Below the radar: A U.K. benefit fraud media coverage tsunami—Impact, ideology, and society

Framed by a literature on media coverage of welfare, “pov erty porn” and “poverty propaganda,” and its emphasis on political messaging and ideology, this article explores re porting of benefit fraud in the United Kingdom. In a comprehensive and systematic assessment, it illuminates the structure and extremely large volume of benefit fraud cov -erage, 2008–2017, across national, sub-national, regional and local newspapers, mainstream television, TV documen taries, BBC web output, and BBC radio. With the aid of the 2016 British Social Attitude survey, and utilizing the most current conceptualizations of the processes underpinning media influence, it argues that this tsunami of coverage will have had a significant impact on the public. The results are explored in relation to their political and ideological implica -tions,


| 709
GAVIN on heavily restructuring welfare provision, while reducing welfare expenditure. Furthermore, they explore the incidence of only a tiny fraction of the many synonyms for benefit fraud. More importantly, in an extremely complex communication environment, these studies exclude numerous national newspapers and all local/regional ones. The latter reach more than 40 million adults a month, with what the public considers quality reporting (Walker, 2019). And broadcasting-in the form of news, documentary, general programming, web content, or radio-does not figure at all.
But, also importantly, the literature on benefit fraud (and "poverty porn" more generally), lacks a fully developed processual dimension. It rarely connects forward to an evidenced impact of coverage on public appreciation of the phenomenon-for exceptions, see Baumberg et al. (2012, p. 52) study of the impact of stigmatizing welfare coverage on perceptions of benefit fraud levels, and the assessment embedded in Briant et al. (2013) focus group study of audience reactions to disability coverage. Explorations of benefit fraud coverage and a broader "poverty propaganda" (Barton & Davis, 2016;Garrett, 2015;Morrison, 2019;Roberts, 2017), appear to assume impact. Or effects are implied rather than fully explored-for example, Roberts' (2017, p. 193) assertion that the repetition of messages around benefit fraud, "…renders the preferred reading [by the public] more likely.." In neither case do these researchers incorporate an appreciation of current thinking around the processes involved in "media effects." There are, then, some extremely important gaps in our understanding of coverage around poverty porn and benefit fraud. In the following analysis these are addressed for benefit fraud, and where the principal research objective is to outline the broad contours of the coverage and its prevalence, and assess its likely impact. The paper cannot hope to explore every dimension of benefit fraud coverage. But it looks systematically and exhaustively at a decade-long stretch of national, sub-national, regional/local coverage, across a very wide range of media. This coverage is then assessed in the light of what is currently understood about media effects. And with the aid of data from the 2016 British Social Attitudes (BSA) survey, it evidences some of the likely impacts of coverage on important elements of public appreciation of welfare fraud.

| RE S E ARCH ME THODS
The following assessment looks at coverage from January 2008 to December 2017, focusing on U.K. newspapers and broadcasting-both important information sources for people (Ofcom, 2013a). The internet and social media were excluded. The latter rank below the BBC, the broadsheets and the local/regional press as sources trusted to give balanced and unbiased reporting. 1 And when British people are asked which source are most likely turned to for accurate and trusted news, social media and online aggregators are important for only a tiny segment of society (BBC, 2017;YouGov, 2016). Unsurprisingly, these are primary news sources for only 6% and 2% of citizens, respectively. 2 However, older (Newman, Dutton & Blank, 2012) and younger generations (Turner, 2015) use social media to share material from legacy media, 3 while evidence from a range of sources, including from recent PAMCo data, 4 suggests the public are heavily engaged with conventional press coverage via titles' online portals (Newman et al., 2012). Furthermore, the internet was (Worcester & Mortimer, 2001) and remains (Newman, 2018) Britain's least trusted medium, an important consideration when, "An established finding of communications research is that trusted news is more likely to persuade." (Newton, 2019, p. 107). However, when British people do access the internet, there is clear indication they use the BBC website as their most accessed source (Ofcom, 2013b;Swales, 2018).
Press stories on benefit fraud were isolated using the LexisNexis database. Broadcast reports were identified using the "Learning on Screen" (LoS) database. LoS facilitated searching within all program transcripts, from January to December 2017 (the first full calendar year they were machine-searchable). But LoS allowed more extensive tracking of program titles back to January 2008. Broadcaster's online stories were accessed via the BBC News website, whose search facility also flagged radio programs. Difficulties with these databases, the associated problems in isolating benefit fraud stories, and how these issues were overcome, are outlined inSupporting Information.

