the same net of forces that brought it together. The idea of a proportional party representation in the press is damaged beyond repair if economic arguments are applied. Any other reasonable balance will also be hard to obtain as long as the tradition of party commitments prevails. On the other hand, if it suggested that the current system may be outmoded, the committee would not command sufficient support from the press. Therefore, conspiciously absent from the report are reflections on how to increase diversity in the press to cover a broader spectrum of opinions.

The immediate effects of the recommendations will be a temporary status quo. But by the imperative of technology the only alternative to subsidies will be increasing monopolization locally. A minimum requirement for a diversified press is a competing press within districts, even if some further centralization does not impair the idea of a press alerted to the changing climate of opinions. It was this base-line that the committee tried to defend. This was their real concern even when they realized that future reductions are unavoidable.

8.6. The marked imbalance between the strength of parties and that of the party press was described by Herbert Tingsten as soon as standarized circulation figures were available (Tingsten 1949). A negative discrepancy was most apparent for Labour. By this account the Labour party was "underrepresented" on a national

press basis by 29.3 % in Denmark, 28.5 % in Sweden and by 22.5 % in Norway, during 1946-48. In the same way the bourgeois press was "overrepresented", in the extreme by the Liberals in Sweden (28.3 %) and then by Conservatives in Norway and Denmark (20.0% and 15.2% respectively). Although the exactness of such figures might be increased, more recent computations tell the same story. Only minor fluctuations have occurred in Norway, as will be learned from Table 13. In the period 1949—65 Labour has improved its position, owing to a growing marketshare of its press and to a decline in voting strength.

The discrepancy between votes and circulation varies by region in all three Scandinavian countries mentioned (Westerståhl and Janson 1958, Rokkan and Torsvik 1960, Thomsen 1965). Where the parties have their strongholds, their press is better represented than in the national average.

Another problem concerns the meaning of this discrepancy. As is shown by Table 20 it varies with the method of computation. The negative discrepancy for Labour is considerably diminished when we compare the more meaningful measurements of circulation per household on the one hand with votes in per cent of electorate on the other. When totals vary between the size of the electorate and that of the audience, aggregate data cannot tell more exactly how the electorate is reached by the party press. In the national sample of the Gallup poll in 1960, 53 % of Labour voters subscribed to a Labour paper. Exactly the same proportion of Liberal voters were loyal to their party press, irrespective of the vast difference in votes compared to Labour and with a press equally strong in circulation. Equivalent figures for Conservatives and Agrarians are 71 % and 21 % respectively (Torsvik 1968). The fact that the discrepancy for Labour can be levelled out so much stems from their strength in the local press. Labour newspapers are regionally more dispersed and thus geographically more accessible to party followers.

Recent findings pose a paradox hard to explain in this respect, except in a historical
perspective. Measured by kommune?6 per cent of party votes and circulation

shows a remarkably high correlation.²⁷ Yet, as measured by regions, the consistency disappears when observed over a recent, period. The same paradox is apparent in a detailed ecological study of the Swedish press by circulation districts. Westerståhl and Janson found reasonably high correlations. Yet when changes in the two variables were compared from one election to another, correlations became much weaker (Westerståhl and Janson 1958). In Denmark Niels Thomsen found in a majority of towns an increase of votes inversely related to strength in the party press during the general elections of 1950-64 (Thomsen 1965).

The essence of these findings points to the same conclusion as that emphasized throughout this article: Compared to earlier periods the political factor plays a diminishing role in recent developments within the audience-market. There may be compensating factors, concerning the process of political information spreading, e.g., a more systematic selection of party papers by readers with increased political activity (Rokkan and Torsvik 1960), and a mobilizing effect of political campaigns on TV (Valen and Torsvik 1967). Nevertheless, an overruling tendency in the post-war period is one of depolitization. Countering factors do not seem strong enough to bring trends in the audience-market and the voter-market into closer harmony.

