Stockholm

Introduction

A number of introductory investigations of the Local Government Research Group in Sweden have been focused upon the possibilities of citizens for obtaining informationl l concerning activities of local government authorities and local parties in their own districts.² In order for democracy in the traditional sense to have any true significance in the communes it is necessary for the citizens to be well informed about what decisions have been made in major questions, which persons and parties have advocated alternative solutions to essential communal problems, etc.

This information is communicated to the citizens via several different channels. The meetings of the communal (municipal-) councils are public, the communes are required by law to post certain announcements on the communal notice-board, they announce in the daily press, they can issue local information sheets, arrange information meetings, etc. In addition to this official activity, the representatives and officials communicate directly with the people at party meetings, in their circles of acquaintances, in the family, etc. In the very smallest communes the lastmentioned channels might even be the most extensive. However, with the exception of the very smallest communes the local press must be the most important channel for the distribution of this kind of information. Of course, the average citizen does not know very many representatives. He is not a member of any party organization but is probably dependent upon the contents of the mass media (primarily the local press).

The contributions of the local press in the local political debate and news distribution have been actualized in two connections in Sweden during the last few years. The press has been criticized for taking too small a part in this debate and for handling news distribution concerning communal questions in a routine manner. The communal contents have been said to have consisted of, to a large extent, reports from communal meetings and, to an altogether too small extent, of penetration and critical scrutiny of questions in preparatory stages.³ Lars Furhoff has investigated in a debate tract the frequencies of editorials with communal political contents in a number of rural newspapers and has found it to be exceedingly low.⁴

The Swedish newspaper market is at the present going through the same rapid
structural changes (commonly called "newspaper death") as the press in many

other countries. To date, this has primarily meant that newspapers in so-called
second positions (i. e., the smaller newspaper - the smaller newspapers - at a place
of publication) have been discontinued.s

This development has been watched with concern as regards the formation of political opinion and the dissemination of local political news. A state commission presented a proposal in 1965 for state support for newspapers and a new commission is working at the present with these problems, among others.6 However, an extensive scientific investigation of how the press actually fulfils at the present its news service and the formation of public opinion on the local level has not been made.

In 1966, the Local government research group made a content analysis of the newspapers which were circulated within the sample of communes of the group - primarily as a stage in the efforts to obtain knowledge about which factors govern the possibilities of the citizen to become informed about communal activity in his own district.

When the investigation was expanded to include the local political and communal contents from all communes within these newspapers' areas of circulation it is apparent that the result is also interesting from the more general framing of the question concerning the way the press fulfils its function in the political system that has been actualized by "newspaper death".

This article is a preliminary report on the central parts of this investigation. After an introductory discussion of the factors that have been assumed to influence the scope of the press information from different communes, an attempt will be made with the aid of simple cross-tables to illuminate the importance of these factors. The technical problems with coding and machines will be discussed only to the extent that they have direct importance for the results that are stated in this article.⁷

Conceptual Framework

Introduction

The extent of contents in a newspaper from a given commune can be said to be dependent upon the editorial appraisals that have been made about the newspaper editorial staff during the period studied. The result of these considerations is of course influenced by the purpose and general nature of the type of newspaper. The local press gives reasonably greater priority to local news than the national press, etc. Moreover, the contents from an individual commune is, however, dependent upon some properties of the commune. It is apparent that the production of news, i. e., the number of occurring events of such a general interest that they also interest individuals besides those directly concerned, varies among different communes. It can be quite generally assumed that there is more for a newspaper to keep an eye on in a large commune than in a small one. The competition on the newspaper market may also be of importance for the newspaper's interest in an individual commune. The communes where the number of potential readers is large or where the newspaper is exposed to competition from another newspaper ought to be subjected to a more intensive surveillance than the communes where the newspaper holds a completely dominating position on the market. Henceforth, we can call these latter

conditions the newspaper situation of the commune. The purpose of this study is to illuminate the importance of these three explanatory factors for the amount of information about the political and administrative activity within his own commune that a citizen is provided access to via his newspaper. These three factors will be more thoroughly discussed separately later. Figure 1 shows their mutual interrelation.

General Assumptions

The following reflections are based on a few assumptions that have been taken as
granted in this study.

1. The communal contents⁸ can be regarded as a part of the local contents in the newspaper and has a reading value which does not fall below that of the local contents in general. Therefore, reflections in the future concerning local contents are valid for communal contents too.

2. Local contents have, as a whole, a high reading value in local and regional
press. By local contents we mean in this study contents from places within the circulation
area of the newspaper.

3. The local news is relatively costly for the newspapers in comparison with
national and foreign news since they demand greater contributions in order to be
acquired.

4. The newspapers are regarded in the first instance as businesses, the primary purpose of which is the presentation of satisfactory economic results. In the case of certain newspapers owned or subsidized by organizations the primary goal can instead be assumed to be access to a channel for the formation of public opinion. However, in the present situation the demand for satisfactory economic results is placed even on these newspapers, although this demand is formulated rather as the

minimalizing of losses than as the maximalizing of profits. This economic goal means
that the newspaper strives after an advantageous relation between income and
expenses.

5. The demand for daily distribution (influencing costs) and the existence of a local advertising market (influencing income) make it economically more advantageous for a newspaper to have high circulation in a restricted area than to have the same total number of copies spread over a wider area. A newspaper that wants to improve or maintain its economic result ought consequently, first, to consolidate its position within its area of circulation and, only secondly, to try to expand this area. An attempt to expand the area of circulation more substantially will probably not, in reality, be economically rewarding until a large part of the local advertising market has been won over from the competitor.9

Types of Newspapers

a) publication frequency

The Swedish newspaper market can not be regarded as a homogeneous market containing the same sort of products. The investigations of the press commission concerning double-reading of daily newspapers indicate that reading of the small non-daily newspaper is often supplemented by other provincial papers or a metropolitan paper.10 In general, the non-daily newspapers have restricted areas of circulation and are strongly locally orientated. It is reasonable to presume that they give greater priority to local contents in relation to other contents than do the papers which are issued daily and compose to a large extent the only paper of their readers. However, at the same time the non-daily papers owing to the lower publication frequency have less available space for text than the larger ones. The non-daily papers form a special group in this study.

