When attention drives attention: Issue dynamics in environmental news reporting over five decades

Introduction

A key feature of environmental news reporting over the years has been its apparent fluctuating nature. The overall attention to ‘the environment’ in the news goes up and down with different issues dominating the news agenda at different points in time. A significant factor explaining this oscillating pattern is that events and issues do not appear as discrete entities on the news agenda. Instead, issues tend to interact and this dynamic interaction may contribute to the cyclic or oscillating pattern in news attention over time. Issue dynamics mean that reporting on one issue influences coverage of other issues. Research on issue dynamics so far, however, has been surprisingly scant both in environmental studies and in journalism studies at large. Thus we do not know whether, for instance, the intense coverage of climate change in recent years has contributed to an increase in the attention paid to other environmental issues, or if climate change has occupied so much of the available news space that all other issues are crowded out.

This article investigates the extent and characteristics of the issue dynamics in environmental news reporting over a 50-year period. The case under investigation is Swedish television news. The study analyses all news stories on environmental issues broadcast in the main public service news programme during the period 1961–2010.

Issue dynamics and the environment

Contemporary discourse on the environment is deeply rooted in the notion of ‘holism’ and the systemic qualities of life on earth. However, the extent to which a holistic perspective permeates everyday environmental news reporting is a matter open to discussion. Although the media and journalism are said to be decisive in the creation of a public understanding of ‘the environment’ as a unified concept and domain (Hansen, 2010: 6), there have been few attempts at actually measuring the extent to which the environmental domain displays a holistic or fragmented character in the news over the years.

Research design and methodology

Most studies to date on issue development and issue agendas in environmental news reporting have focused on the press. A problem when analysing newspapers is that the volume and format of papers have changed considerably, making it difficult to find consistent measures of attention that are comparable over time. The present study makes television news the object of investigation. By focusing on a specific news programme that has displayed basically the same ‘format’ (30 minutes, commercial-free, lead by an anchor and mixing news bulletins and longer stories) for over 50 years, consistent and comparable measures of attention are attainable.

The research strategy guiding the study is to investigate the flow of topics on the news agenda over a very long period of time. The study focuses on a single news outlet: the main newscast on public service television in Sweden. Public service television is no doubt a main source for national and international news – in Sweden and in many other countries as well. Aktuellt was the first news magazine on Swedish television (1958) but since the early 1970s Rapport has had the highest news show ratings. Focusing on what can be regarded as the main news programme on public service television, Aktuellt was chosen for the analysis from 1961–1969 and Rapport for the rest of the period, 1970–2010. All the programmes broadcast during the prime-time slot in these years were surveyed for environmental news stories.

To actually watch all newscasts over a 50-year period is an impossible task. Instead the analysis is based on the written documentation produced in association with the airing of all newscasts. These records are logged the day after the broadcast and contain detailed information on the content of the news show, including descriptions of all news stories.

The search for environmental stories was done manually, i.e. no archival search engine was used, by examining all lists and judging each news item individually. All the coding and analysis were conducted by the present author to ensure consistency in coding principles. An intra-coder reliability test was conducted (two years of newscasts were coded a second time), and the test showed a very high degree of consistency (Cohen’s kappa = .92). Lists for the year 2005 were not available in the archive when coding was conducted in 2010, and this year is excluded from the analysis. ¹

Defining ‘environmental issues’

The study departs from a constructivist approach which suggests that environmental news is constructed when journalists frame certain topics and events as ‘environmental’. From a constructivist position it is difficult to devise a general and timeless definition of ‘the environment’, applicable over extended periods of time and in different national contexts. The solution applied in this study is to take the common denominator of all major environmental discourses (Dryzek, 1997) as a point of departure and define environmental news as news that is framed as to concern the negative impact or the risk of such an impact of human activities on the natural world, but also all the societal processes and activities addressing these problems, including research, technical innovation, opinions, activism, protests, politics, policy, decision-making, etc. Industrial production and energy issues (e.g. nuclear, fossil and renewable energy) are included when stories focus on the environmental implications of production and consumption. Natural disasters and extreme weather (e.g. droughts, hurricanes) are part of the environmental domain if they are framed as caused by human activities. The same principle applies to major health hazards and pandemics (e.g. BSE).

Using this strategy for selection, a total of 8867 environmental stories were found during the search. These stories were subjected to a quantitative analysis. All news topics where classified into main categories consisting of similar issues/themes, and the categorization was done in a way that allows for comparisons both over time and across countries. For instance, ‘industrial pollution’ (the general category) is present in all countries, but the specific industries or production processes that are the targets of the debate obviously differ (the specific topics). The main issue categories were chosen so as to cover the most important themes in the environmental debate since the 1960s onwards (see Table 1 for definitions of categories).