8.7. Two essential market conditions favourable to adaptation among senders vis å vis their audience have been present for about a decade now. They are a level of consumption at about 1.5 copies per household and a low rate of growth in consumption. The effects of a low rate of growth, however, have partly been compensated for by sharply increasing advertising revenues. Our third important factor - the barrier to entry — has seemingly proved the strongest one concerning the propensity to conserve the present system of party commitments. A by-product of a closed market and a growing concentration in the industry has been political stability within the party press. Yet, on closer inspection some expected consequences of the prevailing trends in the audience-market become appearant in journalistic behaviour. Latent strains in the transactions between the parties and their press have not yet caused a break with the tradition of stable commitments, but this may be due to the placid political atmosphere which has charachterized most of the postwar era (Torgersen, 1962).

The two last parties to appear in the political arena - the Christian People's Party and the Socialist People's Party - have not been able to make much headway with a press of their own. Both parties were established on a nationwide basis after 1945 and both have had to rely on a weekly journal with limited circulation to make their day-by-day policy known to followers. In 1967 the last remaining communist daily was turned into a weekly newspaper. Thus three of the seven parties presenting candidates in the last general election, and with a combined share of valid votes of 15.5 %, have no spokesman in the daily press, while the other four parties dominate the Norwegian press. In the pre-war era all parties represented in the Storting, and some minor ones outside it, had newspapers committed to their cause somewhere in the country.

Only one defection in party commitment has been registered throughout the whole post-war period. A Liberal newspaper in the Eastern region has become non-committed by its own classification, but may more properly be classified as a Conservative newspaper.²⁸ Through mergers some bipartisan newspapers have emerged, altogether six of them in 1967.²⁹ Rising costs and slimmer profits have also tempted some newspapers to form cooperatives, by which they share technical facilities and can offer a joint service to advertisers and often demand higher prices. Sometimes these cooperative extend to joint editing of national and international news as well. Thus cooperatives often develop into actual regional syndicates which further centralize news dissemination in the press beyond the infrastructure of national press agencies. An indication of this internal restructuring in the press is the fact that the number of cooperatives has increased rapidly in recent years.

8.8. If the present situation does not provide us with any direct evidence of future changes in the party press some clues about underlying tendencies are easily sensed. For one thing, the report from the Commission on the Press started an intensive and for the most part an open debate among journalists on the principles and justifications for the present system of single commitments in newspapers. If a drastic breake with the past is not likely to be effected by the present generation of editors, we may nevertheless settle for a more restricted hypothesis about an ongoing trend towards more lenient propaganda in the party press.

Our hypothesis will be mainly derived from the trend towards concentration in sales at the audience-market. Concentration threatens to topple the delicate balance of countervailing interests in the Norwegian media system. However, our task will not primarily be to outline how this may affect the larger system, but more specifically to point out how newspapers react to the various market situations that concentration in newspaper sales produces.

Local market boundaries have generally been stable throughout the period. As will be seen from Table 3, the number of publishing towns has not decreased between 1949 and 1965 while the net decrease of newspapers was 34. This means that the continuing concentration is likely to increase the variety of party preferences within the audience for market-leaders, which is amply demonstrated by the fact that in 1961 all market-leaders had a level of coverage for households higher than the comparable share of the voter-market that their respective parties had managed to win. This situation has produced two distinct types of positions within local markets, which cut across the traditional classifications in the party press: the dominant position of market leaders with its concomitant morphology of a heterogeneous audience by political criteria, as opposed to the minority position of other newspapers which have a comparably more homogeneous audience. This seems to hold true, irrespective of the various party affiliations of market-leaders and independent of whether their 'platform' represents a majority of the electorate or not.³⁰

The potential feedback from the audience to editors should then work alongside
this pattern, either directly by overt responses from readers on the current performance
or more indirectly through the potentialities for future gains or losses in the

audience-market. As sketched out, market-leaders now have their most promising market among voters marginal or even alien to their present policy, while nonmarket leaders are left with readers more loyal to their present platform. Consequently our hypothesis will be:

H4 The more heterogeneous is the audience of a newspaper paticipating in the
voter-market, the lower will be the frequency of politically committed expressions
in the medium.