It is difficult to get any idea in advance about how these conditions in conjunction with each other influence the extent of communal contents from the individual communes, since they have effect in different quarters. Nevertheless, the adjustment of the contents of the newspaper to the communities within its area of circulation is the most important competitive means of these papers in relation to others. Although reading value of the local advertisements' must not be underestimated, an important incitament for the readers to take in the long run the extra expense that double-reading causes disappears if the non-daily paper does not contain more local news than does the newspaper that is read parallelly.

b) Competition at the places of publication

It has been presumed earlier that local news has a high reading value at the same time as it is expensive for the newspapers to acquire such news. Additionally, we have stated that it is economically important, to an unusually high degree, for the newspapers to obtain a large share of the market. Newspapers which are subjected to competition within their areas of circulation (especially from their own place of publication) have, therefore, greater motivation for using this expensive contents. Newspapers with a large share of the market within their areas of circulation can

instead, to the extent technical resources permit a larger number of copies, possibly
concentrate on an expansion of their areas of circulation.l2

Therefore, the newspapers which are issued six or seven days per week in the sample have been divided into two groups as regards the competitive situation at the place of publication. The newspapers which are subjected to competition are assumed to devote greater space to local news than other papers which we shall hereafter call "mono-newspapers".13

The fact that local contents is expensive for the newspapers means that newspapers with good economic resources have greater possibilities for using this as a means of competition. The newspapers which are subjected to competition within their areas of circulation have therefore been separated into two groups, first-papers and second-papers.

c) the size of the area of circulation

As has been previously mentioned, the trend of development indicates that we are getting more and more newspapers of a regional nature. These newspapers have more communes within their areas of circulation at the same time as the space for surveillance of local affairs can not be increased beyond certain limits for economic and technical reasons. If this means that each individual commune is to be devoted smaller space in newspapers with greater areas of circulation, this trend leads on the average to a decrease of the extent of local information in the press. In the analysis attention will be drawn to these conditions by our controlling of the size of the area of circulation within the groups of six- and seven days per week newspapers.

d) branch editions

A number of "branches", i. e. editions with their own pieces are also included in the investigation.14 The presence of a branch edition involves an adjustment of the contents to a special part of the newspaper's market. The branch editions ought, therefore, to have more contents from the communes within their areas of circulation than other groups of newspapers.

e) Metropolitan newspapers

The study also includes a number of metropolitan newspapers (papers published in Stockholm, Gothenburg or Malmo). These newspapers are local newspapers not only for the municipalities where they are published but also for a relatively extensive number of communes near to them.¹⁵

It is compatible with our earlier argument to assume that the communes in these areas are devoted less space in the press than communes covered by other newspapers, since the metropolitan newspapers fulfil simultaneously a sort of national press function. This condition ought to be more pronounced in the Stockholm newspaper area where the competition during the last few years has brought about several local supplements as well as harder investments in local one day per week newspapers as supplements to the daily editions.

The Production of News

It is difficult to dissect the concept "production of news". It contains both a

quantitative and a qualitative dimension. It is partly a question of the number of events within a certain area - in this case a commune -, and it is partly a matter of how many of these events are sufficiently news worthy, i. e., are considered to be of such a great interest for the newspaper's readers within or outside of the area that they are interested in reading about the event in question.

It is apparent that a very large part of the routine work that is carried on within the communal administrations is not of the nature that it can be judged as having news value. A large part of the work within health and welfare care should probably have a very low news value.

It can be considered self-evident that a correlation must exist between the population of the communes and the production of news. The more inhabitants a commune has the greater activity that is required on the part of the communal institutions. However, this correlation ought hardly to be linear since a large part of the major activity in a large commune is of a routine nature (more persons are affected by the same decisions and more identical decisions or actions occur).

Similarly, it is reasonable to assume that the larger the percentage of the population that lives in highly populated areas, the greater the average demand per individual will be on communal activities, since the special urban demands are asserted to a greater extent.

The differences in communal activity that can be related to differences in the number of inhabitants or different degrees of urban density involve primarily the quantitative dimension of the news production concept. Nevertheless, in certain communes, primarily those which are rapidly expanding, a large part of the communal actions ought to have a higher news value. New housing areas are being planned, a long awaited lower school replaces a temporary one, etc.

Purely theoretically, it is possible of course to measure the amount of work performed in the communal administration, for example, by the number of work hours or by the number of decisions made etc. The qualitative dimension would,, above all, owing to its subjective nature, be considerably more difficult to measure. Of course, for practical reasons it has neither been possible to investigate how much has been done in communal administrations for over 800 communes, but several different operationalisations have been used in the investigation which indirectly indicate a high news production. Unfortunately, it has not been possible to find one which is completely satisfactory and the various operationalisations yield partly different results.

The Newspaper Situation of the Commune

Consideration was taken of competition at the place of publication when the newspaper typology was established. This competition was assumed to lead to the intensification of local surveillance on the part of the newspapers. Competition between newspapers from differecent publication places (the areas of circulation of the newspapers meet) ought reasonably to produce the same consequences in the actual communes of competition. The presence of several newspapers from different areas of circulation ought, thus, to be regarded in the analysis as one of the properties

of the communes that influence only the newspaper's handling of precisely these
communes.16

Conceptual Framework: Summary

Variations in the extent of the surveillance of the press from different communes have been assumed to lie to a large extent in the following three factors: the classification according to type of the existing newspaper(s), the production of news and the newspaper situation within the commune. Since the surveillance is a result of considerations on the part of the newspaper editorial staff and since doublereading of rural newspapers is relatively infrequent, it is practical and reasonable in the study of the importance of these three factors to investigate how much communal contents an individual newspaper contains from each separate commune. In order to avoid an ambiguous use of language we will from this point on talk about both "communes" and "control communes". With the concept "Control commune" we mean a commune which is situated inside one of the paper's area of circulation. As the papers' areas of circulation are to some extent overlapping each other, a commune where two newspapers are circulated generates two control communes. The relationship between the concepts of control commune and commune is illustrated by Figure 2.

The circles are illustrating the scope of the areas of circulation of the newspapers A and B respectively. Commune I is situated inside B's area of circulation and commune IV inside the one of A. Each of these two communes occurs in the analyses as one control commune. But the commune II and 111 are situated inside the areas of circulation of both A and B and therefore each of the communes occurs in the analyses as two control communes.17 If one is interested in measuring the total amount of press information from one commune, one must weight together what all papers existing in the commune have written about the matter.¹⁸ This type of analysis has not been made in this study, rather the goal has been limited to the

problem of understanding the variations in the individual newspaper's surveillance of the communes within their areas of circulation. The significance of this restriction is illuminated by the result of the Press Commission's investigation concerning the reading of one or more newspapers. Only 8 % of newspaper readers read several rural newspapers according to these figures.19

The Sample of Newspapers and the Time Period

This study has been carried out with the help of the newspapers that are included in the newspaper sample of the Local government research group. This sample consists of all newspapers issued a minimum of once a week²⁰ and with circulation in the research group's sample of communes.21 Since the communes in this sample are spread all over the country the sample of newspapers has also been nationwide.

The newspapers in our sample correspond to 77 % of total newspaper publication in the country. Table 1 shows the division of these into different groups of papers. As is evident from Table 1, the newspapers which are included in the study have on the average approximately twice as large a circulation as other papers.22

In reality, we have investigated the larger rural newspapers with 3-4 days and 6-7 days per week publication,²² respectively. However, possibilities for analyzing the importance of the size of the area of circulation remain, since even within the newspaper sample large differences exist in this respect.²³

An explorative study

It is evident from the account of the newspaper sample that the investigation as a whole must be regarded as explorative. It is not possible to generalize from the result of this investigation to the Swedish newspaper population as a whole. This limitation depends, however, only on the construction of the sample. The conceptual framework that has formed the base for the analysis seems to be sufficiently explicit in itself and coherent in order to make possible the formation of an hypothesis. The possibilities for establishing, on the basis of the investigation, new hypotheses concerning the surveillance of the local commune in the Swedish press, which will later actually be verified, will probably, therefore, be able to be judged as good.