OPEN IN VIEWER

Table 1.

Issue categories in environmental news reporting 1961–2010 (in percentages, per five-year period).

	1961–1965	1966–1970	1971–1975	1976–1980	1981–1985	1986–1990	1991–1995	1995–2000	2001–2004	2006–2010	Total
Water pollution and water quality	12	11	6	2	4	11	4	5	4	4	6
Air pollution and air quality	5	4	1	1	2	2	1	1	1	0	1
Acid rain and acidification	1	1	2	1	7	2	1	1	0	1	2
Climate change and global warming	0	0	0	0	0	1	2	4	8	21	4
Ozone depletion	0	0	0	0	0	2	1	0	0	0	1
Industrial pollution	5	8	9	12	14	13	5	12	6	3	9
Facility sitings and land use controversies	1	6	11	6	3	2	11	1	1	4	5
Fisheries and marine resources	1	0	0	0	1	1	4	1	8	3	2
Biocides, hazardous chemicals and metals, toxic waste	10	14	11	9	6	3	2	2	5	1	5
Workplace environment	6	4	22	10	7	4	2	1	2	2	6
Environmental conservation	19	3	1	4	12	5	8	9	9	10	7
Nuclear power and nuclear waste	3	0	7	26	11	25	13	13	5	10	14
Fossil fuel and non-renewable energy	0	0	1	2	1	0	1	0	2	0	1
Oil spills/ disasters	4	13	3	10	6	3	5	3	6	3	5
Hydro-electric power and renewable energy	5	11	4	4	5	1	2	4	3	6	4
Deforestation, forestry, clearcutting	0	0	1	1	2	1	2	2	1	1	1
Environmental effects of food production and agriculture	2	4	2	2	3	4	5	18	21	10	8
Traffic and transport	2	5	4	4	4	5	4	5	6	9	5
Environmental effects of war	16	4	6	1	3	1	10	1	2	0	3
Consumer society	5	5	3	2	2	2	7	6	4	7	4
Energy conservation	0	0	1	1	1	2	1	3	1	2	2
(Continued)
General environmental politics and legislation	2	5	2	0	4	8	8	7	4	3	4
Other	1	2	3	2	2	2	1	1	1	0	1
Sum	100	100	100	100	100	100	100	100	100	100	100
N (seconds)	21,776	55,861	103,744	142,761	90,565	150,347	99,931	114,844	94,013	135,041	1,008,883

Notes: N = seconds of broadcast time. Data for 2005 are missing. Definition of categories: 1. Air pollution and air quality, including levels and types of pollution of the atmosphere and effects of air pollution (on humans, animals, plants), urban air quality and smog, regulations and management of air pollution. 2. Water pollution and water quality, including the levels and types of pollution of different bodies of water (lakes, rivers, streams, oceans, ground water) and the effects of water pollution (on marine life), regulation and management of water pollution, water purification, water distribution and water resource management, sewage, water contamination, water deoxygenation and eutrophication; 3. Industrial pollution, pollution, emissions and waste from specific industries or production processes, such as mining, aluminium plants, electronic and automobile industries, medical waste, e-waste, industrial disasters (not nuclear disasters); 4. Facility sitings and land use controversies, including the localization of industries, mines, power plants, traffic routes; 5. Fisheries and marine resources, including whaling, overfishing, illegal fishing, fishing methods, fish farms; 6. Biocides, hazardous chemicals and metals, pesticides in forestry and farming and toxic waste (including specific substances and contaminants such as dioxin, DDT, toxic heavy metals, PCB); 7. Acid rain and acidification, including damaged forests, acidification of lakes (a subcategory of air pollution); 8. Climate change and global warming, including effects of global warming, emissions trading and carbon tax, green house gas emissions, extreme weather; 9. Ozone depletion, including CFC and the ‘hole in the ozone layer’ (a subcategory of air pollution); 10. Workplace environment, including work hazards and safety, pollution and toxic substances in the workplace (not general accidents); 11. Environmental conservation and biodiversity, national parks, wild life management, endangered animals and plants (including protection of specific species such as wolves, birds of prey, seals); 12. Nuclear power and nuclear waste, including nuclear safety, nuclear and radiation accidents, nuclear fallout, nuclear waste management; 13. Fossil fuel power stations and non-renewable energy, including production, resources and environmental impact relating to fossil fuels (coal, natural gas and oil) (not oil spills and industrial accidents occurring in fossil fuelled stations); 14. Oil spills, all oil spills at sea or on land; 15. Hydro-electric power and renewable energy, environmental impact (negative and positive) of renewable energy production, including hydro, wind, solar, wave, biogas, thermal; 16. Deforestation and environmental effects of forestry including clearcutting; 17. Food production and agriculture, including GMO, environmental degradation due to farming and livestock industry, animal rights in farming, farm animal protection laws, use of artificial fertilizers, monocropping, organic food production; 18. Traffic and transport, environmental effects of traffic and transport including emissions from traffic, noise, fumes, lead in petrol, new technologies for fuel-efficient and low-emission vehicles, including electric and hybrid cars; 19. Environmental effects of war, including nuclear weapons testing, dumping of toxic or nuclear waste from weapons production, ‘dirty bombs’, environmental impact of the Vietnam war, the Gulf war, etc.; 20. Consumer society, including environmental impact of consumption, including ‘green’ consumption, eco-labelling, household waste and dumps, littering, recycling, over-consumption; 21. Energy conservation on the consumer side (saving energy), energy-efficient and energy-saving products; 22. General environmental politics, legislation and policy, non-issue-specific environmental debate and politics (the political parties’ general strategies and activities in environmental politics, general environmental opinions and attitudes, general legislation); 23. Other issues, all other topics framed as dealing with ‘the environment’, including news stories where the main topic is impossible to classify.

