Blame Avoidance Strategies of Multiple Administrative Agencies During Policy Failure: A Case Study of Taiwan’s 2014 Waste Oil Scandal

Abstract

This study examines the blame avoidance (BA) strategies that may be adopted after policy failures of multiple agencies. A classification of agencies is proposed based on two factors: the amount of actual responsibility that an agency is obliged for, and the extent of an agency’s perceived responsibility. Agencies can thus be classified into four types, which may have different tactical goals and specific BA strategies that contradict other agencies’ goals, thereby leading to a complicated interaction between them. The results show that one of the three related agencies does not adopt suggested BA strategy and therefore fails to avoid blame.

Keywords

blame avoidance policy failure crisis waste oil

Introduction

Societies face various risks, including those related to the environment, food safety, labor safety, and financial order. To avoid or mitigate the potential damage due to these risks, governments must adopt specific public policies. However, if a significant incident happens and human factors are involved in the cause of the crisis, the media, public, politicians, and other supervisory agencies will likely question which agency is responsible and to what extent. Hence, an incident may cause the agency to be blamed, which may seriously damage its reputation. To protect the agency’s reputation, which is a precious asset (Carpenter & Krause, 2012), certain blame avoidance (BA) strategies can be adopted by the agency before and after a crisis erupts (Busuioc, 2016; Hinterleitner & Sager, 2019; Hood, 2010; Hood et al., 2009).

Since Weaver (1986), many scholars have studied the strategies used by agencies and politicians for BA. However, most studies focus on a single agency and its BA strategies before or after a crisis (Hinterleitner & Sager, 2015; Hood et al., 2009). Although some studies analyze the interactions between the trustee and trustor (Brändström & Kuipers, 2003; Hood, 2002) or examine the network failure of multiple agencies (Brändström & Kuipers, 2003; Moynihan, 2012), little attention has been devoted to the interactions between the culpable agencies. Thus, this article seeks to provide a deeper understanding of the BA strategies taken by multiple agencies.

In many circumstances, risks are managed by multiple administrative agencies. Several government agencies are usually involved in a government’s risk governance, seeking resolution through cooperation and division of labor (Hood, 2010; Hood et al., 2009). In BA studies, an agency may adopt strategies such as redefining the issue, passing the buck, finding a scapegoat, and circling the wagons (Weaver, 1986, 2013); these strategies imply that a crisis is caused by multiple agencies, making it possible to pass the responsibility to other agencies or to find a scapegoat, but only when they are involved.

This study’s main purpose is to identify BA strategies adopted by multiple government agencies after a crisis caused by a policy failure, the definition of which is a policy program that has failed to achieve the results desired by the government, because of poor implementation by related agencies (Brändström & Kuipers, 2003; McConnell, 2010). The research questions are the following:

Research Question 1: What factors affect the BA strategies of these agencies?

Research Question 2: What are the BA strategies expected to be used by different types of agencies?

Research Question 3: What is the possible impact of the BA strategies of one agency on other agencies’ subsequent BA strategies?

To answer these questions, the staged BA model developed by Hood et al. (2009) is used to analyze the strategies used by various agencies, because their model can show the BA strategy changes when these agencies deal with a crisis. Next, this study proposes that the BA strategies expected to be used by government agencies vary systematically depending on levels of actual responsibility an agency is obligated to bear, and the perceived responsibilities the public and media believe an agency is liable to undertake. Each different combination of actual responsibility and perceived responsibility produces a different type of agency and its own specific BA strategy. For example, agencies with less actual responsibilities and more perceived responsibilities will face the greatest challenges, because of the huge pressure from the public and the media. They need to respond appropriately; otherwise, it is easier for them to become the scapegoat. They are expected to spend more resources to identify the true causes of the incident, transfer responsibilities to other agencies, admit their own small responsibilities, and perform some remedial activities. On the contrary, agencies with less perceived responsibilities and more actual responsibilities are fortunate and want to escape from the attention of the media or the public. They are not expected to transfer responsibilities to other agencies many times, because this may anger other agencies and attract the attention of the media and the public. While they are expected to take remedial activities whenever necessary, it is not easy for them to admit their own responsibilities unless it is unavoidable. Thus, this article contributes to a deeper understanding of the BA strategies used by different agencies, especially when the perceived responsibility of some agencies does not match their real responsibility.