The time

All newspapers were examined during the period Ist January-31st March 1966. This was a period with the usual amount of political activity characteristic of periods between elections. Since the proportion of communal news increases in periods close to elections this was a fairly important point. The selection of the time period was also guided by our desire to obtain the possibility for comparing the results from a survey of citizens that was made in September 1966 in connection with the communal and municipal elections.24

Coding

In all newspaper copies from this period the communal contents have been scanned and classified. In order for an article to be classified as a communal one it was necessary for it to have a connection with the communal institutions (or the local parties) or their activity in a certain commune(s). Contents which could not be connected to a certain commune, for example articles concerning the role of communes in national planning, were not codified. The definition of local communal questions is fairly broad. Communal activities of a service nature like snow-clearance have, for example, been classified as communal to the extent that it is clearly evident that the work has been done under the management of the commune. The principal rule was that it should be evident from the article that the commune was involved in some way.^24a All articles which were classified as communal were measured as to length and were assigned to the commune involved. If the article concerned more than one commune it "was divided" into equal parts which were assigned to each commune mentioned in the article.

Moreover, it was noted in the case of each article whether it was placed on the
"front page," which subject it treated, whether it was a news article and whether
it was critical towards the communal institutions (negative tendency).

Article which treated more than one subject area was divided if every subject took more than five col. cms. Otherwise, the article was entered under the contents category "mixed". This category consisted primarily of references of meetings, small announcements cut from agendas or published documents, etc.

The categories news articles with and without interviews, can be regarded
as examples of active surveillance on the part of the newspapers. Articles which

have been coded into these categories were not allowed to contain reports from communal meetings and therefore these categories contain news articles from the communes that the newspapers have introduced without the question involved having been actualized by a meeting.

To obtain an idea about the nature of the newspaper's news surveillance an attempt was made to draw attention to articles with a clearly negative tendency. We must in this connection be very restrictive in classifying an article as negative if we desire to obtain satisfactory reliability. Therefore, it is possible that the percentage of the newspaper's communal surveillance that according to a general understanding should be classified as negative in this study has been underestimated. Nevertheless, there is no reason to assume that this underestimation, if it occurs, influences the result of comparisons between different newspapers or different communes ²⁵ A more detailed account of the coding work and the contents categories is given in the research group's publication series no. 11.

The contents from the communes that the newspapers give priority to ought to be classified to a large extent as news articles with or without interviews in addition to the fact that they are exhaustive to a large extent simultaneously as they have been classified to a small extent as "mixed".

Findings

Introduction

The report on the results begins with an account of the surveillance from various control communes without taking into consideration factors other than communal size. Thereafter, the explanatory factors were successively introduced into the analysis. First, news production was analyzed, next, the type of newspaper and finally, the newspaper situation. However, this systematism was interrupted by the fact that the attribute of being a place of publication, which can be connected primarily to the commune's newspaper situation owing to its strong manifestation on the scope of the contents, was introduced into the analysis already directly after news production.

Communal contents from all communes

Material is included in the following from all communes where newspapers in the sample are circulated in a minimum of 5 % of all households. However, the metropolitan newspapers are not included since information is lacking about the circulation in households for these in each separate commune. 1268 control communes remain,

Table 2. shows how these are distributed into different classes according to size and also shows the average number of column cms. the number of column cms. without advertisements and the number of column cms. of news articles with or without interviews. The total number of column centimeters is a measurement of the amount of information that a reader of the newspaper can maximally acquire in this manner. Thus, this information is primarily interesting from a functional point of view when one studies the extent of varous information channels in the

commune. On the other hand, the column cms. of non-advertisements is a measurement of the space that the editorial staff of the newspaper has chosen to devote to the respective commune, and is, therefore, of greater interest in the analysis in connection with the conceptual framework that has been discussed. The column cms. of news articles are, as has been mentioned previously, to be regarded as a more ambitious form of surveillance that occurs to a larger extent from communes that the newspaper subject to a more active surveillance. Therefore, upon analysis a study of the amount of column cms. of news articles ought to yield greater variations than the total amount of col.cms. of non-advertisements.

The large variations in contents between control communes of different sizes are immediately evident from Table 2. These variations are so large that the mean for the entire population becomes fairly non-revealing. The central values in table 2 in the column for all control communes are in reality fairly misleading owing to the central measurement employed.26

For these reasons statistics will from here on be thoroughly accounted for in the
case of the control communes, partly in total and partly divided into the various
classes according to size.27

The large variations in col.cms. from control communes in different size-classes that are evident from Table 2 must reasonably be to a very large extent a result of dissimilarities in news production. Although considerably differences should be expected between the largest and the smallest control communes, the differences are, nevertheless, strikingly large according to Table 2. The communes in the largest class have been devoted 26 times more space in the newspapers than the smallest communes. A part of the difference can, however, be explained by the co-variation between news production and any of the other explanatory factors, especially the attribute of being a place of publication. Thus, it is proper to study the relation between different "surveillance variables" and the operationalisations of the production of news, and afterwards to investigate possible differences between the places of publication and other communes.

News Production

We have already stated that the news production concept is difficult to perform

analyses with. In the attempts to find a suitable operationalisation four variables
have been studied.

1. The number of inhabitants on January l, 196628

2. The percentage of the population living in densely populated areas on January
1, 196029

3. The development of the population from January 1, 1960 to January 1. 1966.30
4. The number of communal civil servants on Ist August and Ist October 196531

It was assumed that the first three together reflect both the quantitative and qualitative dimensions of the concept, while the fourth through its more direct relationship with work performed within the communal administration was assumed to correspond in itself to the variables 1 and 2.

An investigation of the mutual relationship of these four variables shows a high
co-variation. The linear correlation between the variables (the product moment
correlation coefficients) can be read from the correlation matrix in table 3.

The linear correlation between the number of inhabitants and the number of civil servants in the communes is, consequently, very high. The other relations are weaker but obvious. Thus, if from this point on we use only the number of inhabitants or the number of civil servants in the analysis, a part of the "effects" we are measuring can instead be "effects" of differences in population development or degree of urban density. However, the correlations are too high for permitting, with the use of the contingency method in the future, the holding constant of more than one of the news production variables at the same time.

If the higher figures for communal contents that we found in table 2 only reflect differences in news production, the differences between control communes in different size classes ought to disappear upon control for differences in size - the number of inhabitants - and the number of communal civil servants, respectively.

It can be seen from Table 4 that the differences disappear, by and large, upon measurement of all contents without advertisements per 100 inhabitants. However, the news articles still show a weak correlation with the size of the population of the commune. Upon use of the number of civil servants as an operationalisation of news productions we get a steadily falling trend. Theorectically, the explanation for this can be that the large communes are watched least when controlled for dissimilarities in news production. However, it appears more realistic to assume that the number of civil servants is a good measurement of the quantity of communal activity,

but reflects to far too small an extent the qualitative dimension of the concept. The smallest control communes were devoted 61% of the equivalent amount of communal contents from the largest control communes per 100 inhabitants. The differences between the two extreme groups is about 25% when all material except advertisements is measured, but a relationship to size can not, nevertheless, be considered existing when the largest size class differs pronouncedly from the others.

The percentage of communal contents that has been categorized as news article
is evident from diagram 1 which shows a higher percentage in larger communes.

The tendency in diagram 1 that a larger share of the contents from the larger control communes have been classified as news articles indicates that the larger communes are devoted a more active interest on the part of the newspapers than the smaller ones. If such is the case, the contents from the larger communes ought also to be placed to a greater extent in prominent places in the newspapers.

