Abstract
This study content analyzed how elite news sources framed Taiwan’s housing policy and inequality that underlies its major social problems in the press. Results show that reports used pro-market rhetoric, not pro-social equality justification. Official sources causally assigned unaffordable housing to individuals, whereas nonofficial sources blamed the government’s failed policy. This research added to the scarce literature on framing of economic issues. It’s also among the first framing studies incorporating a time element in analysis.
News sources can affect public policymaking through agenda-building (McCombs & Shaw, 1972) and frame-building (Entman, 1993) functions. Meanwhile, news media are often criticized for being agents of power (Altschull, 1995), indexing the viewpoints of those with power (Bennett, 1990) and becoming a “guard dog” for elites (Donohue, Tichenor, & Olien, 1995). Therefore, political elites as news sources can easily mediate news content (Shoemaker & Reese, 2014) and help construct social reality (Herman & Chomsky, 1988; Tuchman, 1978).
Housing is a critical public policy area in any society as it reflects the wealth and well-being of the public (Cockerham, 2007; Stafford & McCarthy, 2008). Many nations do not offer adequate and affordable housing to their citizens (M. Chen, Tsai, & Chang, 2007; Graddy & Bostic, 2009; Rahman, 2010) because of an unjust social structure (Feuerherd, 2018; Fisher & Rohter, 2007; Office of the United Nations High Commissioner for Human Rights, n.d.; United Nations Human Rights Office of the High Commissioner, 1966/1976). Housing inequality due to a failing policy not only undermines individuals’ overall health but can push a nation into social disorder, confronting a myriad of social problems and a resultant staggering social cost.
Housing inequality, among social problems and social injustices, can be a direct outcome of news representations of wealth distribution, social resources, market forces, and media exposure (Cwikel, 2006). It is documented that news frames built by social forces as news sources and embedded in the news content can effectively shape public cognition and forge attitudes and behavior toward a controversial issue or public policy (Scheufele, 1999). News coverage of the housing issue could thus mold the public’s understanding and perception of the issue and its monetary activities and allocations on the housing market, creating a spiral effect on the market orientation. Thus, a framing analysis can help explain why housing policies, in essence, often fail to promote affordability and equality because (a) framing patterns reveal the engagements of various social actors (b) whose frame-building efforts help shape or even determine government policies (Kenterelidou, 2012). Moreover, a framing analysis of Taiwan’s housing policy can provide hindsight on pragmatic corrective measures for other major cities in the world, such as Hong Kong, Sydney, Vancouver, and Auckland that face a similar plight against which public complaints have increasingly mounted (e.g., “Housing in All Main NZ Markets Now ‘Severely Unaffordable,’” 2019).
This research analyzed how four major Taiwanese newspapers allowed powerful social actors to frame-build the country’s housing policy. This research helps fill a void in the frame-building literature by examining how the powers of news sources influence (Entman, 1993) its housing policy by taking a temporal analytical approach (Gottlieb, 2015). Most studies on framing focus on one event or issue that lasts for a relatively short period of time, and the event or issue in question often does not directly affect the majority of citizens in a given society. This study is a notable exception because it takes a temporal approach by investigating a critical issue affecting a large number of citizens in an international setting over a 5-year span (Chyi & McCombs, 2004).
Housing as Underlying Social Foundation for Social Justice
Housing as “the foundation of family life” (Bratt, 2002, p. 14) requires a committed policy (Erasmus, 2010) that fulfills truthful happiness, empowering individuals to feel secure (Clapham, 2010). A decent home—reasonably roomy and affordable—as the core of a new social contract (Bratt, 2002) fulfills and enhances socio-psychological well-being (Clapham, 2010). Insecure, unstable or poor housing undercuts well-being and health equality via growing exposure and vulnerability to risk factors of illnesses and discriminations (Arangua & Gelberg, 2013; Catalano & Kessell, 2003; Hastert, Ruterbusch, Beresford, Sheppard, & White, 2016,), family disintegration and psychological insecurity (Erasmus, 2010; P. Fowler, Henry, & Marcal, 2015), and social isolation (Tach, 2009). Bryant (2009) and Shaffi (2017) stressed that housing is the quintessential health determinant and lack of adequate housing underlies many social problems. Thus, housing is the underlying social foundation and housing policies dictate social development and stability in that housing also has intermediating influences on suicide (K. A. Fowler, Gladden, Vagi, Barnes, & Frazier, 2015; Houle & Light, 2014; Jones & Pridemore, 2015), fertility, and work safety, among other key social issues. Rising housing prices predict a plunging birthrate and labor force as young couples cannot afford a home that provides for their basic need (Chen, 2013; Strom, 2010). Taiwan today has the lowest birthrate worldwide (Population Reference Bureau, 2012). A plummeting birthrate means a crippled labor force but intensified global competition for talent hunts (Yahya & Kaur, 2010) in which Taiwan is being out-competed. Moreover, persistent housing unaffordability and debased income lead to overwork for the bulk of the labor force in an unequal society (Baxandall & Breslow, 1999), which can lead further to workplace injuries, mental stress, and cardiovascular or other fatal diseases.
Problems With Taiwan’s Housing Market
The capital, Taipei, is among the most unaffordable housing markets worldwide (Global Property Guide, 2014). Demographia (2014) placed Taipei and New Taipei as No. 1 and No. 3 in worldwide housing unaffordability, with a ratio of 15.01 and 12.67 (Feng, 2014). The ratios for Hong Kong, Singapore, and Japan were 14.9, 5.1, and 4, respectively. Taiwan’s housing prices escalated ever since the adoption of the government’s housing market salvage plan, initiated by the construction industry amid the 2008 financial crisis (“Saving the Housing Market to NT$3 Million per Ping in Taipei,” 2009). That plan drove Taipei City and New Taipei City to become among the top-three priciest markets worldwide (Global Property Guide, 2014) in terms of house-price-to-income ratio, indicating a country’s housing affordability. Housing prices in the two major cities kept breaking records. For instance, housing prices in the Taipei market grew by 13.4% in the single year of 2013 (Tsai, 2014) when the average house price across the city was about US$13,000 per square meter, and, more stunningly, luxury houses were priced at US$30,211 per square meter.
The surging prices were due mainly to the dramatic cut in the inheritance tax rate—from 50% to 10%, in 2009—by the reigning Ma government, luring the bulk of some US$26 billion in cash flow from overseas, mostly China, to domestic housing markets. The optimal inheritance tax rate is deemed between 50% and 60% (Piketty & Saez, 2013). It was thus widely blamed as the culprit (B. F. Chen, 2015; Hsiao, 2014) for ballooning housing prices and widening wealth disparity (Chu, 2015). This ill effect was compounded with lowered mortgage interest rates and marginal taxes (Hsieh & Chou, 2016) in owning and selling a house (1.2% annual property tax, based not on the market value), in effect making the buying, selling, or owning of a new three-bedroom house less costly than owning a 10-year-old Toyota Corolla.