| The press
One objective was to get a rigorous, comprehensive and longitudinal picture of the prevalence of "benefit fraud" press coverage. But which keywords are most appropriate to search for? Assessment began with the national press and clusters of two-word phrases broadly denoting "benefit fraud," drawn from existing benefit fraud literature (Lundström, 2013;Sage, 2012) and from "poverty porn" research: benefit-, welfare-and dole-fraud(s), cheat(s), scam(s), fiddle(s), scrounger(s), sponger(s), and skiver(s), alongside a separate, but common, political discourse trope, "skivers and strivers." The "fraud(s)," "cheat(s)," "scam(s)," and "fiddle(s)" terms clearly identified reports dealing with obtaining benefits, illegally, through fraud or deception. However, the "scrounger(s)," "sponger(s)," "skiver(s)," and "skivers and strivers" terms dealt primarily with broader themes around general welfare and poverty, and the indolent or undeserving poor, rather than "benefit fraud." These search terms were, therefore, dropped, leaving 12 phrases: benefit-, welfare-and dole-fraud(s), cheat(s), scam(s), and fiddle(s). These were assessed for their incidence embedded anywhere in stories. LexisNexis at this time differentiated print and online stories (see Supporting Information). However, we sought to isolate only unique reports, counting those appearing in either the printed or the online version, but not double-counting stories appearing in both.
This exercise identified the synonyms most commonly denoting benefit fraud, and allowed movement to isolate stories primarily about the issue, that is, moving to "benefit fraud" appearance in report's headline or first four paragraphs ("at the start" as LexisNexis designated it). This approach is consistent with the "inverted pyramid" journalistic convention: a "…structure that guides readers in finding the most important elements [of a story]." (Bejerman, 2011, p. 95). But to avoid double-counting stories carrying more than one search term, we enter them into LexisNexis simultaneously, separated with an "OR" operator. The database allows only five terms to be entered, so the most common synonyms were chosen (our "core five"). 5 This approach was adopted for nationally circulating daily newspapers, and their weekend or Sunday companions (the broadsheet Times, Telegraph, Independent and Guardian, the mid-market Mail and Express, the tabloid Sun, Mirror, Star and People, plus The Metro a free sheet circulating nationally), alongside the hundreds of local and regional newspapers also carried in LexisNexis (see Supporting Information).

| Broadcasting
To determine benefit fraud's incidence in general TV programming, we surveyed the transcripts of all broadcasts on the five channels where the vast bulk of the audience watch their television: BBC1, BBC2, ITV, Channel 4, and Channel 5. Here, a search was run on the "core five" terms derived from the assessment of newspapers: that is, "benefit fraud OR benefit cheat OR benefit scam OR welfare cheat OR dole cheat," but including "welfare fraud," since-unlike LexisNexis-LoS had no meta-level descriptors of story themes tagged on at the end of stories and clouding the analysis (see Supporting Information).
This process also identified a particular programming vehicle-documentaries. These are important in giving issues sustained attention. The titles of these documentaries were search for, to identify instances of their broadcast. But we were also aware of the titles of other documentaries bearing on benefit fraud, from existing research on "poverty porn" (Barton & Davis, 2016;Huws, 2015;Jensen, 2014). These titles were also searched for. Finally, we sought to ascertain the incidence of benefit fraud coverage on the BBC News website. Our "core five" terms, plus "welfare fraud" were entered individually into its search facility. As with the newspaper, stories were only counted where one or other of the search terms appeared in the headline or first four paragraphs of reports.