To test our expectations, this study examines one major policy failure in Taiwan, the waste oil scandal in 2014. The incident is traced to an elderly farmer in Pingtung, who reported to various authorities the stench and pollution emanating from an underground factory. After investigations, the Pingtung District Prosecutor’s Office found that the underground factory collected waste oil, processed it, and then sold products to downstream oil factories, including the Qiangguan Company, Taiwan’s second largest lard oil supplier at that time. Lard is an important ingredient in Taiwanese pastries and cookies. This waste oil incident shocked society, because Taiwan’s last waste oil scandal occurred in 1985. In addition, before the waste oil crisis in 2014, several food safety incidents occurred in Taiwan, and public confidence in the government began to decline. The polarization of political elites at that time was very serious in Taiwan. Therefore, the incident was widely reported in the press from September 4, 2014 and generated more than 1,000 news reports in about 1 month.

Although waste oil products were manufactured by a private underground factory, this reflected a serious failure on the part of the government’s food safety policies. When the incident broke out, the public found that there were many food products made from waste oil, and so they worried about the health effects of these products and began to question the government’s capacity to handle the problem. For example, the public wanted to know how these products made from waste oil could pass government inspection standards? Why did other downstream manufacturers dare to buy products made from waste oil? Why has the government never found out that this was a problem? In addition, waste oil is a kind of waste and is expected to be managed by at least one government agency. How can waste oil flow into the underground factory without being uncontrolled?

In Taiwan, food safety risks are comanaged by the Taiwan Food and Drug Administration (TFDA) of the Ministry of Health and Welfare (MOHW), the Council of Agriculture (COA), and the Taiwan Environmental Protection Administration (TEPA). The TFDA of MOHW are responsible for the planning and implementing of food safety policies, including food import control, the regulation and examination of additives, and the random inspection of food manufacturers. COA is responsible for the safety of agricultural produce from its production to the wholesale market. TEPA is responsible for the protection and supervision of environmental and ecological issues. It also established a system to guard against toxic chemical substances from being added into food products. These agencies adopted various BA strategies to switch the responsibility and attention of the media and the public to other agencies. During this crisis, the heads of MOHW and TFDA resigned (TFDA & MOHW, 2014).

The case analyzed in this study can be categorized into the “presentation strategy” mentioned by Hood (2002, 2007, 2010) and the “reactive BA strategy” as proposed by Hinterleitner and Sager (2017). In addition, from the perspective of a case study design, although we think that there are many similar cases around the world, few scholars have investigated the interactions of multiple agencies throughout a crisis. Hence, this case serves as an illustrative or revelatory case and provides useful descriptive information (Yin, 2009) for testing our expectations.

This study proceeds by first identifying the factors affecting agencies’ BA strategies by means of a literature review, and then proposing the BA strategies expected to be adopted by different types of agencies. The research strategies, data, and method are then introduced, and the results, discussion, and conclusion presented.

Factors Affecting the BA Strategies of Multiple Agencies

When an agency chooses its BA strategies, it considers many factors (Hinterleitner & Sager, 2015), which can be classified into four categories, as listed below:

(1) The characteristics of the crisis, such as the extent of negative impact on society and the salience of the crisis event (Gilad et al., 2013; Hinterleitner, 2018; Hood, 2010; Hood et al., 2009, p. 714);

(2) The political and economic environment, such as the governmental structure (Hood et al., 2009), and the manipulation by the media and politicians (Brändström & Kuipers, 2003; Liu et al., 2012; Schillemans, 2016; Weaver, 2013);

(3) The factors related to the agency level, such as the main cause of the crisis (Boin et al., 2010; Brändström & Kuipers, 2003; Gilad, 2012; Hinterleitner & Sager, 2015; Hood, 2010; Maor et al., 2012), the perceived connection between the crisis event and the agency (Hinterleitner, 2018; Hood, 2010; McGraw, 1991; Mortensen, 2013), the cost of risk control (Gilad et al., 2013; Weaver, 2013), and the reputation, resources, and information of the agency (Boin et al., 2010; Busuioc & Lodge, 2016; Carpenter & Krause, 2012; Gilad et al., 2013; Hood, 2010; Maor et al., 2012; Weaver, 2013);

(4) The agency’s leaders’ characteristics and leadership style (Boin et al., 2010; Hood, 2010).