As housing increasingly became a superior venue for lucrative capital investments (Ng & Feng, 2016; Ren & Yuan, 2014), many individuals and businesses jumped onto the bandwagon. For instance, at one point, the number of the real-estate agencies and agents outnumbered the omnipresent convenience stores and staff across the country (Lin, 2011), a phenomenon that evinces what Bridge (2001) dubbed about real-estate agents: imposing boosters of the housing market.
The oft-criticized housing policy seemingly took an unexpected turn when a luxury tax, arguably aimed at the housing market, was adopted in June 2011 (C. Y. Hsieh, 2014). The Ministry of Finance was, nevertheless, singled out as the culprit for failed policies that had strategically and covertly pulled up housing prices (Sun, 2012). Was the slow, reactive policy shift reflected in news coverage over time?
Framing and Frame Building
A frame is widely understood as the central idea that threads through an entire news article (Gamson & Modigliani, 1987). Framing is “to select some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation for the item described” (Entman, 1993, p. 52, italics in original). A frame allows journalists to emphasize or de-emphasize key elements of social problems and facilitate audience discourses to make sense of behaviors or issues in question (Gitlin, 1980; Goffman, 1974; Pan & Kosicki, 1993; Tuchman, 1978).
A complete framing process contains frame-building and frame-setting stages (de Vreese & Lecheler, 2012). Frame-building refers to external forces that construct media frames (Scheufele, 1999) or frames in communication (Chong & Druckman, 2007). Frame-building reflects the dynamic endeavors of the vying communicators imposing their belief systems or views in news texts; it can be best described by Entman’s interactive forces between frame functions and communicators and texts as frame locations. Political communicators who exercise their cascading political powers (Entman, 2003; Valenzano, 2009)—from the president at the top to legislators to the least politically endowed—vie to establish, define, diagnose, and interpret a social problem through filtered facts or viewpoints in communication platforms. At issue is how unequal rival communicators vied to determine the nature, cause, and solution of the heated contentious housing policy in news coverage.
News frames prevailing from the contested frame-building process are the “imprint of power” (Entman, 1993, p. 55). Wolfsfeld’s (2004) political contest model similarly emphasizes how invaluable resources are that enable political actors to effectively leverage their political clout that further enables them to dominate news media access and, thereby, control the political climate. Gandy’s (1982) cost-effect notion of information subsidy details the impacts of resources on news gathering. He argued that resource availability, propriety, and participants’ wealth would dictate the news information decision-making process and news media access.
Bennett’s (1990) indexing hypothesis posits that journalists essentially reflect “legitimate” news sources’ breadth of views in public debates. This practice in an implicit manner, normatively and routinely gate-keeps whose and what voices are heard in the “democratic” political process. Official views are normally indexed with consensus, and nonofficial views are allowed only into the gate widened by competing elites. Official news sources—in a political or business arena—often are influential social actors given power through the government or other social institutions (Cross, 2010; Watts & Maddison, 2014). Only a limited number of viewpoints in public policy debates will pass through this filtering process. As a result, news sources who are official—primarily political or business elites—have much more influence on indexed frames in news content (Groshek, 2008). Compelling evidence on sourcing patterns (S. T. Kim & Weaver, 2003; Wahl-Jorgensen, Berry, Garcia-Blanco, Bennett, & Cable, 2016) thus far has been in line with Sigal’s (1973) seminal research concluding that official news sources dominate news reports. For instance, Wahl-Jorgensen et al. (2016) revealed that BBC cited overall political sources—among 16 overall source types—more than half of the time and devoted nearly one-fifth of its media access to prominent British government officials, confining the purview of opinions on controversial, critical issues.
These conceptions essentially come down to the underpinning notion of elitism. The elite theory of public policy analysis argues that public policies are enacted upon the prevailing values and interests of the “thin stratum” of elites (Key, 1967) and that top-down public policies change incrementally, not radically (Dye, 2013). This thin-but-cutting echelon notably includes political and business leaders. In combination, this explains why it took a while for Taiwan to adopt policies that might reduce housing prices. While a president usually tops the influence hierarchy, business leaders can initiate or adjust policies in their favor through overt and covert maneuvers (Dye, 2013; Kingdon, 2011). In the United States context, a thin elite circle encompasses prestige managers and journalists in the news business, particularly those closely allied with the White House and Congress (Mills, 1956). As predominant news sources, political and business elites can help build news agendas and issue frames by effectively using their political or financial resources (Chang, 1999; K. K. Kim, 2003; Koduah, Agyepong, & van Dijk, 2016).
Shoemaker and Reese (2014) explained the essence and significance of news sources in a nutshell. They conceptualized five descending hierarchies of influence on news content: social systems/ideology, social institutions/extra media, media organizations, routine practices, and individual journalists. These five layers of influences can interactively dictate news production procedures. In particular, the overarching influence of ideological subsystem is exerted through social institutions/news sources leveraging their resource muscles on news organizations usually through advertising. It follows that the interactive influences between resourceful news sources and news organizations become the most enforcing on the news process. News sources, they state, have a staggering, permeating influence on shaping messages across the levels as the indispensable origin of information and a source of finances. At the same time, constrained organizational financial policy further restrains news media routines (Shoemaker & Reese, 2014; also see Wu & Lambert, 2014). In short, proactive and predominant communicators as news framers can be commanding in the news production process due to their indispensable roles as news sources for both media information and finance (Dimmick, 1974; Gans, 1979; Tuchman, 1978). Thus, when “private ownership, pursuit of profit by self-interested entrepreneurs, and free market” (Shoemakers & Reese, 2014, p. 71) become the overruling ideological belief for any typical capitalistic nations like the United States and Taiwan, it is natural and rational for the construction industry to aggressively promote a free housing market that helps maximize profits via actively building and imposing a pro-market news frame on the press.
Influences of Media Ownership on News Content
News organizations are one layer of influence on news content (Shoemaker & Reese, 2014). U.S.-based researchers have found empirical evidence to support this argument. For example, the financial interests of corporate owners were related to how newspapers covered the 1996 Telecommunications Act (Gilens & Hertzman, 2000). The ownership structure of a newspaper’s parent company can make a difference in coverage of election campaigns. For instance, “newspapers owned by corporate shareholder controlled companies and large chains are more likely to offer coverage with a negative tone” than their privately owned counterparts (Dunaway, 2013, p. 35). In addition, Time magazine was found to cover Time Warner entertainment products more often after the merger of these two companies (Lee & Hwang, 2004). Furthermore, Fox News, generally considered a “red” or conservative media outlet, takes a conservative stand in its coverage of global warming by treating this subject as bogus and insubstantial (Fedlman, Maibach, Roser-Renouf, & Leiserowitz, 2012).