It is evident from diagram 2 that the percentage of contents from various control communes that has been placed on the "front page" co-varies with the size of the commune. In the lowest size-class only 6% have been given such prominent placing, on the other hand in the largest commune it is 20%.

From diagram 3 it is evident that the same tendency can be traced concerning the
percentage of contents that has been classified as negative. However, the tendency
is not consistent and the differences are small.

A comparison between the extreme groups shows, however, that more than twice as large a percentage of the communal contents has a negative tendency in the largest group. These percentages ought to be seen against the background of the large difference in space, viewed absolutely. Even if we have had very narrow criteria for negative tendency regarding reliability, it is also evident from the diagram that the negative news surveillance towards the commune is of a small extent in most communes.

News Production: summary

By way of summary it can be stated that the newspapers included in the study contain more contents from larger communes than from smaller ones. When consideration is taken to differences in news production this relations disappears, however. The differences remain for the more qualified news material (news articles with and without interviews). The percentage of this material increases with a rise in communal size, as does the percentage of material that has been given prominent placement in the newspaper. The percentage of contents that has been given a negative tendency is also larger from large communes than from small ones.

All these conditions indicate that the large communes are devoted greater interest on the part of the newspapers than the smaller ones even when consideration is taken of differences in newsproduction. However, it is difficult to imagine that the size of the communes as such can be the cause for this greater interest.

The total field of control communes that have been studied to date is composed of both places of publication and other communes, of communes covered by threeand six days per week newspapers, respectively, of communes where the newspaper studied is the dominant one on the market, of communes where the newspaper studied has a very low share of the market, etc.

The results we have obtained to date can, consequently, be the result of covariation
with another of the explanatory factors. Therefore, we shall next study
the distribution of the places of publication in the total field of control communes.

Place of Publication

A passing knowledge of the situation of the Swedish press already tells us that most of the control communes in the higher size-classes must be places of publication. As regards the dominating importance of the places of publication for the newspapers it is natural to interpret the result stated above in such a manner that the larger communes are more activity under surveillance in the role of places of publication rather than in the role of larger communes. Before we study other attributes of the newspapers we shall, therefore, divide the total field of survey of control communes into two groups: places of publication and others.

As table 5 shows, we have very limited possibilities for discerning "the effect" of
communal size independently of the attribute of being a place of publication. Within
the size-classes 8,000-14,999 and 15,000-29,999 inhabitants we have, however,

sufficiently enough control communes for making comparisons between places of
publication and other control communes.

The predominant significance of the places of publication for newspapers is evident from the fact that in the studied size-classes 30% and 40% on the average, respectively, of the copies of the paper are sold in these areas, while the equivalent percentages for other communes are 5% and 7% respectively.

It is evident from table 6a that the places of publication have received 3 and 2.2 times, respectively, as many column cms. of communal contents as other localities in the same size-class. Now, the class width is large and the number of communes relatively low, therefore these differences can possibly be "explained" by dissimilarities in communal size within the size-classes.

However, as is evident from table 6b the differences remain in the same manner
as earlier when a control of communal, size is made.

It is also evident from table 6a that the larger control communes in the actual size interval do not receive much more space than the smaller ones. In reality, lower values for the larger communes are obtained when we look at space per 100 inhabitants. If we resume the earlier discussion of news production, this is, however, rather natural. An increasing number of inhabitants must, if consideration is taken of not only the number of incidents having occurred but also of the news value of these, have a diminishingly marginal effect on news production.

It was stated earlier that the contents from the control communes in the higher size-classes not only were more extensive but also that they indicated a more qualified surveillance on the part of the newspaper. Then, is even this an effect of the concentration of the places of publication to higher size-classes?

All results in tables 7a-d strengthen the assumption that the places of publication are subjected to a more active surveillance than other communes in the same sizeclasses. A larger part of the contents from the places of publication has been classified as negative; more have been placed in a prominent place in the newspaper simultaneously as the surveillance has been regarded to a larger degree as qualified. The measurement of the share of "routine surveillance" of all material that is recorded in 7d shows that even this plays a smaller role from the places of publications than from other communes.

It is also evident from tables 7a-d, with the exception of the percentage of content
with a negative tendency, that the previously found relationship with size remains.

Even if it is not possible to compare more strictly the differences regarding
communal size with those regarding places of publication, it is immediately apparent,
however, that the latter comparison gives considerably greater result.

Place of Publication

Summary

By way of summary it can be stated that a comparison between places of publication and other control communes in the size interval where a comparison i possible shows differences which are consistently substantially larger than the mean difference between the various size-classes. A citizen in a commune that is covered by this investigation receives, if the commune is a place of publication for a local or regional newspaper, access to three times as extensive communal contents in the press from his home commune than the citizen who lives in another commune of an equivalent size. Moreover, this much more extensive material is placed to a greater extent in a prominent place in the newspaper, contains to a greater extent observations on unsatisfactory state of things and is to a greater extent of the more interestarousing nature than news articles with or without interviews can be said to be.

The differences stated here between places of publication and other control communes are so large that they outweigh "the effect" of both other newspaper conditions and dissimilarities in news production. Therefore, only other control communes, which compose 96% of all such in the investigation, will be included in the continued analysis.

Newspaper with Different Publication Frequency

The control communes which we have studied upto now have come from newspapers
with different publication frequencies. As has been evident from the account
of the sample of newspapers, the publication frequency of the newspapers varies

from one day per week to seven days per week. The newspapers which are issued more often have, viewed absolutely, greater possibilities to introduce material from the communes within their areas of circulation during the period under investigation. However, it does not necessarily follow from this that the control communes from these newspapers have higher values on the average than control communes from newspapers with lower publication frequencies. Newspapers with lower publication frequencies have, as a rule, less extensive areas of circulation and thereby fewer communes to watch. We shall, therefore, study in the beginning the control communes from newspapers with different publication frequencies and afterwards group them according to a more elaborate newspaper type classification where, among other things, consideration is taken of the size of the area of circulation.

It is evident from table 8 that the possibilities for comparing control communes from papers with different publication frequencies are relatively good within the two lower size-classes. In the others, by large, only six days per week newspapers occur in our sample. The following comparisons, therefore, only apply to control communes in the two lowest size-classes.

We find it completely natural from table 9 that the control communes from one day per week newspapers have very low values. These newspapers are not to be regarded as newspapers in the first instance. They are only read along side of another daily paper. They have the whole county (lan) as their area of circulation and have a small circulation.³² The results from these newspapers will be left

outside af the analysis from this point on. In the case of the other control communes we find that the groups 6-7 days per week newspapers have on the average a somewhat higher contents value. The differences between the newspaper groups are, however, a function of differences in communal structure. When a control of communal size is made these differences disappear. In the size class 4,000-7,999 inhabitants the value for the 4 days per week group is surprisingly nearly 300 cm. higher than the next group which is the 3 days per week newspapers. The values for four and seven days per week newspapers, respectively, are based, however, on only two newspapers within each group and must therefore be judged with the greatest caution. Individual distinctive attributes on the part of the individual newspapers can influence the result. Moreover, it is difficult to distinguish any tendencies in table 9. The groups break the order sequence between the two communal size classes. A control of the size within the size groups by comparing the values for column cm./100 inhabitants does not change the picture noticeably. The control communes from 3 and 7 days per week newspapers have, indeed, higher values than the six days per week group in the smallets size class, but the differences are very small.