To test out this hypothesis we decided to measure the intensity of commitment in newspapers by means of counting the frequency of party-committed editorials. Editorials were classified as political or apolitical. 'Political' editorials contained unequivocal statements in favour of their "own" party or attacked other parties. If no parties were mentioned, the occurrence of ideological statements which were slanted and corresponded to present cleavages in Norwegian politics was interpreted as equivalent to party-committed statements. Apolitical editorials, by contrast, contained no such statements. In some very few cases — less than ten altogether - newspapers ran an editorial in which some reservations towards the policy of their "own" party were included.

Editorials were classified according to these rules for a period of two months - March and September - in all non-election years between 1949 and 1961. Our sample consists of all party-committed newspapers published in five towns. In two of them a Liberal newspaper was the market-leader, in another two the local Labour paper was market-leader, and in the capital there were two Conservative newspapers, one of them being the market-leader.

In Table 21 newspapers are ranked according to market position, as measured by circulation, in each of the five towns. As will be seen there are some local variations as to the propensity to campaign in the editorials. Frequencies are obtained by computing days with at least one committed editorial as a percentage of the total number of days analysed. As is thus demonstrated, our hypothesis above is confirmed.

Along the y-axis we may observe that market-leaders show less intensity in their party commitment compared to their competitors. This tendency is even more general, as was indicated by our hypothesis. The wider the circulation of party-committed newspapers, the less expressed are their party commitments in editorials. Observing along the x-axis we see from the Table that expressions of conflicts have apparently decreased in the press during the 19505. With some few exceptions this holds true for our entire sample. By the same process the difference between marketleaders and their competitors also became more distinct.

How far this trend may go before the very idea of single party commitments in the press will collapse is hard to say. What we shall maintain, however, is that depolitization has its limits in another basic idea of newspapers as opinion makers. Faced with conflicts - especially those which do not easily correspond to present cleavages within the party system - newspapers are supposed to come out with an

opinion of their own lest their position as movers in the political arena be seriously impaired. Such a situation in turn will bring into focus the hidden dichotomy between obligations to the parties and demands of the public. For market-leaders at least, one solution of this problem is clearly to become independent of existing ties to the parties.

9. The Impetus to Change

9.1. As an alternative network of communication for the administration, the potential power of the press was early appreciated, first by the Grown and then by the growing Liberal movement.The idea of newspapers as mediators between the public and the executive branch of government was conceived along with the early advance of the press as a medium for the masses. This idea has remained a basis for the defence of the press as an institution independent of political control, but possessing political influence. The bottom gradually fell out of this defence as new institutions wielding political power emerged in a pluralistic society. A simplistic view of the press as an agent in support of either a unified establishment or a public of homogeneous interests is no longer attainable, because both have long since ceased to exist.

In the realm of government the press can be an effective agent of change for a limited part of the decision-making process, to the extent that it sways public support from one party to another, or by its more direct influence on the policy of one party alone. However ephemeral public opinion may be, the idea of an independent press demands that journalists offer an interpretation of it to their readers. In this triangle of public, press and parties the journalists must find their own identity, either in competition with parties or by making an alliance with the political elite. Whether it tries to compete or not, the press will ruin its effort to win influence if it estranges itself from the party elites. Journalists must be close to the political arena, whether their alliance with politicians becomes close or insecure.

9.2. In the case of Norway, journalists have traditionally maintained a stable and loyal alliance with parties from the earliest years of political journalism. Through their alliance Norwegian journalists have been more active in shaping political decisions than those of many other nations. But the mixed roles of journalists and politicians have confused an important function of newspapers as channels of information on domestic politics. The interchange of roles is only possible so far as the professional journalist remains a secondary or intermediate politician. The economic conditions of most Norwegian newspapers has not yet elevated political journalism above party politics, but when publishing gradually becomes an industry it will require journalists to stick more closely to their job.