Knowing that ownership likely affects news content, would the financial and political backgrounds of the owners of Taiwan’s four major newspapers be reflected in the coverage of the housing issue in these news outlets?
When speaking to a group of student visitors led by the lead author of this study years ago, the chief of administrative staff of the Liberty Times claimed that the paper’s editorial policy—pro-“green”—was initially driven by its business strategy for a wide-open market as the market for the other end of the political spectrum was already saturated at that time. The newspaper’s owner was a tycoon in both the real-estate and construction businesses. While widely perceived as politically supportive of the then-opposition Democratic Progressive Party or DPP (the “green” party in Taiwan), this daily could also face the bottom-line challenge. The United Daily News, considered pro-KMT (the Nationalist or “blue” Party, which controlled both the legislative and executive branches of the government during the study period), also was engaged in real-estate activities but to a much lesser degree. The China Times, deemed not just pro-blue but also pro-“red” (Mainland China) for its opulent parent company’s substantive business interests in China, is in neither the construction nor real-estate business. The Apple Daily, often lauded for its political independence, can be financially constrained, as its owner had cashed in from selling some of his other news publications. Of the four, The China Times and Liberty Times are arguably the most financially healthy due to their parent companies’ financial resources and, therefore, probably less vulnerable to business performance pressures. Considering the political and business affiliations of these four papers, it is worthwhile to compare how they covered the housing issue.
In short, both governments and corporations as news sources can effectively exert their leverage with political and/or financial tools, such as advertising, to influence news content. The construction industry in Taiwan spent US$100 million on marketing in 2010, growing some 91% over a 5-year period (Directorate-General of Budget, Accounting, and Statistics, 2011). Would newspapers be more receptive to the viewpoints of these major advertisers?
Because news media are supposed to be the guardians of Western-style democracies (Hallin & Mancini, 2012), it is important to investigate whether the press in Taiwan has essentially become the “guard dogs” of political and business elites (Donohue et al., 1995).
Framing and Responsibility Assignments
de Vreese and Lecheler (2012) conceptualized and underscored the notion of frame valence—positive or negative aspects or outcomes—of the countering choices presented, which is compatible with Iyengar’s (1991) responsibility assignments via episodic or thematic framing, and with Entman’s (1993) treatment recommendations. According to Iyengar, episodic news reporting details an event, while thematic coverage offers its contexts. Furthermore, episodically framed news will lead the readers to assign both causal and treatment responsibilities to individuals involved in the event. By contrast, thematically framed reports enable the audience to ponder the contextual factors that underlie the problem and therefore assign responsibilities to the government or society. The concepts of Entman and Iyengar together suggest that powerful political actors, as news frame builders, can push their views in a way that forges the audience’s perceptions and responsibility assignments of social problems.
Framing has been widely adopted in the analysis of political issues (de Vreese & Lecheler, 2012; Iyengar, 1991) but much less so with public health and social problems (Koduah et al., 2016; MacKenzie, Chapman, & Holding, 2011; Maher, Fraser, & Wright, 2010; Viswanath & Emmons, 2006). While many such issues or problems are power-driven and tied to housing (Berkman, Kawachi, & Glymour, 2014), the authors of this article could not locate any framing research on housing policy or housing inequality. Framing studies on other public health issues are reviewed in the following.
Adopting Iyengar’s (1991) framing concept—episodic versus thematic—and responsibility assignments, McGinty, Webster, Jarlenski, and Barry (2014) examined news story presentations—event versus context—of mass shootings. They investigated public perceptions of serious mental illness (SMI) and reactions to a public policy aiming to regulate the gun ownership of such individuals versus a competing policy to overhaul gun control. They reported that the majority (69%) of news items over a 16-year period were event-oriented and depicted SMI individuals as highly dangerous and detestable. McGinty and associates also found that thematic reports were more likely to not only discuss SMI as often stigmatized and non-violent, but also to include anti-gun violence policy initiatives. They concluded that news framing can affect support for public policies to regulate either “dangerous people” or “dangerous weapons.”
In the same vein, Paterson (2006) applied Entman’s (1993) framing functions to demonstrate how mental illness-related killings were framed differently. The government in one case was held accountable for leaving the general public vulnerable to a foreseeable danger by failing to hospitalize those with mental disorders. In another case, the management and on-duty staff of a mental hospital, where a killing took place, were blamed for negligence. There was no mention of any failing social policies in coverage of the second case. Likewise, using Entman’s (1993) notions of framing functions and locations, Koduah et al. (2016) scrutinized how maternal health policy evolved. They stressed the value of studying the “imprint of power” by revealing how political and legal elites framed Ghana’s maternal health policy.
A Temporal Approach to Analyzing News Content
Most content analyses focus on one issue or incident during a relatively short period and analyze the frames aggregately. For instance, Andsager, Chen, Miles, Smith, and Nothwehr (2015) identified four types of frames (normative, hedonic, health improvement, and time/cost) among nutrition stories in 10 community newspapers in the United States from July 2011 to June 2012. Haysashi et al. (2016) qualitatively compared how television news in five countries covered the 2012 London Olympic Games differently.
A 2004 framing study by Chyi and McCombs, a content analysis of news coverage—between April and May 1999—of Colorado’s Columbine High School shootings, is among the few framing studies with a temporal approach. (Some agenda-setting-based content analyses already had included a temporal approach, such as Claussen [1996]—a study that also investigated time lags in intermedia agenda-setting.) Chyi and McCombs’ space dimensions included individual (e.g., the gunmen and victims), community (the town’s high school), regional (the Denver area or the state of Colorado), societal (nationwide), and international (related problems in other nations). The time frames included three categories: the past (previous events), present (the shooting), and future (long-term effects). Chyi and McCombs also analyzed how the frequencies of frames fluctuated over a month. They pointed out that a news event has a life span, and the news media often reframe an event—by focusing on various attributes of the event over time—to “keep the story alive and fresh” (p. 22).
The theoretical framework of the Chyi and McCombs (2004) study originated from Downs’ (1972) theory of issue-attention cycle. Briefly, this cycle has five stages: pre-problem, discovering the problem and initial enthusiasm, “realizing the cost of significant progress,” “gradual decline of intense public interest,” and post-problem (Chyi & McCombs, p. 23).
Even though examining how news content shifts over time seems logical, it is surprising that relatively few scholars have pursued this approach, whether conducting studies on agenda-setting, framing, gatekeeping, or another theory. One well-known example is Hallin’s (1984, 1986) investigation into how U.S. news media covered the Vietnam War. He observed that opposition to the war first belonged to the “sphere of deviance” that was “unworthy of being heard.” When a political elite—namely a presidential candidate—publicly opposed the war during the primary season in 1968, this viewpoint finally moved into the “sphere of legitimate controversy” (Hallin, 1984, pp. 21-22). From then on, opposition to the war became an acceptable topic for journalists to cover.