The main purpose of this article is to highlight the differences between the BA strategies of multiple agencies involved in a crisis, thus, the analysis unit is an agency. As a result, the factors related to the characteristics of the crisis, the political and economic environment, and leaders’ characteristics or leadership styles is not considered because these factors are the same for the agencies involved in a crisis or difficult to identify at the agency level. Only the factors related to the agency level are considered.

Of the factors at the agency level, Hinterleitner and Sager (2015) argue that the truth, accessible information, interpretation of the information acquired, and public perception affect the possibility and extent of the blame. Therefore, this study takes this observation as the starting point, and uses two factors: the amount of actual responsibility an agency is obliged to undertake and the amount of perceived responsibility the public and media think an agency is liable to bear, to classify agencies. In addition, these two factors are chosen because they affect the BA strategies of the agencies throughout the whole process of the crisis. When considering the BA strategies adopted by agencies and the interaction between these agencies during the crisis, information and interpretation play an important role. These two factors are detailed below.

The first factor is the amount of actual responsibility an agency is obliged to undertake during the crisis; this reflects the importance of the truth. After a crisis has erupted, and as an investigation progresses, more information will be acquired and the truth will gradually emerge. The media and the public will find out about the agency that is responsible. If an agency is obliged to be responsible for a crisis based on the truth, it is difficult for it to evade blame (Boin et al., 2010; Brändström & Kuipers, 2003; Gilad, 2012; Hinterleitner & Sager, 2015; Hood, 2010; Maor et al., 2012). This will affect an agency’s BA strategies.

The second factor is the amount of perceived responsibility the public and media think an agency is liable to take. This reflects the importance of information, interpretation, and perception (Brändström & Kuipers, 2003; Hinterleitner & Sager, 2017, 2019; Hood et al., 2016). Many scholars argue that when handling policy failures, the media has an especially important role to play, because the contents reported, means of interpretation, and framing adopted all affect the public’s perception of the event, including the range of fallout and its magnitude, cause of the crisis, government agencies involved, and agencies and figures that should be held accountable (Brändström & Kuipers, 2003; Djerf-Pierre et al., 2013; Hinterleitner & Sager, 2015).

When a crisis occurs and it involves multiple agencies, the media and the public may lack information and be unaware of the true causes at the beginning of a crisis, because of institutional fragmentation or the complexity of the co-governance structure. The media and the public are unsure which agency is responsible (Hinterleitner & Sager, 2017; Moynihan, 2012). As a result, the media may hold an agency accountable depending on several factors, including the information in press reports, intuition, common sense, the matters under an agency’s control, and the distance between the agency and the public (Boin et al., 2010; Brändström & Kuipers, 2003; Djerf-Pierre et al., 2013; Hinterleitner, 2018; Hinterleitner & Sager, 2015, 2017; Hood, 2010; Liu et al., 2012; Schillemans, 2016; Weaver, 2013). Consequently, this affects the perception of the public and media, and the agency with more perceived responsibility may face more questions, criticism, investigation, and bear more pressure, all of which may affect their BA strategies.

Based on the above two factors, this study classifies agencies into one of the four types shown in Table 1.

Table 1.

Types of an Agency in a Crisis Event.

Agency’s amount of actual responsibility	Agency’s amount of perceived responsibility
Agency’s amount of actual responsibility	Large	Small

Large	Culpable agency	Fortunate agency
Small	Disadvantaged agency	Peripheral agency

Source. Prepared by the authors.

Table 1 shows that when multiple agencies are involved in a crisis, these agencies can be classified into four types. These different types of agencies are expected to adopt different BA strategies, as shown in the following section.