If approved, recommendations from the Commission on the Press can change this situation. When politically conditioned subventions are replaced by more neutrally attached subsidies, much of the power of the parties will be broken and some newspapers may eventually feel more free to invent their own policy. When somewhat relieved in their daily routines, the editorial staff will not be so dependent on featurized commentaries from the party agency. Finally, when political barriers are ex

ceeded by a necessary turnover of personnel with common training in journalism, the press community may find a new identity in their primary function as disseminators of news and commentaries. The Commission on the Press invites newspapers to go even further by pooling resources across the party barriers.3l

These trends slowly counteract a political inbreeding in the editorial board. Some few traces of changes to come, however, cannot deny the vitality of the traditional system. The crucial point is that even market-leaders keep intact their bonds of loyalty with parties. As long as this situation prevails we shall not expect changes from other papers. On the other hand, strains are brought upon stable party commitments from two directions in the present structure. Market-leaders of the daily press may eventually find a rationale for compromise between conflicting loyalities in an expanding audience of many political shades. Parochial newspapers, at the other extreme, find themselves in an almost identical position with regard to the composition of their audience. Only the vanishing numbers of medium-sized dailies are left with an audience similar to that which originally gave the impetus to a party press.

The winds of change blow from many directions - from the audience-market which makes special interest newspapers less feasible, from the electronic media which are uncommitted in principle but nevertheless constantly evoke debates on public affairs, and even from the political elite where compromise, out of accord with traditional cleavages, is the vital condition for working coalitions. Ultimately, all of this may bring the future of the Norwegian press into line with what happened in England some fifty years ago. Whether this will in fact be the shape of things to come remains to be seen.

Summary

In this article we have tried, first to outline the more simple features of growth in the Norwegian press in the course of a century, and then to apply some simple propositions on how expansion and contraction in the audience-market have affected editorial strategies in the voter-market.

As we came to see it, expansion in the press followed from a modernizing economy which started in the 1850s. Technical innovations in the printing trade faciliated this expansion and slowly set in motion a process making the printed media into a social institution. In the wake of economic progress came electoral mobilization and the foundation of a party system.

The timing of these events has a special relevance to our theoretical problems. Technology opened the market for new entries with the consequent spreading of publishing activity throughout the nation. Sales, however, were probably lagging somewhat behind this activity, which required publishers to adopt a more conscious and active market strategy. Electoral mobilization seems to have been instrumental in expanding the circle of newspaper readers, as it provided publishers with a cause which easily combined with a more pragmatic strategy: either to keep a given audience or to evoke attention and demands from a new and accessible, albeit unexplored part of the community market, namely the middle class and the petite

bourgeoisie. The less developed habit of regular reading and the high level of political tension at the time prevented newspapers from achieving both goals simultaneously. This eventually resulted in a diversification of newspapers throughout the country, leaving newcomers on the local markets with the task of expanding readership and of mainly becoming Liberal in outlook. We have termed this situation the first stage in the formation of a party press.

The pattern of establishments in the Labour press is most easily explained from the fact that only a few privately-owned newspapers sided with the Labour party when it appeared at the turn of the century, which left the initiative to the local party representatives. This second stage of development in the party press, then, spring essentially from the dynamics of the voter-market and party politics. It may, however, also be explained from developments in the audience-market proper. The social spectrum of newspaper readers during this phase of expansion probably did not include much of the working class at first. Furthermore, the terms of trade for established newspapers did not force them to go out for an additional target group under the threat of defections from the ranks of present readers. The less inviting prospects of this expansion found an ideal challenger in the more tightly organized Labour party. The requirement of party ownership, however, may have frightened away other newcomers from joining efforts in this respect.

This makes the mechanism of expansion and ensuing mobilization of the audiencemarket during the second stage rather analogous to that of the first one. In both cases newcomers on the market were mainly left with the risks of exploring new and unknown possibilities and in both cases they found a cause and a target group within the voter-market. From this pattern of editorial strategy it also follows that electoral mobilization became an important factor both in expanding and in diversifying the Norwegian newspaper industry.

As long as the audience-market expands it may contain enough resources to allow for a steady diversification of newspapers within districts along with the party system. A dominance of the party press by itself invites initiatives from the parties. But it is also clear that a business rationale for political involvement is primarily confined to the audience-market. Thus a hidden dichotomy between more pragmatic strategies and party commitments was conceived along with the technological and economic progress of the trade. What: has saved the system of commitments from disintegration is the additional resources which were partly found in the parties and partly came from advertising. Parties counteracted declines in their press with a bargain including subsidies to newspaper's and positions within the political elite for journalists. The inertia of the tradition was not seriously challenged, because parties controlled the behaviour of smaller newspapers, because the market became closed to newcomers with potential independent platforms, and because marketleaders continued to grow undisturbed by fluctuations in the voter-market.