| The press
Just how deeply "benefit fraud" penetrates into the public conversation via the national press, is illustrated in Table 1. "Benefit cheat" and "benefit fraud," surfaced most often, with 4,294 and 4,125 stories, respectively, carrying them somewhere within. And, the combined number of hits across all terms was a staggering 10,097. Table 2 shows the results from the follow-on exercise to avoid double-counting items containing more than one search term, and identifying stories where benefit fraud was a central theme.
Obvious patterns emerge. First, coverage was still very extensive, with 4,619 stories published. This is also likely to be an under-estimate. LexisNexis allows only five simultaneous search terms. This precluded counting the six other Table 1 phrases, plus less frequently used synonyms identified in other research. 6 But even 4,619 is significant. The coverage began rising 2009-2012, plateaued for 2 years then declined. But in the years when coverage was greatest, 2010-2015, on average, there were between one and two stories every day, 365 days a year, somewhere in the national press, with only marginally less in the preceding and following years.
Also clear from Table 2, the largest number of stories surfaced in titles typically supporting the Conservative Party: the Express, Mail, Sun, and Telegraph, newspapers among the most widely circulating in their market sector (Mayhew, 2019). However, it also needs to be acknowledged that the left-leaning Mirror also carried 593 stories, averaging over one report a week in a title that can reach 27 million people over a month (Mayhew, 2019). Patently, an interest in benefit fraud stories is not confined solely to those titles most closely associated with conservative politics, an issue we will return to below. Table 3 illustrates another wave of coverage in the local and regional press, and where benefit fraud was the main story theme. Coverage rose steadily from 2009 to 2014, before falling off. And at 10,915, the number of stories overall seems large in comparison to national newspapers. However, Table 3 shows the number of titles carrying reports ranged between 142 in 2008 and 300 in 2014. Therefore, what we are observing, quite often, is an extremely small numbers of stories from many newspapers in any given year, that is, many titles often carry only a handful of reports.
Nevertheless, it is noteworthy that these numerous titles have a very large cumulative reach-in the region of 40 million people per month (Walker, 2019). In addition, for some titles the coverage was significantly larger than a "handful." The bottom row in Table 3 shows that between 6 titles (2016), and 30 (2011) Tables 1 and 2 are anything to go by, there are likely to have been many more stories carrying "benefit fraud" deep within the body of reports, rather than, as in Table 3, just "at the start." Overall, this is still a prodigious quantity of coverage by any standards, and was extremely widely dispersed, geographically. Furthermore, the rows in Table 3 separately representing the Evening Standard in London, and Scottish newspapers, are testimony to very significant geographical overlap in coverage. This will have been most evident in Scotland, where communities will have been exposed to, among others, the local Paisley Daily Express or Airdrie and Coatbridge Advertiser, but also the Scotland-wide Daily Record and Herald, or the Glasgow-based Evening Times, alongside any one of the UK's nationally circulating titles figuring in Table 2

| Broadcasting
Turning to television, the "core five" benefit fraud search terms derived from the press survey (but including "welfare fraud"), were sought for in the transcripts of all broadcasts across the main five channels, from January to TA B L E 1 Benefit fraud synonyms in U.K. National Press ("Anywhere in a story") c "The Metro" circulates nationally, but LexisNexis stored its stories under "Regional Press." After the 2019 LexisNexis 2018 reorganization, the Metro is no longer in the database.
TA B L E 3 "Core Five" stories in U.K. regional/Local Press ("At the start")  This coverage is vanishingly small when it is considered that total broadcast program output across the year is likely to run to a large five-digit number. However, the exercise unearthed a particular programming vehicledocumentaries-which tended to give benefit fraud more sustained attention, with broadcasts often lasting up to an hour. Some of these were available to watch in LoS, and the database provided brief program details for all of them. Both confirmed the specific "benefit fraud" theme in the targeted programs. With the titles of these documentaries ascertained, and others known from the "poverty porn" literature (Barton & Davis, 2016;Huws, 2015;Jensen, 2014), the LoS "programmes title" search was used to track their broadcasting across the whole TV sched- This exercise identified four documentary series, and one one-off. The longest-running series was BBC's Saints and Scroungers. Press stories containing the term "scrounger(s)" were rarely about benefit fraud. But the "Saints and Scroungers" web site described it as a "Series looking at benefit fraud…," and the brief details LoS offers flagged content "…looking at benefits thieves…," something confirmed by viewing those available to watch on  ously not straightforwardly cumulative, but clearly, many, many millions were exposed to one or other programme.
Finally, the BBC's web-based coverage of benefit fraud was explored. totaling 386 overall. This survey also flagged radio broadcasts dealing with benefit fraud. have not always been negligible: Radio Ulster's total weekly audience in this period was over half a million, and Jeremy Vine's radio show could net over seven million. 7 However, the overall number of web-based stories and radio broadcasts were quite small against the likely overall output on these platforms.