Another well known example is Shoemaker’s (1989) book about communication campaigns about illegal drugs, which showed how both government public service announcements and news media coverage were out of sync with the period of heaviest use of illegal drugs.
Gottlieb (2015) also took a temporal perspective—based on Downs’ (1972) five-stage framework—to study news reports. He found that The New York Times changed how it covered the Occupy Wall Street protest movement between September 2011 and July 2014. Journalists first paid more attention to the economics frame and then shifted to the conflict between protesters and city officials. Gottlieb also found that news coverage of the issue entered the stage of decline of public interest after the fourth month. In their meta-analysis of how Muslims and Islam are represented in the media, Ahmed and Matthes (2017) calculated that only a small portion of research (16.49%) used a time frame of more than 5 years.
Various historical and social scientific studies of news coverage of various social movements, such as environmentalism, feminism, LGBTQ rights, civil rights, and others, also have traced changes in coverage over time; successful U.S. movements have tended to experience three phases: being ignored by mainstream news media, then being covered negatively (as deviant), and then—when the social movement was becoming successful—receiving coverage that suggested that the social movement was not so deviant after all and perhaps that the news media had been supportive all along.
This study takes a longitudinal approach to analyze how news media framed a critical issue that affects a large number of residents in a society, and the types of news sources used, over a course of 5 years. As with the studies by Hallin (1984, 1986) and Gottlieb (2015), Taiwan’s housing policy under examination appeared to have gone through a similar process. Moreover, while Downs (1972) specified the attention cycle into five subsequences, Hilgartner and Bosk (1988) argued that the ups and downs of any (competitive) social problems cannot be so concretely divided, with each stage capable of co-occurring with the other on the public arena on which drama, novelty, and politics/culture play a key role. Therefore, a social problem such as housing inequality can have existed for some time and have waved over the course, with alarms sporadically sounding and then muting cyclically. At first, President Ma’s government adopted a market-salvaging plan initiated by industry elites even though occasional outcries of pricy housing had started to emerge; this period appears to include the pre-problem and “minor” alarmed discovery stages. Then came the surging housing prices accompanied by mounting public complaints about a “sphere of deviance” (Hallin, 1984) that the same administration ignored. This period, evincing serious alarm-sounding, reflects an unsolvable and resistant attitude that, as Down mindedly noted, is quite commonplace in older (i.e., Chinese) political culture/society about social problems. Following that, as the issue exploded and became much more dramatic and contentious, a quick remedial but likely ineffective policy was willy-nilly implemented as an effort to accommodate the “sphere of legitimate controversy” (Hallin, 1984) and enable the ruling government to remain in power in the upcoming election. In this third period, which is prolonged by consistently renewed dramas, the government realized that the cost of rectifying this social problem is too overwhelming. The problem cannot be truly resolved without a major restructuring of social resources and redistribution of individual wealth—for instance, heavily taxing the lucrative earnings from homeowners selling housing units whose sharp price rising was due to the construction of such public infrastructure as a nearby metro railway transport. A resultant luxury tax was apparently adopted as a “false play” aiming to soften the opposition voices while likely heard more frequently in the news. The shifts in the policy reflect the attitudinal variations of the political elites holding political leverage.
Hypotheses and Research Questions
Both theoretical and empirical research show the staggering influences of elite news sources. Acting as interest groups, American business elites can overtake the president and Congress as the public policy initiator through maneuvers to either maintain their status quo or to steer policy in their favor (Dye, 2013). Furthermore, framing analysis (Entman, 1993, 2003) indexing hypothesis (Bennett, 1990; Groshek, 2008), political contest (Wolfsfeld, 2004), and the hierarchy of influences (Shoemaker & Reese, 2014) all suggest that resourceful elite sources are predominant communicators who can likely control news media access and content. In addition, news media are probably part of the elite circle (Mills, 1956) and often function as a “guard dog” of the rich and powerful (Donohue et al., 1995). Empirically, the leaders of Taiwan’s construction industry maneuvered to initiate the housing market salvage plan at the early stage of the global financial crisis and tried to undermine or minimize the luxury tax (C. Y. Hsieh, 2014; “Saving the Housing Market to NT$3 Million per Ping in Taipei,” 2009). Moreover, the presidency and executive and legislative branches were under the same party’s (KMT or Kumintang, the Nationalist Party) control. Thus, political consensus remained high among these elites, and associated reporting likely did not substantively change. Therefore,
News sources as frame builders endeavor to establish, define, diagnose, and interpret a social problem through filtered facts or viewpoints in communication platforms (Entman, 1993). Chang (1999) and K. K. Kim (2003) showed how business elites influenced political elites and policy frames in safeguarding their status quo by building a fair trade frame that coerced a more open Japanese auto market or Taiwan’s and Korea’s tobacco market. Chang’s and Kim’s findings effectively demonstrated the impact of the long-held capitalistic ideology of the self-interested entrepreneurs in pursuing and maximizing profits in a free market (Shoemaker & Reese, 2014) on news representations. Therefore, leaders in the Taiwanese construction and real-estate industries would be expected to maneuver to underscore arguments that promoted the housing market to protect their financial privileges. Their viewpoint would likely dominate the debate. By contrast, nonofficial sources would try to define and frame the issue from the perspective of social economic inequality, but their voices would be heard less frequently. Moreover, while it is hypothesized that overall news coverage would be slanted toward the pro-market frame, are there variations among the four newspapers as a result of their political or financial stance? In addition, research (e.g., Chyi & McCombs, 2004; Gottlieb, 2015 Hallin, 1984, 1986) has demonstrated the benefit of taking a temporal approach to analyze the development of news frames. Housing is a major social issue, and any housing policy can benefit or damage a large number of players in an economic system. The positive or negative effects of a housing policy can take some time to emerge and will likely evolve. This study examines whether the developing stages of a social problem—from its pre-problem/minor alarm stage, to seriously alarmed discovery, to realizing the problem but resisting in solving it—are reflected in shifting news frame adoptions. Thus,
Moreover, by effectively defining and interpreting the problem, news sources can also assign responsibilities (McGinty et al., 2014; Paterson, 2006) via news valence (de Vreese & Lecheler, 2012), news presentation (Iyengar, 1991), or treatment recommendation (Entman, 1993). Official sources, particularly the construction and real-estate industries benefiting from high housing prices, would blame those complaining about unaffordable housing for their short-sightedness, hesitation, or financial incapacity. However, nonofficial sources would blame the government for failing to create a public policy that would keep housing affordable and pricing stable. Accordingly,
Finally, heeding advice to incorporate the time element in analyzing news content (Chyi & McCombs, 2004; Downs, 1972; Gottlieb, 2015), the final research question is as follows:
Methods
This study content analyzed Taiwan’s four leading dailies in circulation and political influence: the United Daily News, China Times, Liberty Times, and Apple Daily. The overall study period was from August 1, 2008, through July 31, 2013, a 5-year duration. This overall period was conceptually divided into three stages: (a) “pre-problem/minor alarms”: the outlook of Taiwan’s housing market became gloomy in August 2008, the very beginning of the second global financial crisis, but quickly turned around with a stimulus plan called for by the construction industry to salvage the housing market even though occasional outcries about pricy housing had started to emerge; (b) “serious alarmed discovery”: escalating housing prices became the principal political complaints in early November 2009 and received no positive response from the central government. Such an ignoring and “no-kidding” attitude reflected what Downs (1972) mindedly noted as a commonplace practice in older culture/society as well as what Hallin’s (1984, 1986) regarded as a “sphere of deviance”; and (c) “realizing but resisting to solve the problem”: as a response to the continuing, mounting public reproach, which turned into a “sphere of legitimate controversy,” the ineffective luxury taxes (aimed arguably at the housing market) was proposed in late February 2011 before the campaign started for the 2012 presidential election. However, no truly effective measures that helped check capital flows and speculative investments were ever initiated. Accordingly, the study duration was correspondingly divided into three periods based on the aforesaid stages and time points, using constructed weeks (CW) as the dividing line: Period 1 (August 2008 to November 2009) from CWs 1 to 16; Period 2 (December 2009 to January 2011), CWs 17 to 30; and Period 3 (February 2011 to July 2013), CWs 31 to 60.