A dilemma between alternative strategies has only recently come out into the open. As the audience-market approaches a level of saturation, a trend towards simpler structures of mass communication has also become dicernible, e. g., in readership patterns, which is further amplified by the strategies of advertisers. The demand for

technological and economic efficiency as well as journalistic proficiency has raised the price of status quo in the party press. This extends to the amount of support needed to keep the weaker parts of the party press alive as well as to the potential losses of attention for the surviving giants.

In this situation the government is asked to provide the necessary capital to save the institution of the party press. However, as soon as support is released from its more direct links to the voter-market - in this case by a more automatically functioning system of subsidies - the editorial policies will probably drift closer to developments in the audience-market. Whether the government becomes the benefactor of smaller newspapers or not, the mechanisms of the audience-market under the present terms of trade will drive editorial policies in this direction anyway and will probably do it faster. This development may open a new market for political journalism more consciously geared to supporting the parties among periodicals and magazines, which are rarely committed to any political cause.

---

NOTES

1 The above data on newspapers have been assembled from Fiskaa 1935, Kirkvaag 1960, and Christensen 1964. By modern standard expansion is somewhat slower than indicated, because the aforementioned authors are counting publications which may be classified by different criteria, as weeklies or journals.

2 The definition for the construction of the Tables is as follows, if not otherwise made explicit in the text: Titles found in the list published by the General Post Office in Oslo ('Poststyret' in the Tables), having a minimum frequency of publication two per week or more are considered 'newspapers' unless the title proves a hollow one, e. g., only a proliferation in names noted by our data source or in the reference source (a nearly complete bibliography published by the University Library in Oslo). By judgement some few titles were excluded from counting when it was either known or became evident from the two sources that they covered special purpose publications. By this procedure the places of publication were found.

---

2 The definition for the construction of the Tables is as follows, if not otherwise made explicit in the text: Titles found in the list published by the General Post Office in Oslo ('Poststyret' in the Tables), having a minimum frequency of publication two per week or more are considered 'newspapers' unless the title proves a hollow one, e. g., only a proliferation in names noted by our data source or in the reference source (a nearly complete bibliography published by the University Library in Oslo). By judgement some few titles were excluded from counting when it was either known or became evident from the two sources that they covered special purpose publications. By this procedure the places of publication were found.

3 To classify the size of places, we have used the very thorough study of urbanization from 1875 to 1950 made by the demographer Hallstein Myklebost (Myklebost 1960). His data on residents centrally located in towns are independent of the changing administrative units throughout the period covered. His minimum level for counting is 200 people centrally located, which means houses clustered with a separation of less than 50 metres.

4 These figures are derived from Fiskaa 1935 and do not accord with our definition of 'newspaper'. Fiskaa does not specify frequency, so that Figure 1 cannot be compared with our data.

5 The rate of growth for copies delivered may be even more dramatic than indicated, because relatively more newspapers are distributed by publishers themselves today. Today about 40 % of copies of newspapers are mailed. The general rule is that dailies do not use the postal service for distribution within the postal district within which they are published.

6 Verdens Gang, No. 9, 1883.

7 To decide on the number of newspapers and places of publication for 1885, we followed the procedure described in paragraph 3.1. From the files of the University Library of Oslo we obtained two additional papers launched just before the election. By this source we also discovered 30 items which did not conform to our standard definition of 'newspaper' and which were not included in our sample. Only one of these items, however, could influence the results obtained by coding 'regular' newspapers. In Horten there were two newspapers in 1885, one of them publishing once a week. Commitments in this paper were coded, but as it conformed to H2, and the town to both HI and H2, we decided to keep our definition. By following our procedure we also cut out 13 places of publication with altogether 15 newspapers, all of them having the frequency of one issue per week. The other 14 items found, appeared at places with more than two newspapers as defined by us, and were probably not regular newspapers by any standards.