| A pervasive structure to coverage
Exploration of the detail and contours of all this coverage is well beyond the scope of this paper. However, one dominant pattern was very clearly discernible: a focus on individual cases of benefit crime. Table 5 makes the distinction, in web-based reports, between this type of story and more abstract and general ones. The latter could be about issues like policy ("Crackdown on U.K. benefit cheats who live abroad," 27 December 2010) or costs ("Benefit fraud error at £2.6bn," 31 October 2018). But, as Table 5 suggests, the bulk of items were individual cases revolving around the criminal justice process, like "Psychic and sex chatline worker's benefit fraud sentence" (4 November 2011) or "Czech gang sentenced over £1m benefit scam" (4 November 2013).
And this type of focus was not the sole preserve of "conservative" titles, or the national newspapers. The Mirror-a consistent Conservative Party critic-had an even heavier predominance of stories like "£16k benefit TA B L E 5 "Core Five" (plus "Welfare Fraud") on BBC Website and BBC Radio fraud by TV porn model" (18 February 2012), or "Dole cheats detector in £310k fiddle" (4 May 2012): in 2008, 77%; 2009, 77; 2010, 73; 2011, 87; 2012, 91; 2013, 77; 2014, 85; 2015, 89; 2016, 83; and 2017, 93. And local/regional press stories almost exclusively concerned individual cases, such as "Winner of mum in a million had illegally claimed £4,000 benefits," Derby Evening Telegraph (29 July 2008), or "Cheat claimed more than £110k benefits," Hull Daily Mail (27 April 2016). The same could be said for the TV documentaries identified in Table 4, where the litany of cases of criminals being sussed out, run down, and publicly exposed on camera, was relentless.
Overall, this observed pattern is entirely consistent with other research which finds the predominance of "… individual cases of benefit fraud…" (Baumberg et al., 2012, p. 33), and stories "…typically based on a preliminary charge or court conviction of persons that have cheated on welfare benefits…" (Larsen & Dejgaard, 2013, p. 295).
So Lundström (2013, p. 642) is correct in concluding that U.K. national newspapers "…predominantly construct the issue of benefit fraud using representations of individual cheaters and employing the narrative strategies of crime news…," a finding that can now be extended to local/regional newspapers, broadcast documentaries, and much web coverage.

| The overall picture
The coverage outlined above represents a veritable tsunami of coverage across a range of media, especially in newspapers but also in television documentaries. It is a heavy, cumulative contribution to the national conversation, to which the limited quantities of web-based commentary and radio programming adds a modest, but not totally insignificant, ripple. Indeed, in the period assessed, scarcely a day can have gone by, 2008-2017, when there were not at least some part of the U.K. public exposed to reports dealing predominately with, or touching on, benefit fraud-whether this was in daytime broadcast programming, or in locations served by one or other local, regional, subnational, or national newspaper. The structure of much of the commentary had an extremely heavy emphasis on an often unrelenting procession of criminal cases, something especially prevalent in the local press, but strongly detectable elsewhere.
If "poverty porn" or "poverty propaganda" cast the poor as indolent and feckless, the coverage described here positions many of them as criminally active. And if not "hegemonic," as is sometimes claimed of "poverty porn" narratives, this message is certainly very heavily pervasive. But does benefit fraud coverage play the same ideological and political roles claimed of "poverty porn" and "poverty propaganda?" Answering this question requires a nuanced and sustained assessment of the likely impact of coverage, a feature often absent in debates around "poverty porn" and "poverty propaganda."