Hard copies of news articles were obtained and analyzed. Riffe, Aust, and Lacy (2005) have demonstrated that two CW’ content can effectively represent a whole year’s content. This stratified sampling method, based on the logics of normal distribution, allows efficient and effective uses of sampling data from more normally distributed and long-lasting news issues, such as the housing problem. To ensure the dataset is sufficient while manageable, this research adjusted the constructed-week sampling technique to fulfill its special need. For each newspaper, one constructed-week’s samples were selected from each full month; altogether 60 CW of articles were included for each newspaper. As each month can have up to 5 weeks (beginning with a Sunday), the range of random numbers for any months is either four or five that constructed the stratified week. In cases of a void fifth-week Tuesday (for any 30-day months) or Wednesday (for 31 days), it is replaced with a non-redundant constructed day.
Two undergraduate student assistants searched for, identified, and photocopied all relevant pieces based on the dates generated by the adjusted stratified sampling. In total, 2,350 news articles (as cases) were included for final analysis. Many extra stories from the non-chosen dates were also collected for coding practices. The research assistants were trained by this study’s lead author to become familiar with the coding protocol and instructions. The assistants did multiple coding exercises using the extra articles before inter-coder reliability was calculated. After they achieved an acceptable level of reliability, they performed coding for the final analysis. While coding units varied depending on the nature of variables, the unit of analysis was a single article or case; 20% of the total items were randomly selected to calculate inter-coder reliability, using Krippendorff’s alpha (α). The alphas by category are reported below.
Concept Operationalization
Coding and analysis focused on two key sets of variables: news sources and news frames.
News Sources—Official and Nonofficial: Social actors or communicators as news sources are operationally defined as types of people, agents, organizations, or government units that are quoted or cited in the news by way of both routine and non-routine channels, such as interviews, news releases, press conferences, and corporate reports. A preliminary reading of some articles helped identify 14 types of communicators as news sources: (a) government officials in the central executive branch (i.e., the prime minister; α = 1.0), who took and executed presidential orders; (b) central government legislators (α = 1.0), who normally followed the will of the president, who also served as party chair; (c) central bank (officials; α = 1.0), whose decisions on maintaining or further dropping the already-low interest rates facilitated housing market investments; (d) the banking industry (α = .95), whose businesses were directly affected by the central bank’s decisions; (e) insurance companies and investment firms (α = .92), whose bulk of businesses involved property transactions; (f) the construction industry (α = .93), whose main goal was to sell its products at the highest prices possible; (g) the real-estate industry (i.e., real-estate agents; α = .96), whose main goal was to earn the maximum commissions by selling properties at the highest prices possible; (h) local/city government officials (e.g., a mayor; α = .95), whose relevant role was usually limited to proposing city rezoning; (i) city council (α = .91), whose equally limited role was to decide on the proposed city rezoning; (j) professional speculators (α = .92), who could make a good living by aggressively speculating on an active, booming housing market; (k) housing-oriented publications (e.g., Housing Outlook; α = .94), whose circulation and revenue from advertising and readership relied decisively on the prosperity of the housing market; (l) realty scholars, that is, Dr Ching-Ow Chang (α = .93), a leading advocate for affordable housing, whose academic expertise is to responsibly discuss the potential impacts of a high-priced housing market on society and citizenry; (m) regular homebuyers (α = .94), who planned to purchase property as a home, not as an investment; and (n) the non-home-buying general public (α = .90), who either could not afford or did not plan to buy a house.
As defined, the first 11 types of political and business elite communicators were categorized as official news sources, with the remaining three coded as nonofficial. Official sources are inclined to act as promoters and/or practitioners of capitalism and typically do not advocate for a benign living environment for the economically disadvantaged (Bryant, 2009). Nonofficial news sources (the 12th to 14th categories), however, are those who see or possess (or fail to possess) housing units as a necessity, not intending to engage in speculative trading of properties.
The citation frequency in the news was recorded as a continuous variable; each type of actor was counted as many times as cited in any story. For instance, if the prime minister was cited once and the chair of the central bank was cited twice in the same article, a coder would record one citation in the category of central executive branch and two in the category of central bank officials. Certain sources, such as Dr Ching-Ow Chang, had their positions or titles changed during the study periods; their coding was changed accordingly. For instance, when Chang spoke as a college professor, he was cited as a realty scholar; when speaking as a deputy mayor of Taipei, he was coded as a local (city) government official.