Eventual "effects" of differences in publication frequency are counteracted by, from all appearances, either differences in the percentage of the newspaper's space that is devoted to communal content on the whole or by the differences concerning the number of communes to be watched.³³ Before we continue the investigation of the quantitative differences between the various control communes we shall, therefore, investigate whether the newspaper groups differ regarding the qualitative attributes we have registered in the contents.

No tendency concerning the percentage of contents that has been classified as negative or news articles, respectively, can be traced. On the other hand, tables lOa and lOb show differences between the newspaper groups that are consistent and large. The newspapers issued three days per week have in comparison with those issued six days per week placed approximately three times as large a percentage of contents from each control commune on the front page. The percentage of contents that has been classified as mixed show less differences between the groups, but even here the differences between the three-and six days per week newspaper groups occurs throughout.

Publication frequency: summary

By way of summary, we can state that newspapers with different publication frequencies
do not appear to differ as regards the quantity of contents on the average
from the communes within their areas of circulation.

The investigation of the qualitative properties of the contents confirms, however, the assumption that communal contents play a more important role in the non-daily newspapers. The three days per week newspapers give a larger percentage of communal contents from different communes a prominent placement at the same time as a smaller share has been classified as routine surveillance. Therefore, communal contents in these newspapers ought to be easier to observe by the reader and possibly also more interest-arousing.

Different Types of Newspapers

Earlier, it has been assumed that local contents are more costly for the newspaper to acquire, compared with national and foreign news and agency material. At the same time, the adaptation of the product that goes with surveillance of local activities involves an important competitive means on the Swedish newspaper market.34 Therefore, it can be said that newspapers which are subjected to competition within their area of circulation have greater motivation for staking in this costly material. At the same time, newspapers with good economies have greater possibilities for using local contents as a means of competition. If we now view the communal contents as a part of local contents, as a whole, this communal contents ought to ofliow the same pattern as the local contents in general, even if the more brief announcements pertaining to communal activity come to the newspaper without any major efforts on its own part35

The newspaper sample is considerably heterogeneous concerning the size of the areas of circulation. However, it is apparent that precisely the number of communes within the area of circulation must form an important limitation for a newspaper's possibilities for introducing news from each individual commune. Therefore, from this point on the control communes from newspapers with six-and seven days per week publication have been combined into one group which has afterwards been divided into subgroups in respect to the competitive position of the newspapers and the size of the areas of circulation. Also included in the analysis are three- and four days per week newspapers, merged into a group "non-daily", branch editions and metropolitan papers. The last mentioned group has not been included in the earlier calculations, but is introduced here as a comparison with the rural press since even the metropolitan newspapers form the local press in certain areas.

When the circulations statistics do not give an account of the metropolitan newspaper's circulation per commune, the content from the communes within the respective newspaper area is included in the investigation.36 The types of newspapers obtained in this manner are evident from the arrangement in table 11.

The papers within each competitive class have been ranked according to the area
of circulation and divided according to the median. Newspapers with more communes
to watch than the median number have large areas of circulations and newspapers

with less than the median number have small areas of circulation. As is evident from the table this mode of procedure led, nevertheless, to the fact that the areas of circulation which have been classified as large are, on the average, equally large within different groups. The co-variation between the publication frequency and the size of the area of circulation that has been assumed previously is also clearly evident. The non-daily newspapers have a lower number of communes to watch than other groups with the exception of the branch editions.

Thus, among the control communes from these different types of newspapers those from the first- and second-papers ought to obtain higher spatial values than those from mono-papers. A large area of circulation yields lower values for contents per control commune than a small one.

Since a special edition involves the most far-reaching adjustment of the product to a certain part of the market at the same time as this part of the market is normally relatively limited (14 communes), the control communes from the branch edition group ought to have higher values than all other groups.

Since the comparisons must still take place under control of communal size (the same class limits as used earlier), the possibilities for analysis decrease in the larger size-classes. If we accept cells with a minimum of 5 observations, we can make comparisons concerning the three lowest size classes. Lower frequencies than 10 exist, then, in the size class 8,000-14,999 inhabitants for the newspaper groups 5-8.37

It is evident from Table 12 that control communes from newspapers with small areas of circulation receive higher values in each size-class than control communes from newspapers with large areas of circulation. The average values are between approximately 20 % and 40 % higher for control communes from newspapers with small areas of circulation in comparable competitive situations. It is apparent that newspapers with larger areas of circulation, viewed totally, contain more communal content than newspapers with smaller areas of circulation. The larger space that is devoted in these papers to communal surveillance is not, however, sufficient to compensate for the higher number of communes to hold under surveillance.

The assumption that control communes from newspapers subjected to competition
at the places of publication should have higher content values than mono-newspapers

does not prove to be correct. A ranking of the mean column cms. with both communal size and the area of circulation constant (6 rankings in total) gives the "rank sums" 13, 13, 10 for first-papers, second-papers and mono-papers, respectively. Thus, it is rather the mono-papers which have, on the average, somewhat higher values. However, the differences are small and goes in different direction within different size-classes. Therefore, it is not possible, based on these results, to declare any relationship between the competitive position of the newspapers at the places of publication and the extent of contents from different control communes in the area of circulation. However, the value of these results is dubious since the places of publication themselves were not included in the calculations. The small number of observations made it impossible, however, to perform an analysis of these.

The assumed positive effect of the branch edition on communal information via the press is also clearly evident from Table 12. In the smallest size class the control communes in the branchedition group receive the highest mean value and in the other two size classes only one or two groups, accordingly, have higher mean values.

The table also shows the particularly low values from the control communes group of metropolitan newspapers. The information that a citizen in a suburban commune receives via the press is thus practically non-existent if he does not supplement his reading of the daily press with one of the special suburban papers.³⁸

When size within the size classes is controlled by dividing the column cm. values by hundreds of inhabitants the same picture is obtained for the entire group of non-places of publication as regards to the communal size variable, i. e. no relationship can be traced. On the other hand, the same relationship with the size of the newspaper's areas of circulation reoccurs. We have also investigated the percentage of material that has been classified as negative and that has been placed on "front pages". The tendency is the same as earlier, that is, the larger part of the contents is negative or well-placed, respectively, from larger control communes. Both groups of control communes from first-papers consistently receive the highest percentage of contents that have been classified as negative but the differences between

the different groups are small. Moreover, no thorough tendencies can be detected. The percentage that has been placed on the front page is largest for the group from non-daily newspapers. Next come the group of first-papers and then others without it being possible to detect any pattern of differences between them.

Type of Newspapers: summary

As has been assumed, large differences can be pointed out between control communes from newspapers with large areas of circulation and those with small ones when the control communes are divided according to newspaper type. The results hold when a control is made of news production (column cms. per 100 inhabitants).

It is not possible to point out any consistently quantitative differences in any size-class when the size of the area of circulation is held constant. The assumption that the competitive situation at the places of publication influences the extent of local communal contents in the newspapers have consequently not been confirmed in this investigation which does not, however, include the places of publications themselves.