8 With its final victory over the old regime the Liberal party proceeded to extend the franchise to male taxpayers in 1884. Under the income criterion, the number of qualified voters rose by 25 per cent from 1882. The new regulations had their strongest impact on the electorates of the larger cities which almost doubled, compared to an expansion of less than a quarter for the rest of the country. Qualified voters had still to register by giving an oath to the constitution, and a slight drop in the registration level indicates that the newly enfranchised voters turned out less frequently. Overall turnout rates at the election, however, dropped by less than two percentage points from 1882. An estimation of the composition of the 1885 electorate, basedon a special tax census carried out by the Central Bureau of Statistics, shows the following distribution of cities: 11.9% of voters belonged to the administrative establishment — the core of conservative resistance; another 22.4 % belonged to the upper-middle to upper class of businessmen and related professions; 36.5 % may be classified as belonging to the petite bourgeoisie, while workers made up 27.6 % of the city electorate. In rural districts the distribution was as follows: the administrative and business elites made up 5.4 % of the electorate, 75.2 % of it was farmers, while 11.0 % was workers (Rokkan 1967 and 1966). All told, the electorate was probably a cross-section of existing and potential audiences for newspapers by 1885.

9 Oslo provides us with an interesting case in this respect. The first Norwegian newspaper which still published here in 1885 remained neutral beyond the tide of rising political tension. In the meantime the next four establishments which survived until 1885 came out in favour of the Conservatives. In turn the first establishment was faced with a fait accompli, and the conservative electorate was well furnished with sympathetic newspapers. Eventually the newspaper decided to become Liberal after 1885.

9a The main source for this paragraph has been the annual reports of the Labour party. I am also indebted to Johan Ona of the Norwegian Labour Press Ltd. for valuable information.

10 Manhood suffrage was decided on in 1898. It has been estimated that the turnout rate for newly enfranchised voters in 1900 was about 44 %, and 52 % in 1903, in the cities, while corresponding figures for rural areas were about 33 % and 28 %. (Stein Rokkan 1966).

11 Quoted from Verdens Gang, No. 9, 1883. As our data from 1885 show, his assessment of the situation was somewhat optimistic. However, we may take some precautions as to newspapers with less than two issues per week. If he is correct, on the other hand, this indicates that the Liberal newspapers first came out with a commitment.

12 "Kristiania Høire", no. 7, 1924.

13 'First copy price' is equivalent to production costs for each edition. Through subscription this is partly paid for in advance by the readers.

14 The component data are: amount of investments and running production costs as related to single copy prices and minimum circulations. On neither of these points are we able to present data. Accounting statistics and data on circulation are only available for the post-war period.

---

15 The assumed correlation between circulation and weights of volumes is basedon the idea that newspapers behave like normal competitors and improve their product along with increased incomes, or anticipate increased incomes by improving their news-service. A fair amount of these improvements may be registered by more pages in the average edition, which adds to the weights of volumes.

16 Table 14 is adapted from Torsvik 1968, with minor alterations. His number of newspapers does not correspond with ours.

17 The correlation between circulation arid weights of volumes can be distorted by two mechanisms in the market. I) The volume of advertising, which may be a means of subsidizing - for example, big business may only advertise in bourgeois newspapers, or advertizing may be distributed according to the assumed buying-power of the audience, which does not always correspond to the circulation of a newspaper. II) The climate of competition on local markets will tempt newspapers to become alike in their news-service.

18 Circulation for 1940 is confirmed by Werner Erichsen who says that circulation increased 105 % between 1939 and 1950, when it reached 1,513,000 (Werner Erichsen 1960).

19 Computed from data of theAssociation.

20 Towns which we have been able to control for show two-thirds of market-leaders among dailies to conform to the rule of concentration. This computation was made for competitive towns on the basis of sales output over the period 1949 to 1965 with four sub-periods.

21 By the courtesy of Norwegian Advertising Statistics we have been able to compute per cent of advertising incomes as compared to per cent of circulation in a sample of altogether 109 newspapers for 1961 and 1965. For market-leaders among dailies the discrepancy between these percentages showed + 20.6 in 1961 and + 21.1 in 1965 in favour of advertising.