Measuring attention and interaction

To be able to compute and compare attention over the years, a consistent measure of the level of attention is necessary. Two initial measures were produced: (1) the number of stories on different issue categories and (2) the time allotted to different issue categories. The latter is computed by weighing stories by their duration. Since the length of the stories changes considerably over time, this measure provides the best comparative estimate of attention. In order to chart the fluctuations of attention, both measures were divided by the number of analysed newscasts in order to generate two measures that are consistent and comparable over time: (1) the average number of environmental stories per newscast and (2) the average broadcast time allotted to environmental news per newscast. The latter measure was used to calculate the yearly average of environmental attention in the news (Figure 1) and the relative distribution of attention to different issues per five-year period (Table 1).

OPEN IN VIEWER

Figure 1.

Attention to environmental issues in public service television news 1961–2010 (seconds per newscast each year).

Aggregated measures of attention were computed in order to examine the degree of issue interaction. This was achieved by summarizing the seconds of broadcast time devoted to each issue for every month (and subsequently quarter and year). These aggregated measures were then correlated to estimate the degree of co-variation between issues. Issue interaction is thus measured as the zero-order correlations between aggregated measures of time allotted to different issues in a specific news outlet (programme) in a specific period of time (month, quarter, year). First monthly aggregates were correlated, then quarterly, and finally yearly measures (Tables 2 and 3).

OPEN IN VIEWER

Table 2.

Correlates of attention to 23 different categories of environmental issues 1961–2010 (Pearson’s r of monthly aggregates, N = 568).

	Water	Air	Acid	Climate	Ozone	Indust	Sitings	Fish	Biocide	Workplace	Conservation	Nuclear	Fossil	Oil spill	Renew energy	Forestry	Food and agricul	Traffic	War	Consumer	Energy cons	Politics
Water	1
Air	.109**	1
Acid	.045	.014	1
Climate	−.039	−.039	−.080	1
Ozone	.131**	.054	−.018	−.039	1
Industrial	.173**	.240**	.097*	−.077	.148**	1
Sitings	−.001	.018	−.017	.032	.025	−.006	1
Fisheries	−.034	.051	−.051	.116**	.032	−.032	.027	1
Biocide	−.003	−.012	.060	−.078	−.029	.091*	.050	−.056	1
Workplace	−.047	−.015	.025	−.072	.073	.075	.078	−.078	.127**	1
Conserv	.109**	−.042	.009	.120**	−.011	−.009	.024	.074	−.063	−.115**	1
Nuclear	.013	−.019	−.012	−.028	.021	.079	.051	−.007	−.005	.059	−.030	1
Fossil	−.058	.094*	−.003	.027	.055	−.007	.027	.104*	.005	.143**	−.007	.112**	1
Oil spill	−.026	.013	−.055	−.072	−.018	.011	.063	.048	.019	.053	−.054	.114**	.020	1
Renew	−.015	.002	−.057	.084*	−.001	−.028	.043	.018	.011	−.002	.021	.017	.001	.078	1
Forestry	−.005	−.001	.176**	.088*	−.025	.002	−.020	.055	.031	−.030	.063	.032	.067	−.021	.015	1
Food & agricul	−.014	−.013	−.044	.108**	.016	.010	−.067	.013	−.043	−.050	.119**	.005	−.046	.007	.010	.024	1
Traffic	.029	−.067	−.035	.268**	−.020	−.022	.031	.099*	−.043	−.074	.146**	−.001	−.045	.017	.013	.063	.057	1
War	.106*	−.054	−.011	−.052	−.008	−.037	.027	−.056	.013	−.037	−.024	−.008	−.042	−.032	−.043	.000	.028	−.050	1
Consumer	−.005	−.082	−.008	.295**	−.056	−.057	.027	.096*	−.045	−.047	.087*	.013	.012	−.028	.081	.106*	.080	.165**	.080	1
Energy cons	.021	−.054	−.046	.107*	−.049	.000	.076	.044	−.009	−.006	.074	.180**	.014	−.010	−.014	.038	.087*	.039	−.035	.034	1
Politics	.207**	.025	.016	−.003	.146**	.019	−.004	.048	−.051	−.081	.078	−.003	−.033	–.090*	−.025	.132**	.054	.092*	.109**	.069	.049	1
Other	.158**	−.023	.017	−.054	.092*	.032	.022	−.052	.031	.063	.009	.005	.045	.013	−.029	−.005	−.041	−.005	.489**	.107*	.021	.171**