The Newspaper Situation of the Commune

In this analysis two of the explanatory factors have been introduced heretofore. Consideration has been taken to the third, the commune's newspaper situation, by our having excluded, after a comparison between places of publication and other communes, places of publications from the further analysis. The newspaper's areas of circulation were limited in order that all communes where the newspaper had a circulation equivalent to at least 5 % of all households have been investigated. In this way the control communes which have been studied upto now have represented far different market situations. Communes have been included in the analysis in which the newspaper studied has had a dominating market position and communes have been included in which the newspaper has been upto the fourth in size (measured in the number of subscribed copies) within the commune.

We have, however, assumed in the introductory discussion concerning the conceptual framework that it is important to an unusually high degree for the newspapers to obtain a high share of the market within the area of circulation. This is so owing partly to thetendency of local advertisements to seek out the largest newspaper and partly to the nature of the costs of distribution. Therefore, it was assumed in the conceptual framework that the surveillance of communes where the newspaper had been subjected to competition from another paper will be more extensive than from other communes.

In order to test this last assumption the control communes have been divided into three classes with respect to the newspaper's share of the newspaper market within the commune. Where the share of the market is low - under 25 % - the competition from other newspapers can be judged as being so hard that their dominance is practically impossible to upset.³⁹ In communes where the newspaper's share of the market is very high, on the other hand - over 75 %, - the newspaper is completely dominating.40 The other twenty-five percents are often divided among

several newspapers and primarily represent only a few newspaper copies. However, in the share of the market of the middle group, 25-75 %, the competitive situation ought on the whole to be of such nature that none of the existing newspapers dominates the market. Thus, there are relatively many potential readers. Control communes where the market share of the newspapers which were studied lie within this middle group ought, therefore, to have higher values than other control communes within the same newspaper type and size-class.

When the following calculations have been primarily performed as preparatory runs for an attempts towards stricter analysis of the reciprocal importance of the three explanatory factors, a punched card deck, intended especially for this latter purpose, has been used. This card deck differs from the earlier one only by the fact that all communes where the newspapers are circulated in at least 50 copies are included. This limitation is slightly broader than the earlier one (5 % of the households) and the difference only influences the result in the lowest share of the market class. Table 13 shows the number of col. cm. (without advertisements) from the control communes in the group of six- and seven days per week publication, divided into size classes and market-share classes. The figures within parentheses give the number of observations in each cell.

The classification of the size variable is more minutely divided in Table 13 and in the following than earlier in the lower size classes. Table 13 shows that our assumption is confirmed in two of the communal size-classes but that a high share of the market leads to the highest value in the three largest classes. However, the assumption does not hold att all in the smallest class. The connection which was found earlier with communal size remains within all market-share classes.

Upon a study of the news articles variable our assumption would have been confirmed
even in the size class 6,000-7,999 inhabitants. Moreover, no changes in the
relation between different cells would have occurred.

When a control of news production is made, the correlation with communal size disappears, as earlier. However, the number of col. cms./100 inhabitants in table 14 shows the same correlation with market-share as the col. cms. values in Table 13. Thus, the condition that control communes with high shares of the market in the three largest size classes received higher values than the next lower market-share

class can not be explained by differences in the size of the communes within the
size class.4l

Table 13 and 14 are now based upon observations from all six days and seven days per week newspapers and the number of observations is low in the deviating classes. It is possible the deviating values in these size classes can be influenced by the distribution of the type of newspapers within the different cells. The deviating control communes can, for example, stem exclusively from newspapers with small areas of circulation.

The possibilities for simultaneously holding also the newspaper type constant are now limited since the number of control communes "does not suffice" for observations in all cells. If it is accepted to analyse as few frequencies as five, certain possibilities exist nevertheless. The average number of col. cms. in every cell with both type of newspaper, communal size and market-share held constant is evident from Tables 15 a-g. The values within parentheses are based upon 3-5 observations. Cells with lower frequencies than 3 have, as have cells where there are no observations, been maked with a dash (—). Henceforth, only the values based on at least five observations will be commented upon.

It is evident at once from the table that the control communes in the deviating size classes are not concentrated to any special type of newspapers. More than three observations from these size-/market-share classes exist only within the newspaper type four and five, i. e., mono-papers. Quite naturally, these are responsible for most control communes in the highest market-share class. Of the observations from the deviating size classes which do not appear in Table 15 four are from the monopaper group and four from the first-paper group.

It is also evident fromTable 15 that the second-papers do not have, in any instance,
three or more observations in the highest market-share class. Purely parenthetically,
this illuminate the market situation of these papers.

When the type of newspaper is held constant, how valid is the assumption that the control communes in the middle market-share classes have higher values in contents than other control communes? With the exception of the smallest size classes the assumption is confirmed in all size-/newspaper-type classes where comparisons are possible. In the smallest size class, on the other hand, the assumption does not hold at all. An ad-hoc explanation for this dissimilarity between size

classes lie, however, close at hand when consideration is taken of the conceptual framework. These communes are each so small that they are relatively uninteresting as markets for the newspapers and, therefore, are not objects for market type considerations. The surveillance of smaller communes has previously shown possession of greater elements of routine than that from larger ones.⁴²

It is also evident from Table 15 that the earlier found correlation between col. cms. of communal contents and communal size remains when both the type of newspaper and the share of the market are held constant. The relation between the type of newspaper (read: the size of the area of circulation) and the col. cm. of communal contents also remains when both communal size and market size are held constant.

The Newspaper situation: summary

The correlations which were found earlier between the extent of communal contents from different control communes and communal size or newspaper type, respectively, are not changed upon an introduction of the market-share variable within the respective commune. On the other hand, the communal contents also show a correlation with this variable in accordance with the assumptions made. The fact that contents from control communes where the market-share is higher than 75 % is lower than in control communes where the market-share lies between 25 % and 75 % is thereby the strongest support for the assumption of the importance of competition within the commune since the values in the lowest market-share class can be partly influenced by different judgements on the part of the editorial staff concerning where the area of circulation "end".

Summary

The aim of this study has been to illuminate somewhat the factors that determine the extent of information the citizens can obtain via the press about communal and local political activity in their home communes. These factors have been assumed to be related partly to attibutes of the respective commune, news production and the newspaper situation within the commune, and partly to the choice of newspaper or access to a newspaper (the newspapers were classified in different types of newspapers with respect to publication frequency, the competitive situation at the place of publication and the size of the area of circulation).

After an introductory discussion of these three factors in which the newspapers were primarily viewed as commercial enterprises as regards the present hard economic conditions, the three explanatory factors have been successively introduced into the exposition. The unit of investigation has consistently been composed of the contents in a given paper from a given commune, i. e., the contents the readers of one newspaper have access to from one commune. This unit of analysis has been described with a special term "control commune" in order to avoid an ambiguous use of language.

When the control communes, distributed according to size classes, were studied,
a high correlation with communal size appeared. However, upon a control of dif

ferences in news production by studying the col. cms. values per 100 inhabitants, large and small communes received similar values. Therefore, the correlation with communal size was interpreted as a result of dissimilarities in news production - the number of communal civil servants (upon use of the second operationalisation of news production) received instead a falling trend with rising communal size: a fact that was interpreted as a reflection by the number of civil servants of the qualitative dimension to an altogether too small extent.

The contents from the larger control communes were to a greater extent placed in a leading place in the newspaper, consisted to a greater extent of news articles and were to a greater extent classified as negative. This indicated that the larger communes were devoted more interest on the part of the newspapers than the smaller ones. However, a large part of these differences proved upon a comparison between control communes from places of publication and other localities to be connected with the attribute of being a place of publication.