BA Strategies Expected to Be Used by Different Types of Agencies

Scholars have proposed different classifications of BA strategies adopted by agencies or politicians. For example, Weaver (1986) argues that politicians can employ eight different BA strategies. Hood (2002, 2007, 2010) proposes three different types of BA strategies: agency strategies, presentational strategies, and policy strategies. In addition, other scholars divide BA strategies into two categories: anticipatory and reactive strategies (Hinterleitner, 2017; Hinterleitner & Sager, 2017). Hood et al. (2009, 2016) propose that the agency of concern will adopt staged retreat BA strategies after the outbreak of an incident. Specifically, when the agency chooses the BA strategy, it will use the method of trial-and-error, by first adopting more convenient, lower cost, and more favorable BA strategies. If these BA strategies are not effective, then the agency will be forced to adopt BA strategies that are costlier or may damage its reputation (such as taking responsibility).

This article uses the BA classification developed by Hood et al. (2009) for analysis, because their model can be used to examine how a government agency changes its BA strategy over time. They divide the BA strategy into three major categories: problem denial (PD), problem admission but responsibility denial (PA+RD), and problem and responsibility admission (PA+RA).

Each major category has three subcategories. For PD, the three subtypes of BA strategies are: A1, pure PD; A2, admission that there is some problem, but denial that the problem is serious; and A3, PD accompanied by a counterattack. For PA+RD, the three subtypes are: B1, an open stance on investigating who is responsible; B2, assert others to take responsibility; and B3, admission of some responsibility, but denial of major responsibility, or responsibility denial by offering excuses and accepting only technical and minor responsibility. For PA+RA, the three subtypes of BA strategies are: C1, explanation-only response, but not accepting culpability; C2, institutional action-taking response and offering institutional apology, compensation, and remedial action; and C3, admission of personal culpability, which may include resignation or an acknowledge of error.

The ultimate goal of the above four types of agencies is the same—that is, to protect the reputation of the agencies by minimizing the danger of adverse reaction (Busuioc, 2016; Carpenter, 2002; Maor, 2010), which this study measures by the number of news reports criticizing them. However, their tactical goals and BA strategies differ. The tactical goal of the BA strategy of culpable agencies is to shift responsibility to the agencies that the media or public considers responsible and to make them the scapegoat, or at least make them accept a share of the blame. Hence, they may delay the investigation or mess up the information and then transfer the responsibility to disadvantaged agencies. Because culpable agencies are those that are mainly responsible for the crisis, they may make no response or admit responsibility outright when certain information indicates that they are responsible; otherwise, more criticism will follow. Thus, B2 (assert others to take responsibility), C2 (offering institutional apology, compensation, and remedial action), and C3 (admission of personal culpability) strategies are expected to be adopted by culpable agencies. B1 (an open stance on investigation) and C1 (explanation-only response) strategies are not expected to be used often.

The tactical goal of the BA strategies of disadvantaged agencies is to avoid being made the scapegoat. Significant pressure from the media and public means that these agencies must answer many questions, respond to criticisms, and deal with the investigations in a short time. The best strategy is to move responsibility over to the culpable or fortunate agencies. This is not easy, because it is only possible if the agency can show the true causes of the crisis. As a result, these agencies need to stay calm, keep information transparent, and try to find the truth as soon as possible. In addition, they must avoid doing something wrong, such as making verbal faux pas, becoming the center of the crisis or interrupting the investigation. From the perspective of the staged BA model, it is expected that disadvantaged agencies will use many types of BA strategies, such as B1 (an open stance on investigation), B2 (assert others to take responsibility), B3 (admission of some minor responsibility), C1 (explanation-only response), and C2 (offering institutional apology, compensation, and remedial action) strategies.

The main tactical goal of the BA strategies of fortunate agencies is to avoid becoming the center of media attention. Even when these agencies are obliged for much of the actual responsibility, they may escape the attention of the media and public for various reasons. Therefore, their best strategy is to keep a low profile, and they will not take any activities that may stimulate or anger the public, the media, or other agencies. They will try to delay the investigation and to not take responsibility on their own. They will admit responsibility only when certain information indicates that they are responsible and they cannot escape from it. They sometimes may admit some minor responsibility or make some indefinite explanations to increase the vagueness of the crisis. Hence, it is hard for the fortunate agency to be held accountable unless other agencies provide reliable information. As a result, the fortunate agencies are expected to adopt BA strategies, such as B3 (admission of some minor responsibility), C1 (explanation-only response), and C2 (offering institutional apology, compensation, and remedial action). They are not expected to use A2 (denial that the problem is serious), B1 (an open stance on investigation), or B2 (assert others to take responsibility) often.