Notes: The table reports zero-order correlations. Significant correlations are highlighted in bold and by asterisk(s): **p < .01, * p < .05. The definitions of the 23 categories of environmental issues are described at length in the notes to Table 1. The level of attention is an aggregated measure of seconds of broadcast time allotted to different categories of environmental issues in the news programme Aktuellt (1961–1969) and Rapport (1970–2010), computed per month (April 1961–June 2010). N (number of months) = 568. Data for 2005 are missing.

OPEN IN VIEWER

Table 3.

The number of negative and positive correlates (Pearson’s r) of attention to 23 categories of environmental issues 1961–2010 (measured monthly, quarterly and yearly).

	Attention measured monthly		Attention measured quarterly		Attention measured yearly
	All months (April 1961–June 2010)		All quarters (3rd quarter 1961–2nd quarter 2010)		All years: (1961–2010)
	Count	%	Count	%	Count	%
Positive correlations	142	56%	160	62%	157	62%
Negative correlations	111	44%	93	38%	96	38%
Sum (total number of correlations)	253	100%	253	100%	253	100%
Significant positive correlations (significant at p < .05 or better)	42	16%	35	14%	41	16%
Significant negative correlations (significant at p < .05 or better)	2	0%	3	1%	3	1%
N (months/ quarters/years)	568	568	193	193	49	49

Notes: Count = the number of correlations. N = number of observations (months, quarters, years). The total number of possible zero-order correlations between 23 different issue categories is 253. The level of attention is the seconds of broadcast time per news programme allotted to different categories of environmental issues in Swedish television news computed per month (April 1961–June 2010), quarter (3rd quarter 1961 to 2nd quarter 2010) and year (1961–2010). Data for 2005 are missing.

Issue interactions: Crowding-in or crowding-out?

When examined over an extended period of time, one of the most conspicuous characteristics of environmental news reporting is its oscillating pattern. Over the 50 years the attention to the environment in the news has fluctuated considerably, and periods of heightened attention have been followed by considerable drops (Figure 1). Environmental reporting started at a very low level in the 1960s but has since displayed a typical wavelike pattern. The years 1979, 1986 and 2007 are all time highs, and 1982, 1991 and 1999 represent the low points of the attention curve. The issues on the news agenda display a similar fluctuating pattern (Table 1).

The major peaks in reporting are similar to those found in previous studies – albeit limited in numbers and comparability – both in time and in the issues reported (cf. Hansen, 2010: 22). Here we find the focus on environmental conservation and biocides in the 1960s; the attention to industrial pollution and nuclear power in the 1970s; acid rain in the early 1980s; the environmental impact of food production in the late 1990s and early 2000s; and finally the strong focus on climate change later in the 2000s (see Djerf-Pierre, 2012b, for an analysis of the specific issues covered).

This clearly suggests that national attention dynamics – from the outset of the environmental debate – are part of the global flow of public attention, often fuelled by the same global crises events (Cottle, 2009). To be sure, some of the noticeable surges in attention in Swedish television have appeared in conjunction with disasters occurring overseas, such as the nuclear disaster at Chernobyl in 1986. But if one issue attracts much attention, does it crowd out other issues from the news agenda or do issues interact, increasing attention to the entire domain of environmental issues?