The differences between places of publication, both qualitatively and quantitatively,
were so extensive that the places of publication were from that point on left
outside of the calculations (they composed only 4 % of all units of analysis.

When several qualities of the newspapers were introduced into the analysis a strong correlation with the size of the areas of circulation was noticed. Control communes from newspapers with limited areas of circulation consistently obtained higher values than control communes from newspaper with large areas of circulation. On the other hand, no correlation could be stated between publication frequency and communal contents. During our three month period the non-daily newspapers contained, by and large, as many col. cms. from the individual communes as did the six- and seven days per week newspaper. The non-daily newspapers gave communal contents a more prominent placement at the same time as a smaller part of it was classified as routine surveillance.

When the newspapers were classified in different types with consideration being taken also to the competitive situation at the places of publication, the assumed ties between the extent of competition and the extent of surveillance could not be pointed out with certitude. The value of the results on this point are, however, dubious since the places of publication themselves were not included in the calculations. However, the small number of observations made it impossible to perform an analysis of these. The existence of so-called branch editions proved, on the other hand, to be accompanied by a more extensive material from the individual communes within the areas of circulation of these branch editions. The citizens in the suburban communes were nearly completely referred to other information channels during the period embraced by the investigation.

Another way to study the importance of competition for the extent of local content is to compare control communes with different market situations. When the control communes were class-distributed as regards the papers share of the market of them it could be stated that both high and low shares of the market appeared to lead to less extensive surveillance than the middle group where the competition is strongest. However, the possibilities for making these comparisons were limited by

the fact that there were an insufficient number of observations for many newspaper
type/size classes and market-share classes.

By way of summary, we can state that we have, by and large, found the correlation between the extent of the surveillance and the "explanatory factors" that we have assumed in the conceptual framework. The information about his own commune that a member of a commune can obtain via the press is dependent partly upon the production of news - surveillance of the commune shows a strong correlation with communal size - and partly upon the qualities of the paper read, primarily the size of the area of circulation - newspapers with limited areas of circulation devote greater space to the individual communes than newspapers with extended areas of circulation. The surveillance of individual communes proved also to be dependent upon the competitive situation on the newspaper market. The newspaper devoted most space to the communes where the competition between different newspapers was strongest. However, it is impossible to declare with such simple methods anything certain about the reciprocal "importance" of the different factors for variations in the extent of the surveillance or about how large a part of the total variation in the dependent variable can be traced to the undependent variables studied«

An attempt to investigate more thoroughly with the help of regression analysis "the addition to the explanation" of the variance in the surveillance of these and other operationalizations is now being conducted. The result of this work will be reported on in the Communal Research Group's dissertation series.

---

NOTES

1 In the following by information we mean communications concerning a condition not the knowledge that individuals pocess which can possibly be the result of the fact they have been acquainted with the communication.

2 Professor Jorgen Westerståhl has presented the work of the Local Government Research Group in its entirety, Westerståhl 1967.

3 See 22 ff.

4 See Furhoff. 1963. D. 39 ff.

5 SOU 1965 :22 D. 87.

6 SOU 1965 : 22.

7 A report of the entire investigation is given in the Local Government Research Groups' disseration series nr. 11.

8 See page 38 for definition of communal contents.

9 A through discussion of the economic situation of the newspaper market in Sweden can be found in SOU 1965 : 22 p. 50 ff. This presentation has played a great role in the development of this conceptual framework.

10 SOU 1965 : 22, p. 13 ff.

12 This assumption concurs also with the conclusion that the majority of the Press Committee made concerning the importance of competition on the newspaper market. "In the opininon of the Committee, one must, however, proceed from the basis that competition creates better and more effective newspapers for the benefit of both the newspapers' readers and political life. In the opposite case, the conditions under which the newspapers work differ in a fundamental respect from the laws which govern all other economic activity". (SOU 1965:22, p. 90).

13 By "mono-papers" we mean only newspapers issued six or seven days per week without competition at the places of publication. Most newspapers in the non-daily grovparepare also alone at their places of publication, but are not included in the "mono-papers" group in this study.

14 Lars Furhof f has a more detailed discussion of the "branch edition" concept. (Furhoff 1965).

15 The results from the metropolitan newspapers will only be reportedon sporadically, since the available circulation statistics for these are less detailed than those for rural newspapers, Hereafter, the metropolitan newspapers will be assumed as being local papers within their respective statiscal "newspaper areas," i. e., the areas where these newspapers are in the majority over newspapers from other places of publication. A more detailed account of how the boundaries of the "newspaper areas" are drawn can be found partly in the Newspaper Statistical Annual, Stockholm every year, and partly in Arpi, Hall and Wiren, 1966, p 100 ff.

16 The Press Committee appears to be prepared to assist towards an increase in the size of the areas of circulation if it would be possible to watch the competition within the area in doing so. (SOU 1965:22, p. 68).

17 The control communes assume the qualities of the parent commune concerning size, degree of population density, etc.

18 The measurement which was then obtained has, however, its limitations as Johansson and Nilsson, among others, have shown. With the support of studies of literature and their own investigations they find that newspapers published at the same place show large similarities in contents, (Johansson and Nilson, 1965). Thus it is likely that different newspapers from the same commune report the same news for the most part.

19 SOU 1965 : 22, p. 14.

20 At the sample of newspapers, information in the TS (Newspapers' Statistics) publication "Landsortspressens spridning per kommun, 1966" (The Circulation of the Rural Press by Commune, Stockholm 1966) was followed. Thus, the newspapers whose circulations have not been controlled by TS are not included.

21 As the lower limit for circulation 10 % of all households was chosen. In the cases where the newspapers are published in several editions only the edition which is circulated within the selected commune is included in the sample of papers and in the investigation.

22 The different "sampling quotas", which of course are not desirable from the investigation point of view, are a natural result of the sampling procedure. As a result of our having chosen a number of communes and having investigated the newspapers which are circulated in the former, newspapers with large areas of circulation have greater probability of being included in the sample than newspapers with limited areas of circulation.

23 The sample includes only one newspaper issued two days per week and no papers issued five days per week.

24 If we had not been bound by this desire it would have been natural to investigate a random sample of days of publication during 1965, for example, which was a year completely free of elections in Sweden. 24a Examples of communal questions: 1. Minutes from communal meetings. 2. Requests to state authorities and decisions by that authorities concerning requests from communes on the granting of loans by the government, appeals of communal decisions, for example, etc. 3. News and reports concerning companies owned by the communes. 4. Communal activities of a service nature, for example, snow clearance, electrical distribution, gas works. 5. Reports concerning localization of industries where a communal representative has expressed an opinion or where the commune has promised to build housing units or the like. 6. Minutes from meetings from local party organizations (also women and youth organizations). 7. Personalia concerning communal representatives where the communal work duties are treated.