22 Commercial radio and TV does not exist in Norway.

23 Data given by the courtesy of the Association of Authorized Advertising Agencies show a turnover totalling 16.1 million kr. in 1949 compared to 292.5 million kr. in 1966. During this period the share of advertising capital channelled to newspapers went down 19.2 %, from 66.6 % in 1949 to 47.4 % in 1966.

24 The telephone and telegraph are public utilities in Norway.

25 A 'newspaper' was defined as: 1) a publication with a general news coverage, that 2) had a minimum frequency of two editions per week, and 3) was member of the Norwegian Newspaper Association. The last of the three criteria was not explicitly mentioned, but was nevertheless a rule for the application of the definition. Subsidiaries, on the other hand, were not ruled out by this definition.

26 The smallest administrative unit in Norway.

27 Per data from the Institute for Pressl966.

28 This paper ("Varden") is today a member of the Conservative press agency.

29 All of the bipartisan newspapers are bourgeois. Bipartisanship is arranged so that one member of the staff serves as editor-in-chief while one or two others serve as political editors, one for each party.

30 The discrepancy between votes and circulation for market-leaders was demonstrated on the basis of aggregate data. Lack of space does riot permit us to present the necessary amount of Tables to prove this point. The same goes for the complementary proposition that nonmarket leaders have a comparably more homogeneous audience. This was demonstrated with data from a community survey in the Stavanger area. See Valen and Katz 1964.

31 Journalists generally seek jobs within the same party press. However, some transitions between socialist and non-socialist newspapers are found. Among those changing jobs, 9 % had crossed this barrier in the 19605, compared to 6 % in the 1950s and only 4 % in the 19305. As may be guessed, changeovers of this kind are seldom possible among high-ranking journalists. Political activity is closely related to positions within the editorial board. This amounts to a requirement that political journalists become active in the parties at some point in their career before they go in to become editors (Werner 1966).

REFERENCES

Central Bureau of Statistics, Norges økonomi etter krigen. (The Norwegian Post-War Economy.)
80S no. 12, Oslo 1965.

Central Bureau of Statistics, Langtidslinjer i norsk økonomi 1865-1960. (Trends in Norwegian
Economy 1865-1960.) SØS no. 16, Oslo 1966.

Christensen, Chr. A. R., Fra "Tiden" i 1814 til vår tids presse. (From "Tiden" in 1814
to the Press of Today) in: Dette er Norge, Vol. 3, pp. 373-387. Gyldendal, Oslo, 1964.

Commission on the Press, see "Stat og Presse".

Diesen, Emil, Norske aviser og tidsskrifter med fortegnelse over norske pressefolk. (Norwegian
Newspapers and Periodicals.) Norsk industri og næringshåndbok, Vol. 11 one,
Kristiania 1920.

Fasting, Kaare, Den gamle by og den unge boktrykker. (The Old Town and the Young
Printer.) Bergens Tidende og J. W. Eides Bogtrykkeri A.S., Bergen, 1964.

Fiskaa, Haakon, Den norske presse før 1850. (The Norwegian Press Before 1850.) in:
Daniel Grini (ed.), Norsk Bladeierforening 1910-1935. Oslo, 1935.

Grini, Daniel, (ed.) Norsk Bladeierforening 1910-1935. (The Norwegian Newspaper Proprietor's
Association 1910-1935.) Oslo, 1935.

Johansen, Per, Men — vi lever. Norsk presse gjennem krig og okkupasjon. (But - We Live.
The Norwegian Press During War and Occupation.) Gyldendal, Oslo, 1945.

Kirkvaag, Olav, Arbeiderpressen i Norge. (The Labour Press in Norway.) Oslo, 1935.

Kaartvedt, Alf, Kampen mot parlamentarisme 1880-1884. (The Fight Against Parlamentarism
1880-1884.) Universitetsforlaget, Oslo, 1956.

Myklebost, Hallstein, Norges tettbygde steder 1875-1950. (Urban Settlements in Norway
1875-1950.) Universitetsforlaget, Oslo, 1960.