| Reinforcement of predispositions
What was that impact? This cannot be determined definitively here. However, the processes at work certainly can, and appropriate conclusions drawn. Researcher are divided over which processes are most dominant. Some suggest a commanding role for the media in influencing citizens' perceptions (Gavin, 2018), while others focus on the conditionality of such impacts (Philo, 2008;Wolfsfeld, 2011). However, a strong strand in research suggests coverage has a weak impact on people's perceptions (Newton, 2019)-it only reinforces preexisting sentiments, and is, therefore, socially or politically inconsequential (Bennett & Iyengar, 2008). From this perspective, individuals will not passively change their opinions in response to coverage. Instead, in a process termed "confirmation bias" or "directed reasoning," they will actively choose media sources reinforcing their preexisting sentiments, and reinterpret incoming information in a manner consistent with those ideas.
This perspective does not fully or adequately explore where such "pre-existing sentiments" come from or the media's possible role in this respect (Gavin, 2018, p. 833). But even if such considerations are set aside, these reinforcement processes would see individuals who are already disinclined to support welfare provision (or think benefit fraud is unacceptably high) buy a newspaper, like the Mail, which reflects or flags numerous fraud cases.
Alternatively, people predisposed to positively evaluate welfare provision and its recipients, or who consider systemic fraud exaggerated, will buy the more left-leaning Guardian or Mirror. Such directed reasoning is certainly signaled in broader research on "welfare coverage." Barton and Davis (2016, p. 193) see "poverty porn" as "… reinforcing fear and mistrust of the economically marginal…." Mooney (2011, p. 6) considers it to "…reproduce dominant attitudes to poverty-and to welfare in general…." McEnhill and Byrne (2014:105) conclude the Sun's benefit fraud coverage contributes to "…misinforming its readership, or reinforcing their existing prejudices by reflecting already-held opinions." Some of these "opinions" are known to be woefully misinformed. For instance, the U.K. government estimates that, out of a total welfare spend of £148bn, £1.2bn is defraud per year (Baillie, 2011), that is, less than one penny per £100 spent. However, surveys suggest the public think fraud levels are much, much higher. In 2013 the average guess was around 24 pounds per £100, prompting the researchers to conclude that, "…people estimate that 34 times more benefit money is claimed fraudulently than official estimates…" 8 a finding broadly consistent with other studies (Baumberg et al., 2012, p. 51;Briant et al., 2013, p. 882).
How are such misperceptions explained, and is benefit fraud coverage implicated? People certainly recollect being exposed. Baumberg et al. (2012, p. 51) report that half the public had seen a benefit fraud story in the last 2 months. And the 2016 BSA survey suggests only a minority (26.9%) could recollect having seen no stories across the media, but nearly a third has seen 1 or 2, another third saw between 3 and 10, and over 1-in-10 registered more than 10 articles-figures unsurprising given the coverage Tables 1-7 exposes (see Supporting Information).
Was this exposure capable of influencing public perceptions? Some focus group research suggests an affirmative answer. When Briant et al. (2013, p. 882) asked their scattering of focus groups to justify their mis-estimates of disability benefit fraud, "…participants cited newspapers as their primary source and all participants were gen- The more formally representative 2016 BSA survey allows a nationally oriented exploration of the coverage-perception relationship. It asked people their estimates of national benefit fraud levels, and also their newspaper readership (see Supporting Information). The cumulative weight of newspaper stories, 2008-2016, is already known from Table 2. And, we calculated the average estimate of benefit fraud levels for the various cohorts of newspaper readers. The relationship between this cumulative commentary exposure and levels of fraud (mis)perceptions can thereby be observed (see Figure 1). It is a remarkably close, near linear one. The greater the quantity of coverage, the higher the estimate of fraud (and vice versa), a highly statistically significant relationship, even with only nine observations. 9 But is this simply a re-expression of "directed reasoning?" Do individuals already left-liberal in orientation simply buy the Independent or Guardian, and confirm their sense that fraud is relatively low? And do those already conservative in predisposition buy the Sun or Mail, to reinforce their opinion that fraud is relatively high? This can be conjectured, but there is absolutely no direct survey evidence that people choose their newspapers in this manner, that is, because of their politicized messaging (Gavin, 2018, p. 830-831). Furthermore, politicized directed reasoning does not explain the low national fraud estimates for Times readers (a newspaper consistently backing the Conservative Party and their politics), or the high estimates of their Mirror counterparts (a title consistently criticizing the Conservative Party). Additionally, a directed reasoning premise does not straightforwardly explain why the average estimates for all newspaper cohorts, are so strikingly far above the actual level, that is, less than one pound in every hundred, which would be close to the x-axis in Figure 1.
Finally, even if directed reasoning underpins and fully explains Figure 1 (which, at the very least, is doubtful), this would not undermine the notion that such reinforcing of preexisting attitudes is highly significant, attitudinally, politically, socially, and ideologically. A recent critical reconceptualization of the "weak effects" thesis challenges the idea that "reinforcement" is intrinsically inconsequential, politically, or socially. Reproduced there is a rarely quoted, but pointedly important, comment from the doyen of "weak impact" theory, Joseph Klapper: It must not be assumed, however, that reinforcement which involves no change in opinion, is synonymous with lack of effect. Reinforcement is an effect, and an extremely important one. It strengthens the resolve in question, produces a type of immunity to counterpropaganda, and nurses straying sheep back into the fold.
(Klapper, 1968, emphasis original, quoted in Gavin, 2018, p. 834) This is in line with Feldman et al.'s (2014, p. 595) assertion that "…attitude reinforcement should not be overlooked as an important persuasive effect…." And Flynn et al. (2016, p. 128) point out that, across many issues, "…political misperceptions are typically rooted in directionally motivated reasoning…," a process also thought to be directly relevant to opinions around issues like climate change (Bolin & Hamilton, 2018). And if we are concerned about messages that are ideologically inflected, it is also significant that the Conservative-supporting newspapers where benefit fraud stories figured most prominently have a combined circulation five times greater than their nonsupportive counterparts (Ponsford & Turvill, 2015), and are, therefore, much better positioned to disseminate such "reinforcing" commentary.