---

24 If we had not been bound by this desire it would have been natural to investigate a random sample of days of publication during 1965, for example, which was a year completely free of elections in Sweden. 24a Examples of communal questions: 1. Minutes from communal meetings. 2. Requests to state authorities and decisions by that authorities concerning requests from communes on the granting of loans by the government, appeals of communal decisions, for example, etc. 3. News and reports concerning companies owned by the communes. 4. Communal activities of a service nature, for example, snow clearance, electrical distribution, gas works. 5. Reports concerning localization of industries where a communal representative has expressed an opinion or where the commune has promised to build housing units or the like. 6. Minutes from meetings from local party organizations (also women and youth organizations). 7. Personalia concerning communal representatives where the communal work duties are treated.

25 It is necessary that one of the following criteria be fulfilled for an article to be classified as negative: a. An account of illegal acts by communal authorities. b. Report of an appeal of a communal decision. c. Accusation of neglect on the part of the commune (for example, from the newspaper or from the person interviewed). d. Accusation of waste or ineffectiveness in administration, lack of foresight by the commune (from the newspaper, person interviewed or person cited, for example).

26 When the calculation has been done by electronic computer the mean has been consistently used as a central measurement. In the cases when the skewness of the distributions make this misleading attention is drawn to this in the text. The two skew distributions which regularly appear are partly the surveillance variable and partly the communal size. By consistently making a certain control of this, the disadvantages of the central measurement used are reduced, however.

27 All three operationalisations of surveillance show a strong co-variation with the size of the commune. The co-efficient for the product moment correlation for the relation between the total column centimeters and the col. cms. of non-advertisements is as high as + .99. Thus, it does not noticeably influence the result if we study from this point on the variable col.cms. in total, even incertain cases when it would actually be theoretically most correst to study the col. cms. of non-advertisement or vice versa.

28 Source: Arsbok 1966

29 Source: Årsbok for Sveriges kommuner, 1966

30 The information has been calculated with the help of information in Årsbok for Sveriges kommuner (1960 and 1966)

31 Source: Statistiska meddelanden AM 1967 : 19 and AM 1967 :30

32 Many of these newspapers support the Center Party and have a close connection as far as contents are concerned with the farmers' organizations movement.

33 The total space of communal contents/newspaper can not be counted only by adding the values for the control communes, since some of the newspapers have several editions which are not included in our investigation.

34 The competition on the Swedish newspaper market can be regarded as duo-political. As the Press Committe shows, the most important means of competition of the newspapers are: 1. the form of the product, 2. distribution, 3. active sales investments, and 4. the price. In the case of distribution and price newspapers at the same place publication do not differ as a rule. Consequently, the form of the product and sales investments remain. (SOU 1965:22, p. 73 ff.)

35 In the final report on this investigation an account will be given of the extent of the endeavours of the communal authorities to inform the citizens in a number of communes. An important instrument for this activity is the sending of documents to the newspapers.

36 Concerning the limitation of this newspapers' areas, see note 15.

37 The number of control communes in the different classes is evident from the table below: ™ r Number of inhabitants Newspaper Totøl _399g 4000-7999 8000-14,999 1 74 38 21 14 2 345 192 98 41 3 65 34 17 11 4 163 84 57 17 5 101 67 28 5 6 161 106 41 8 7 120 86 28 5 8 107 55 42 6

---

38 After the investigation was carried out one of the Stockholm newspapers has started local suburban supplements. This increases, of course, very manifestly the newspaper's possibilities for publishing local news.

39 The newspaper choice of the rural population should probably be dependent to a nonnegligible extent on in which central locality purchases of durable consumer goods takes place. The communes where the newspaper's market-share is low are presumably to a large extent such communes where the population travels to another locality than the newspaper's place of publication for its purchases and where the newspaper's possibilities for expanding considerably its circle of readers ought to be small.

40 The class limits 25 % and 75 % respectively are arbitrarily chosen out of consideration of the desire to get fairly similar frequencies in the classes and partly to get a class division which differentiated. If the correlation appears as been assumed, the differences between the means become greater the further "out" from the middle points of the scale that the class limits are chosen.

41 A measurement of the percentage of col.cms. that has been classified as negative is given in the table below. If the percentage of the contents that is classified as negative surveillance can be considered to indicate greater interest on the part of the newspapers vis-å-vis the commune, the assumption holds in three cases of six (in the size class 8,000-14,999 the percentage of negative material is the same as in the two higher classes). Of course, there is no reason to place too great importance on these figures. Since the percentage is consistently so low the margin of uncertainly should probably be large and several of the differences stated in the table ought to reasonably lie within this margin. Table 4. Percentage of the contents classified as negative. Control communes from newspapers with six- and seven days per week publication divided according to size and marketshare classes. Number of Inhabitants Market share -1999 2000- 4000- 6000- 8000- 15,000-3999 5999 7999 14,999 29,999 -24% 2.4 7.1 7.9 15.4 7.7 12.9 25 %-74 % 8.2 10.1 8.8 18.8 8.8 5.6 75 %- 4.4 21.1 7.0 0.7 8.8 13.3

42 The reasonableness of such an interpretation is illuminated by the table below which shows the average percentage of the newspaper's circulation that each control commune represents. The values in the tables are thoroughly expected since the percentage of the circulation that is sold in a commune is by and large a function of three factors, communal size (the number of households), the newspaper's market-share within the commune and the size of the circulation. Number of Inhabitants Market share -1999 2000- 4000- 6000- 8000- 15,000-3999 5999 7999 14,999 29,999 -24% 0.55 0.87 1.13 1.33 2.17 3.02 25 %-74 % 0.82 1.67 2.75 3.81 6.24 9.70 75 %- 1.26 2.83 3.78 6.33 6.50 16.83

43 It is apparent that it will be possible to gain more certain conclusions about the newspapers' editing principles, above all, when the local surveillance of them as a whole is placed in relation to other contents. However, this is complicated by our newspaper sample where as a rule only one edition is included from newspapers with several editions and this part of the investigation is not concluded.

REFERENCES

Birgersson, Bengt Owe: Kommunal bevakning i svensk landsortspress (Communal Surveillance
in the Rural Press in Sweden), The Local Government research Group dissertation
Serie No. 11, Stockholm 1968.

Westerståhl, Jb'rgen: Swedish Local Government Research, in Scandinavian Political Studies,
Vol. 11, Helsinki 1967.

Furhoff, Lars: Kommunal Skendemokrati (Communal Sham Democracy), Uddevalla 1961.

Furhoff, Lars: Pressens forrdderi (The Treachery of the Press), Stockholm 1963.

Statens Offentliga Utredningar, (State Official Investigations), 1965 : 22 Stockholm.

Furhoff, Lars: Definitionsforsok for dagspress.
Statsvetenskaplig Tidsskrift, Stockholm 1965, p. 439 ff.

Newspaper Statistical Annual, Stockholm, Annual.

Arpi, Hall & Wiren: Region och Marknad (Region and Market), Stockholm 1966, p. 100 ff.

Johansson och Nilson: Konkurrens och innehåll. En innehållsanalys av dagstidningarna i Eskilstuna och Orebro - "Competition and Contents. A Content analysis of daily newspapers in Eskilstuna and Orebro". Beredskapsnåmnden for psykologisk forsvar nr. 27, Stockholm 1965.

Statistiska Meddelanden, Arbetsmarknaden 1967 : 19 and 1967 : 30. The Statistical Review,
Labourmarket 1967 : 19 and Labour Market 1967 nr. 30, Stockholm.

Årsbok for Sveriges kommuner 1960 och 1966 (annual for the Swedish Communes), Stockholm.