Rokkan, Stein and Torsvik, Per, Der Wåhler, Der Leser und die Parteipresse. Eine Analyse
iiber politische Einstellung und Lesen von Zeitungen in Norwegen. Kolner Zeitschrift
fiir Soziologie und Sozialpsychologie, 12. Jahrgang, heft 2, 1960.

Rokkan, Stein and Valen, Henry, Regional Contrasts in Norwegian Politics: A Review of Data from Official Statistics and from Sample Surveys. In: Erik Allart and Yrjo Littunen (eds.), Cleavages, Ideologies and Party Systems, pp. 162-240. The Academic Bookstore, Helsinki, 1964.

Rokkan, Stein, Electoral Mobilization, Party Competition, and National Integration. In:
Joseph LaPalombara and Myron Weiner (eds.), Political Parties and Political Development,
pp. 241—266. Princeton University Press, 1966.

Rokkan, Stein, Geography, Religion and Social Class: Crosscutting Cleavages in Norwegian
Politics. In Seymor M. Lipset and Stein Rokkan (eds.), Party Systems and Voter Alignments.
Cross-National Perspectives, pp. 367-444. The Free Press, New York, 1967.

"Stat og Presse" (Report from Commission on the Press), Innstilling om tiltak for å opprettholde
en differensiert dagspresse, (Recommendations on How to Keep a Differentiated
Press.) Orkanger, 1967.

Tingsten, Herbert, Pressen. (The Press.) In: Hal Koch and Alf Ross, Nordisk Demokrati.
Westermanns Forlag, Copenhagen, 1949.

Thomsen, Niels, Partipressen. En undersøgelse af danske dagblades politiske bindinger.
(The Party Press. An Investigation into the Party Bias of Danish Dailies.) Institut for
Presseforskning og Samtidshistorie, Århus, 1965. (See also his article in this volume).

Torgersen, Ulf, The trend Toward Political Consensus: The Case of Norway. In: Stein
Rokkan (ed.), Approaches to the Study of Political Participation. Chr. Michelsens Institutt,
Bergen, 1962.

Torgersen, Ulf, The Formation of Parties in Norway: The Problem of Right-Left Differences.
In: Scandinavian Political Studies, pp. 43-68, Vol. 2, Helsinki, 1967.

Torsvik, Per and Valen, Henry, Økingen i valgdeltagelsen i 1963 og i 1965. (The Increase
in Turnout Rates at the Elections of 1963 and 1965.) pp. 187-219. Tidsskrift for Samfunnsforskning,
Vol. 8, no. 2/3, 1967.

Torsvik, Per, Massemedia i Norge. (Mass Media in Norway.) In: Natalie Ramsøy ed., Det
norske samfunnet. Gyldendal, Oslo, 1968.

University Library of Oslo, Norske Aviser 1763-1920. (An Annotated Bibliography of Norwegian
Newspapers 1763-1920.) Universitetsbibliotekets årbok for 1923, Christiania, 1923.

Utheim, John, Statistik over Storthingsvalgene. (Statistical Review over The Elections of
1882 and 1885.) Christiania, 1886.

Valen, Henry and Katz, Daniel, Political Parties in Norway, A Community Study. Universitetsforlaget,
Oslo, 1964.

Venstres Presselag 1894-1919. (The Association for Liberal Editors and Journalists 1894-1919.
A 25th aniversary report.) Christiania, 1919. (Author unknown.)

Westerståhl, Jørgen and Janson, Carl-Gunnar, Politisk Press, Studier til belysning av dagspressens
politiska roll i Sverige. (The political Press, Studies Illustrating the Political
Role of the Daily Press in Sweden.) Statsvetenskapliga Institutionen, Goteborg, 1958.

Werner, Anita, Norske Journalister. (Norwegian Journalists.) Tidsskrift for Samfunnsforskning,
Vol. 7, no. 1, 1966.

Werner Erichsen, Rolv, Samarbeide og konkurranse i norsk presse gjennom femti år. (Cooperation
and Competition in the Norwegian Press over Fifty Years.) J. W. Eides Boktrykkeri,
Bergen, 1960.