| Media agenda-setting and beyond
"Reinforcement" is an important media effect in its own right, but we also need to look at another media effect that is worthy of consideration, "agenda-setting"-the notion that increased media coverage of a topic induces the public to consider it "an important issue facing the country," that is, it elevates public concern (McCombs, 2014).
There is solid, consistent, representative survey evidence about the percentage of Britons categorizing "pensions, social security and benefits" as their most significant concern. 10 When this data are averaged across each year of our coverage survey it shows the distinctive, incline-decline pattern in concern (Figure 2, in gray). However, this pattern does not closely align with shifts in the quantity of local/regional press coverage, or of BBC documentary, radio or web commentary. Nevertheless, we should not expect it to. Local or regional coverage was highly dispersed geographically, while "concern" here is measured at national level. And documentary coverage was sporadic and, although occasional watched by large numbers, was not as large a part of the public's staple media diet as national press coverage. In contrast, the newspapers in Table 2 carried much more benefit fraud reporting, circulate nationally and were regularly read by many tens of millions.
We can see that the weight of national press coverage ( Figure 2 in black), closely matches the national trend in "concern." Their close correspondence does not, in itself, evidence a causal connection, However, it is statistically significant. 11 And the "incline-plateau-then-drop-to-above-pre-incline-levels" pattern of press coverage, perfectly matches Downs' (1973, p. 100-112) "issue attention cycle," a feature closely associated with the media's agenda-setting capacities. Furthermore, the correspondence between this pattern and public concern around welfare issues is certainly consistent with the impressive weight of research evidencing the agenda-setting power of the media (McCombs, 2014). At the very least Figure 2 suggests that national press agenda-setting around benefit fraud is certainly worthy of consideration, and should prompt a more statistically formal, multi-variate exploration.
Finally, the structure of much of the benefit fraud coverage in Tables 2-5 is relevant to an important distinction between "episodic" and "thematic" reporting (Iyengar, 1994). The latter is characterized by in-depth coverage of issues, and which heavily contextualizes them, giving detailed, rounded explanation, and interpretative analysis, involving causes, effects, and exploration of implications. In contrast, "episodic" reporting describes concrete, isolated events, individual illustrations or specific instances, and cases around an issue. The latter is precisely the structure of commentary found to be very widely prevalent across much of newspaper and television coverage, that is, a parade of separate, individual instances of crime, pursuit, and punishment. This episodic dominance is important. In respect to crime-related news, it has a well-evidenced impact on those exposed (Iyengar, 1996). It tends to elicit a strong sense of only the individual's decontextualized blame for a crime, rather than-as thematic coverage does-encouraging people to believe that such problems have complex, deep-seated, societally located roots that connect strongly to political and social structures, processes, and institutions. These perceptions, in turn, have important implications for people's appreciation of appropriate treatment options, that is, retributive punishment and individualized redemption, versus complex, societally located interventions. The results in Tables 2-5 should be viewed in this context.