News frames—Pro-market and pro-social equality
Based on a preliminary reading of articles, 10 concrete news frames were identified: (a) government shall promote the market to save the economy (α = .86)—the housing market has become the locomotive of the economy and governments shall facilitate and promote the housing market; (b) shifting environment or mentality (α = .87)—former economic policies indispensable to sustain housing prices are no longer essential nowadays as the overall environment and mentality have altered; (c) optimistic market outlook (α = .85)—the market outlook is highly optimistic due to the tremendous impact of economic Memoranda of Understanding (MOUs); (d) housing investment profitable (α = .93)—housing market investment is lucrative and safe; (e) infallible housing prices in Taiwan (α = .90)—prices will not fall due to a diminishing land supply and an increasing job demand; (f) bubbling housing market (α = .87)—housing market is bubbling and will burst due to lack of accompanying economic growth (e.g., increased personal income) justifying rising prices; (g) corruption or failed government policy (α = .92)—the government’s policy has failed because it has benefited only the construction firms, real-estate industry, and speculators, while undermining those in greatest need; (h) worsening the M-shaped society with increasingly unaffordable housing (α = .86)—the M-shaped society has become more economically divided as the rich became richer and the poor poorer due to the seemingly incessant rises of housing prices; (i) governments shall rein in price and speculation (α = .86)—governments shall create initiatives that suppress inflated housing prices and rein in speculative activities; and (j) sovereignty is undermined (α = .86) by the invasion of massive Chinese capital—the massive in-pouring of money from China will allow its Communist government to take control of Taiwan’s properties and political independence. Each of the 10 original concrete frames was initially coded individually as a (yes or no) dichotomous variable. Coders decided initially whether any frame was present thorough reading each article. When a frame’s denotation (e.g., optimistic market outlook) or connotation (in a parallel, “bull” or booming market) was present, for instance, it was coded as 1; when omitted, as 0. For the purpose of analysis, the first five frames were later collectively re-coded as the more general pro-market frame, with the other five as the more general pro-social equality frame. The re-coded pro-market frame and pro-social equality frame were each run as a continuous ratio variable via summation.
News frames built by news sources not only define the problem but also diagnose its causes and assign responsibilities to stakeholders (Entman, 1993; Iyengar, 1991). Under the broader pro-market frame, shifting environment/mentality and government shall promote the market are considered, respectively, the causal and treatment responsibilities of the housing market. In this frame, individuals crying about unaffordable housing are themselves to blame because they have kept missing the optimal purchase timing due to their unbendable attitudes. However, the specific frames of corrupt government/failed policy and the government shall rein in the price/speculation would work, respectively, as the causal and treatment responsibilities of the unaffordable housing problem. Under the general pro-social equality frame, the government ought to be held accountable because it first adopted a housing market salvage plan initiated by the construction industry, then failed to adopt countering policies that would have helped rein in rising prices. It should be remembered that the government hastily and dramatically cut inheritance taxes from 50% to 10%, which was generally considered the culprit for the booming housing market and increasing prices (B. F. Chen, 2015; Chu, 2015; Hsiao, 2014; Piketty & Saez, 2013).
Findings
Overall, official news sources remain dominant. A series of one-sample t-tests shows significant difference in uses of two overall competing types of social actors across all three time periods, t = 31.92, df = 2,349, mean difference (MD) = 2.06, p < .001, and within each period (t = 19.60, df = 479, MD = 2.28, p < .001 for Period 1; t = 22.51, df = 716, MD = 2.12, p < .001 for Period 2; t = 18.01, df = 1,152, MD = 1.93, p < .001 for Period 3). H1, predicting that more official sources than nonofficial sources would be used, is thus supported.
RQ1 asks how the news media used official and nonofficial sources. Findings provided by post hoc Tukey analyses in analysis of variance (ANOVA) tests, and summarized in Table 1, reveal that, above all, an interesting pattern concerning the consistently dominant but descending use of the real-estate agency/agent/analyst and an ascending use of central government officials across the time periods. The real-estate industry personnel were used most frequently, albeit decreasingly, throughout the studied periods with an across-period average use of 1.02 (SD = 1.61) per article, but its gradual skidding from Period 2 (M = 1.10, SD = 1.61) to Period 3 (M = .90, SD = 1.43) and more notably from Period 1 (M = 1.20, SD = 1.97) to Period 3 (M = .90, SD = 1.43) was statistically significant. The use of central government officials (M = .40, SD = 2.31) across the periods was insignificant even though it nearly doubled from Period 1 (M = .24, SD = .82) to Period 3 (M = .46, SD = 3.13). Farglory/construction firms (M = .26, SD = .84) and the banking industry (M = .19, SD = .83) took the No. 3 and No. 4 spots, both of which were sliding, and the drop in the latter group was noticeably significant, as exactly was the real-estate group. City government officials (M = .16, SD = .70) were increasingly used, similar to central government officials, doubling the frequency from Period 1 (M = .10, SD = .51) to Period 3 (M = .21, SD = .78); this growth, alongside the rise from Period 2 (M = .12, SD = .66) to Period 3 was, on the contrary, statistically significant. However, the use of parliament legislators (M = .05, SD = .47) was almost halved from Period 1 (M = .07, SD = .64) to Period 3 (M = .04, SD = .47), which failed to reach statistical significance, though. Also noteworthy is the significantly increased use of housing speculators from Period 2 (M = .02, SD = .20) to Period 3 (M = .06, SD = .36). This group appeared to make an effort to oppose the luxury taxes aiming at the housing market.
Mean Scores of the Social Actors Used by Periods
p < .05. **p < .01. ***p < .001.
As for nonofficial sources, which tend to support social equality, university realty scholars (M = .16, SD = .75) were the most relied-on social actors but shown no significant pattern of usage. The frequency of the group improved slightly from Period 1 (M = .16, SD = .77) to Period 2 (M = .18, SD = .79) but dropped lower in Period 3 (M = .14, SD = .14). More critically, regular homebuyers (M = .06, SD = .35) significantly and consistently lost their ground from their already-little heard voices from Period 2 (M = .07, SD = .39) to Period 3 (M = .04, SD = .27) and from Period 1 (M = .08, SD = .43) to Period 3. In comparison, the general public (M = .06, SD = .32) gained significant voice (from M = .01, SD = .13 in Period 1 to M = .07, SD = .38 in Period 3) in the power competition of influence.
Noticeable Framing Competitions
The pro-market frame prevailed as the dominant frame. A one-sample t-test shows a statistical difference in the use of the competing news frames overall (t = 9.01, df = 2,349, MD = .15, p < .001) and within periods (t = 11.02, df = 479, MD = .41, p < .001 for Period 1; t = 4.74, df = 716, MD = .15, p < .001 for Period 2; t = 1.67, df = 1,152, MD = .03, p < .048 for Period 3). Therefore, H2, positing that pro-market frames would dominate over pro-social equality frames, is supported. Findings provided by post hoc analyses in ANOVA tests, as summarized in Table 2, reveal the shifting uses of frames.
Sum Scores and M Values of the News Frames Used by Periods
p < .05. **p < .01. ***p < .001.
Above all, use of the overall pro-market frame significantly decreased from Period 2 (with a mean use of .42 mention per article) to Period 3 (.25 mention apiece), while use of its countering frame rose from Period 1 (.12 mention apiece) to Period 2 (.26 mention apiece).
Most conspicuous among these sub-frames’ periodic uses is the persistently prevailing but statistically significant lowering use of the shifting environment/mentality sub-frame across the periods. Specifically, it dropped from .33 mention per article from Periods 1 to .21 mention apiece in Period 2, and then to .14 mention each in Period 3. Moreover, use of the housing investment/market profitable sub-frame decreased from Period 2 (.10 mention apiece) to Period 3 (.06 mention apiece); the decreasing use of the government promotes market sub-frame is noticeable only from Period 1 to Period 3.