The long-term fluctuations in issue attention are a key aspect of issue dynamics, but the relative distribution of attention to issues (Table 1) provides only limited information on how issues ‘vary together’. When there is a surge in the overall attention to the environment in the news (e.g. 1979, 1986 and 2007) the total amount of attention (seconds) to an issue may increase while the relative amount decreases (percentage of seconds). Also, issue interactions would be expected to play out in shorter periods of time (months, quarters, years) and not in five-year periods.

Issues interact if the level of attention to one issue is associated with an increase or decrease in the attention to other issues. By computing zero-order correlations between aggregated measures of attention to all issue categories, we measure the degree to which a high level of attention (time allotted) to one issue category is associated with higher or lower levels of attention to other categories. Table 2 shows the monthly correlates of attention to the 23 categories of environmental issues.

There are 253 possible zero-order correlations in a correlation matrix containing 23 variables. Out of the 253 correlations 142 are positive (56%) and 111 are negative (44%), which means that the overall pattern indicates a positive relationship between issue categories (Table 2). However, if we only count the correlation coefficients that are flagged as significant in the table, the positive coefficients dominate even more: 42 are positive and only 2 negative. Thus, there is clear evidence supporting the hypothesis that positive correlations exist between the different categories of environmental issues on the news agenda. When one issue category receives a high level of attention, the attention to other environmental issues tends to increase as well.

From Table 2 we can see that there is a recurrent pattern of positive correlations among the categories of environmental issues. This does not mean that all issues are positively associated to the same extent. The issue category that displays the greatest number of significant positive correlations with other categories is ‘Climate change and global warming’. Eight significant correlations with climate change can be observed and this is a clear indication of the pervasiveness of the climate change debate in recent decades. It is not surprising that the debate on climate change triggers a number of related environmental concerns, such as energy conservation, consumer choice, traffic and transport, food production and agriculture and renewable energy. These constitute issues that can be regarded as factually related to each other (cf. Kepplinger and Habermeier, 1995).

The correlation patterns tend to correspond to expected factual linkages between issues. For instance, acid rain is positively correlated with industrial pollution and forestry; biocides with industrial pollution and workplace environment; energy conservation with nuclear power and climate change; and so on. Some correlations, however, do not seem to originate from a direct factual link. The most obvious example here is the positive relationship between fisheries (mainly focusing on overfishing and the depletion of fish stocks) and traffic/transport. These are issues that have displayed surges in attention at similar points in time, but they do not share an obvious factual component.

The simultaneous co-presence of issues that are not factually linked supports the claim that there is indeed a holistic dimension to environmental news. The fact that there are just two instances out of 253 where a significant negative correlation can be observed offers additional evidence that a high level of attention to one category of environmental issues does not crowd out other environmental concerns. On the contrary, high levels of attention to one issue tend to ‘crowd in’ even more issues and attract additional attention to the entire environmental domain. This pattern remains highly stable when testing for other techniques of aggregating data. Table 3 summarizes the results of correlation matrices computed for monthly (same correlates as presented in Table 2), quarterly and yearly aggregates of attention. Regardless of whether we aggregate the data by year, quarter (three months) or month, the correlation pattern remains the same. Thus, issue interaction is a persistent feature of environmental news reporting, observable on a monthly, quarterly and yearly basis.

Conclusion: Attention drives attention

This article substantiates the fluctuating quality of environmental news reporting. The five decades of reporting clearly display an oscillating pattern in news coverage of environmental issues, as peaks in attention are followed by steep declines in interest. The categories of issues on the news agenda have also varied considerably over time, but the study provides clear evidence that an interaction exists among environmental issues in the news. There is indeed a positive correlation between different environmental issues with regard to the level of attention. An intense focus on one category of environmental issues does not crowd out all other environmental concerns in the news. On the contrary, a surge in interest for one category of environmental problems generates attention to other environmental issues as well.

The fact that attention seems to attract more attention indicates that issue interaction can help us explain the long-term cyclic pattern of news attention to environmental issues. Previous research on news-waves has established that a surge in attention to key events lowers news thresholds for similar and related events, thus increasing the overall attention to the general theme for a period of time. This study suggests that similar mechanisms come to play in the reporting of issues and sub-domains of issues as well and that they persist over longer periods of time. Collective spirals of attention mainly arise among factually related issues, but in some cases also pertain to non-related issues within the broader domain of environmental news. This indicates that there is indeed a holistic perspective that permeates everyday environmental news reporting and that some ‘relatedness’ exists among the various issues reported over the last 50 years.