| CON CLUS IONS
In their assessment of "What drives attitudes to the welfare state?" Park et al. (2013, p. 52) scarcely consider the media, and seemed decidedly incurious about "poverty porn," "poverty propaganda," or benefit fraud coverage. This is understandable if the media are thought to only, and inconsequentially, reinforce people's preexisting F I G U R E 2 Issue salience and total national press coverage: 2008-2017 a attitudes, the implication being that little is gained from sustained attention to the weight of, or patterning around, long-term coverage. But in light of the preceding analysis of coverage and its likely impact, this incuriosity is decidedly unhealthy. This analysis suggests that, over a 10-year period, there was a deep wave of benefit fraud coverage, with a recurrent and heavy focus on specific cases of criminal behavior. This was highly likely to have driven public concern about the benefit system, reinforced or influenced politically significant perceptions of the poor, and help to generate woeful misperception of benefit fraud incidence.
All of this is decidedly consequential, for a number of reasons. The preceding assessment does the sort of heavy evidential lifting often missing from the "poverty porn/propaganda" literature, in terms of rigorous, exhaustive, and longitudinal analysis of coverage across platforms, married to a sustained exploration of current research on impacts. In the light of the evidence presented above, such connections between media coverage and their political or ideological corollaries are much more firmly and plausibly supported. For instance, if we consider politics, the preceding analysis flagged the role of coverage in generating misperceptions about welfare fraud levels.
So, there is considerable overlap between "poverty porn/propaganda" and benefit fraud discourses, and the ideologically politicized messages they convey. But the pervasiveness of benefit fraud coverage highlighted in the preceding analysis, should prompt us to consider other politically and ideologically charged discursive overlaps.
For instance, one provisional assessment of benefit fraud coverage has found that the individuals involved were often "immigrants" (Gavin, 2018, p. 837), a connection also surfacing in Berry's (2019) study of media coverage and its reception by the public, around the time of 2008 financial crisis. Another study shows that "scrounger" stories connected to an anti-European discourse, where newspapers suggest the United Kingdom is "…infiltrated by job-poaching, benefit grubbing bogeymen…" from the European Union (Morrison, 2019, p. 152). And as Gavin (2018) argues, even if the role of coverage connecting these topics is only to reinforce audiences' preexisting attitudes toward benefit claimants, immigrants and Europe, this can have considerable ideological, political, and societal significance. This is forcefully underscored when it is appreciated that the "Leave" campaign during the Brexit referendum run-up were consciously and openly strategizing around, and hoping to capitalize on, the discursive connection between immigrants, their benefit claims and the "intolerable strain" this put on public services (Shipman, 2016, p. 40).
It should be admitted that there is still a lot to learn about the structure, discursive detailed and framing of benefit fraud stories. By the same token, we need more direct exploration of journalistic attitudes and practice around benefit fraud reporting. For instance, in a context where the business models of commercial media are inordinately strained (Freedman, 2010) are journalists in different sectors of the press consciously rehashing stories from other titles about benefit fraud cases, in order to draw eyes-on, and eschewing duller but more illuminating stories about the surrounding processes or related policy issues? Likewise, did the uptick in television documentaries in 2013 owe less to political developments like the 2012 Welfare Reform Act, than to the fact that repeats from previous years could be re-broadcast to fill airtime cheaply? This is especially important in an environment where the sort of "episodic" reports featuring in much benefit fraud coverage "...do not require reporters with subject-matter expertise; and, being devoid of interpretative analysis, they are less likely to be labelled as biased by media critics." (Iyengar, 1996, p. 62), but also where all media are under relentless financial pressure and axing specialist reporters (Freedman, 2010), and where the broadcasters are highly sensitive to accusation of bias (Hill, 2013).
Finally, we need further exploration of what moves and influences benefit fraud perceptions, perhaps involving a more multi-variate, formal statically approach, and a broader range of prospective drivers. But the type of assessment offered in preceding analysis would be an important precursor to this type of advanced exploration.
And it would also be a necessary foundation for the evaluation of other "beneath the radar," but as yet unexplored, potentially important politically and ideologically charged themes that overlap in public discourse. In this context, "incuriosity" about the long-term patterning of coverage and its likely impact would surely be inappropriate, and decidedly so.

E THI C S APPROVAL
The analysis in this paper used publicly available sources, and anonymized and publicly available survey data, therefore, formal ethics approval was not sought.

CO N FLI C T O F I NTE R E S T
The author declares that they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are publicly available from the LexisNexis (https://www-lexis nexis -com.liver pool.idm.oclc.org/uk/legal/ news?sourc eid=1061905) and the BUFVC's Learning on Screen (http://bufvc.ac.uk/) databases, and via the BBC News website (https://www.bbc.co.uk/news). The 2016 British Social Attitudes data are also publicly available (https://beta.ukdat aserv ice.ac.uk/datac atalo gue/studi es/study ?id=8252), reference number SN8252.