More noteworthy are the inconsistent shifts in use of the optimistic housing outlook sub-frame, increasing from .03 mention apiece in Period 1 to .06 mention apiece in Period 2, but decreasing to .02 mention apiece in Period 3. While use of the government promotes market frame dipped somewhat, use of the government shall rein in the market (government regulates market) sub-frame surged exponentially, showing an identical pattern as the overall pro-social equality frame, jumping from .03 mention per article in Period 1 to .11 mention apiece in Period 3. Use of the housing market bubbling sub-frame was halved from Period 2 (.04 mention apiece) to Period 3 (.02 mention apiece). Perhaps most insightful in light of interdisciplinary considerations is the scarce across-period uses of the worsening M-shaped society sub-frame, which would enable better understanding of the issue in an extended and contextual sense.
Newspapers and time element matter in framing. In answering RQ2, a two-way ANOVA was conducted. The outcomes suggest that newspapers (F = 37.79, p < .001) and the time element (F = 23.31, p < .001) each exerted a significant main effect on the overall framing decision; an interaction effect (F = 9.96, p < .001) of the two factors also occurred. In descending order, coverage in the China Times (M = .45, SD = .03), Liberty Times (M = .16, SD = .03), and Apple Daily (M = .03, SD = .04) were overall slanted toward the pro-market frame while coverage in the United Daily News (M = −.02, SD = .04) favored use of the pro-social equality frame more.
Post hoc analyses suggest that this main effect among newspapers likely lies in China Times, which revealed a statistically greater slant than all three other dailies (with the p-values all less than .001). The Liberty Times slant was significantly greater than the United Daily News but no different from that of the Apply Daily. One noteworthy finding is that while China Times and Liberty Times have steadily declined in their use of the pro-market frame across the time periods, the United Daily News gradually increased in the same regard, with Apple Daily fluctuating. The time element showed a statistically significant cascading decline—from Period 1 to Period 2 (MD = .26, p < .001) and from Period 2 to Period 3 (MD = .12, p < .01)—in tilting toward the pro-market frame. Moreover, the significant interaction effect is effectively reflected in the noticeable temporal declines in frame favorability in the China Times and Liberty Times. The former had a more moderate drop in slant from Period 1 to Period 2 but a sharper sinking from Period 2 to Period 3, while the latter plunged from Period 1 to Period 2 but dropped less sharply from Period 2 to Period 3. This contrasting pattern could be due to their varying responses to the changing political atmosphere.
News Sources Point Fingers at Each Other. Hypothesis 3 predicts that official sources tend to assign responsibility for unaffordable housing to individuals, while nonofficial sources would causally assign the blame to failed government policies and would demand strict government regulations to rein in the housing market as a treatment. Findings from Pearson correlations between news sourcing (official vs. nonofficial) and frame competitions (shifting environment/mentality minus corruption/failed policy as the cause, government shall promote the market minus government shall rein in the housing market as the treatment) support the hypothesis. In causal responsibility assignment, official sources are more inclined to blame individuals, with competing frames slanted toward shifting environment/mentality (r = .06, p < .01). Simultaneously, their counterparts faulted the government, with vying claims tilted toward corruption/failed policy (r = −.07, p = .001). That is, the former blames the complaining individuals for failing to adjust their mentality to the shifting environment, while the latter holds the government responsible for the unfettered housing prices due to its corruption or failed policy. Moreover, in treatment responsibility assignment, nonofficial news sources deem the government responsible to fix this problem, with arguments leaning to government shall rein in the price/speculation (r = −.08, p < .001), but the association between official sources and the competing frames failed to reach the significance level.
The third research question investigated whether shifts in news sources and frames occurred across the periods. As illustrated by Figures 1 and 2, overall, official sources—namely business and political elites—were used as news sources much more frequently than nonofficial sources. The appearance of official sources declined over time, while the appearance of their nonofficial counterparts remained somewhat steady. The use of market-oriented frames peaked in Period 2 and dipped slightly afterwards. By contrast, social inequality-oriented frames steadily increased over time. Nevertheless, the latter category was still outnumbered by pro-market frames.
Mean numbers of two types of new sources by period
Total numbers of two types of frames by period
Discussion, Conclusions, and Limitations
These outcomes contain several interesting aspects that merit more detailed discussions. The unmatched uses of overall and specific news sources across and within the periods effectively reaffirm previous sourcing patterns as a result of varying resources and powers. Elite theory (Dye, 2013; Key, 1967; Mills, 1956) in general and framing (Entman, 1993), indexing (Bennett, 1990; Groshek, 2008), and political contest (Wolfsfeld, 2004) concepts in particular all highlight that resourceful political and business elites can have a strong or even decisive say in public policy decision-makings. Associated theoretical cum empirical evidence also shows that the highest executive leaders are the ultimate decision-makers, and business elites are tremendous power-wielders, particularly when they maneuver to collaborate with politicians to rally society around their preferred flag (Chang, 1999). But when politicians are accused of collusions or corruptions and there is a backlash over domestic public policies, it is highly probable that the existing consensus becomes split between the government and the interest group (Banting & Myles, 2016; Entman, 2003; Kingdon, 2011). Clearly, a cooperative relation has an intensified impact on their media influence, but a less harmonious division might not hurt their media access much if possessing financial leverage (Shoemaker & Reese, 2014).
At issue is the potential impact of the seeming division in Taiwan’s housing market policy on key social political actors’ media access and frame-building capability. As the findings show, official news sources as frame-builders—more specifically real-estate personnel, central government officials, the construction industry, and the banking industry—dominated the intensive news coverage. While use of real-estate personnel, the most relied-on type of news source, statistically significantly decreased over time, these dips were quantitatively minor and, in essence, truly unimportant. This group alone substantially outnumbered the aggregate use of all nonofficial sources both across and within periods. These vocal real-estate personnel are, in effect, the interpreters of the housing issue (Bridge, 2001) and become the effective “front organization” (Manning, 2001) of the construction industry as they were often upbraided for aggressively pursuing lucrative earnings via sales commissions.
It is fairly easy to explain the continued dependence on central government officials because they hold the greatest political resources as news information subsidizers and also are responsible (Gandy, 1982) for the heated policy. As the third-most frequently used, elites of the construction industry obviously did not need to aggressively vie for news media access. Their political clout is exerted and viewpoints adopted covertly but effectively through their economic and financial leverage (Shoemaker & Reese, 2014), accounting for some 30% of the market-driven (Bagdikian, 2004) newspapers’ advertising incomes in section-one pages. On the contrary, the statistically but unimportantly increased use of the general public as news sources helps reconfirm the indispensability of resources and accompanying media access and political influence (Koduah et al., 2016). As a whole, nonofficial sources in this study amounted to little more than nothing in selling their arguments, particularly when regular homebuyers—the most relevant group concerning the policy—lost more of its already slim ground. For them, dramatic strategies—highly eye-catching, newsworthy acts—are perhaps necessary to acquire “back-door” news media access (Wolfsfeld, 2004). In short, these findings effectively reflect Entman’s (1993) claim about frames as the “imprint of power” (p. 55) in the frame contest (Benford & Snow, 2000).
While our findings fail to demonstrate the powers imprinting the contested frames due to our inconsistent uses of measurement levels, the pattern of news coverage provides insights into the framing of the issue, with competing news frames moving exactly the same way as does the use of news sources. As frames help define a contending social problem (Entman, 1993), use of the pro-market frame and the shifting environment/mentality sub-frame outweighed the overall pro-social equality frame and thus legitimized a booming and prosperous housing market. It is thus reasonable to believe that many were brainwashed to quickly jump onto the booming-market bandwagon when potential homebuyers’ (and surely the speculators’) mentality about the housing market was being willy-nilly restructured. More critically, the clear downplaying of the worsening M-shaped society sub-frame inevitably confined the countering arguments and understanding of the causes of rising housing prices.
Moreover, the four newspapers appeared to be responsive in shifting their news frames over time, but in different degrees and for different reasons. The overall favorability of the pro-market frame significantly declined in the China Times and Liberty Times. While the two newspapers are equally free from financial constraints, they are politically polarized. And it appears that this peculiarity is due much more to political considerations than to financial calculations. While the pro-KMT China Times reacted strongly to accommodate the call of the ruling Ma government under attack, the pro-DPP Liberty Times was very likely to aggressively use the mounting public complaints against the unstoppable housing prices to hammer the regime.
This finding is in line with political and communication theories. First, the conspiracy theory suggests that elites would maneuver to retain their status quo or gain a vantage point from a crisis. In response, these two newspapers function as the guard dog for their respective political party, promoting their viewpoints and news frames accordingly.
These crucial observations have strong implications for many Taiwanese—in individuals and those in academic disciplines—in that housing inequality is both an outcome and a main cause of a variety of associated social problems (Berkman et al., 2014; Bryant, 2009; Shaffi, 2017). Foremost, as Gasher, Ross, and Dunn (2007) wrote, the press that helps build news agendas and frames has relentlessly ignored or is unknowledgeable about the health determinants—housing as its core (Shaffi, 2017)—that facilitates explanation of the issue and understanding of how housing inequality can undermine individual health and well-being. Losing a house and therefore the bulk of one’s wealth (Wahowiak, 2016) can result in suicides (K. A. Fowler et al., 2015; Houle & Light, 2014). Soaring housing prices undercut a country’s birth rate and labor force as married couples cannot afford a home that meets their basic needs (Chen, 2013), leading Taiwan to become the nation with the lowest birthrate worldwide (Population Reference Bureau, 2012) and also lag behind in worldwide talent hunt (Yahya & Kaur, 2010). Its work force with lower incomes and poorer housing is forced to routinely work more, suffering more job-related injuries (Kao, Spitzmueller, Cigularov, & Wu, 2016). The vicious impact of housing inequality as a consequence of income inequality and as a risk factor steadily widens social injustice, in terms of wealth, health, and well-being inequalities. Finding solutions to this significant social issue undoubtedly requires committed efforts from the academics in public policy, economics and finance (taxes), sociology, public health, and mass communication. It requires collaborative endeavors by interdisciplinary scholars to provide viable initiatives to fix this long-lasting problem. And the government, alongside other stakeholders, should acknowledge the dire impacts of housing inequality and commit to create a policy that is acceptable to the citizenry and fulfills this new social contract (Bratt, 2002).
While our dataset cannot correlate power (news sources) with the (frame) imprint, it effectively connects news sources with responsibility assignments, recommended by news frames (Entman, 1993). The findings show that official sources, particularly members of the real-estate and construction industries, criticized regular and potential homebuyers for resisting the altering nature and precarious environment of the housing market, repeatedly losing their opportunities to find homes. However, nonofficial sources accused the government of corruption and initiating harmful public policies—slashing the inheritance tax—that made housing prices spiral upward. As a result, the latter camp of social actors demands that the government find ways to rein in market speculations. In light of Entman’s (1993) framing notion, power-wielding political and business elites as dominant communicators master interpretations of social issues favorable to their position and blame the pig-headed public as a loser in this zero-sum capital game. Resource-constrained nonofficial sources, however, blamed the central government as the culprit and being responsible to fix the problem. This finding raises a question about the rank order of the four framing functions—problem definitions, cause diagnoses, moral interpretations, and treatment recommendations (Entman, 1993)—in conjunction with the power location or news sources. If assigning responsibilities (also consider Iyengar, 1991) is the ultimate function and goal of framing, then this aim may have failed in the overall coverage because nonofficial sources seemed to be more efficient and effective than their counterparts, who could only effectively assign causal responsibility, but not both. Future empirical and theoretical research on this aspect is merited to conceptually refine framing.
A few limitations and suggestions for future research need to be further addressed. This study used four major newspapers to represent all news media in Taiwan. Even though newspapers remain a major source of news for Taiwanese audiences (Lee & Fujioka, 2017), future scholarship should include additional types of news media. Second, our study period was August 2008 through July 2013. Expensive housing remains a problem, however, and an updated study might be necessary soon, if not already. Perhaps more critically, the varying levels of measurement used for content analyzing news sources and news frames eludes an effective correlation analysis, namely how the former builds the latter. A valid correlation analysis may prevent the pitfall of analyzing frames by simply quoting selective paragraphs here and there (Reese, 2007). This drawback in precision or even validity was duly compromised by our initial resource constraints in trying to manage the huge amount of data; greater effort and caution shall be taken in the future to enhance the theoretical discussions and usefulness. Next, as this study demonstrates, comparing shifts of news frames over time is a worthwhile approach, so we urge more scholars to adopt a longitudinal analysis more often.
In conclusion, this study examined a critical, multi-perspective social issue—the cost of housing—that framing scholars have not covered. The results confirm previous findings (e.g., Altschull, 1995; Entman, 1993, 2003) that, through framing, powerful social actors with a capitalist ideology dominate news coverage of this issue. Considering the less-than-desirable results that nonofficial sources had very marginal voices, journalists are strongly urged to take the side of the masses, instead of siding with business and political elites. Again, using a canine metaphor (Donohue et al., 1995), in a democracy, the news media should serve as people’s watch dog, instead of a guard dog or a lapdog for social actors and groups with resources and power. Otherwise, the housing price increases could eventually lead to enraged public uprisings against corrupt elites, as occurred when the 2011 Occupy Movement expanded to become the “99/1” social movement that created considerable social upheavals or even dethroned ruling governments.