The tactical goal of the BA strategies of peripheral agencies is to avoid attention from the media and public. They have fewer actual responsibilities, meaning that if any information indicates their culpability, they may acknowledge it and make corrections quickly to reduce possible blame. They thus may adopt BA strategies, such as B1 (an open stance on investigation), B3 (admission of some minor responsibility), C1 (explanation-only response), and C2 (offering institutional apology, compensation, and remedial action). They are not expected to adopt B2 strategy, because there is no need to transfer responsibility, and this strategy may anger other agencies or the public.

The above propositions show that different types of agencies have different tactical goals and BA strategies. For example, disadvantaged agencies want to expedite the investigation and to make the information complete and transparent, while culpable and fortunate agencies want to delay the investigation and make the information murky. Hence, the culpable and fortunate agencies may reluctantly respond to demands for transparency and openness, which is in line with the findings of Hood and Rothstein (2001). However, the disadvantaged agency would like to have more transparent information. The above findings show that different types of agencies have different attitudes toward information transparency and openness.

Another example is that the difficulty of these agencies admitting their own responsibilities varies. It is easier for peripheral agencies to admit their responsibility, but it is very hard for fortunate agencies to admit their own responsibility. In addition, it is expected that fortunate and peripheral agencies will not often adopt the B2 strategy, because it may attract the attention and counterattack of other agencies and make them the focus of crisis attention. In other words, not every agency will try hard to shift responsibility to other agencies. Finally, the BA strategies adopted by one agency may contradict other agencies’ tactical goals or BA strategies, leading to a complicated interaction between them. To test our expectations, the case study chosen is the waste oil scandal in Taiwan.

Research Strategies, Data, and Method

To test the above expectations about the BA strategies used by different types of agencies, three elements are required. First, the actual causes of a crisis must be known and the degree of actual responsibility of the agencies involved must be reviewed. Second, the degree of perceived responsibility of the agencies involved must be analyzed. Third, the BA strategies adopted by different agencies must be identified. Only when these three different elements are ready can the whole picture of the BA game between these related agencies become clearer.

To investigate the real cause of the waste oil crisis, this study first collected relevant government documents, such as the investigation report by the Control Yuan, the indictment by the district prosecutor’s office, and the verdict of the courts. This helps to understand the causes of the crisis as well as the interaction of government agencies during the incident. In-depth interviews were then conducted to further confirm the whole picture and the true causes of the entire incident. The reasons behind the actions taken by these agencies were also explored.

With regard to the choice of interviewees, the officials of the key agencies were selected based on the government documents mentioned above. The authors conducted interviews between October 2015 and March 2016. In total, 17 government officials were interviewed, most of whom were responsible for the planning, implementation, and audits of food safety policies. This study also interviewed an investigator and a prosecutor related to the investigation of the crisis.

To explore the perceived responsibility of the agencies and their BA strategies, this study collected and analyzed secondary data such as newspaper articles and official press releases.¹ For newspaper articles, the United Daily News (UDN), China Times (CT), Apple Daily (AD), and Liberty Times (LT) were selected because they are the largest news groups and have different political stances. When the crisis erupted in 2014, the Kuomintang (KMT) was in power. Sorted in the order from highly supportive to highly unsupportive of the government is the CT, UDN, AD, and LT, where the first two newspapers were supportive and the last two unsupportive.

The research time period is from September 1, 2014 (3 days before the waste oil crisis erupted) to October 7, 2014.² Through the news search system of these newspapers, keywords such as “waste oil”, “Guo Liecheng,” and “Qiangguan” were used to manually search and collect news reports related to the incident. After excluding duplicate and irrelevant news, 1,295 articles in total were selected and used as the basis for the analysis. Each article was analyzed by one of the authors to identify the administrative agencies mentioned and to determine whether the agencies mentioned were criticized or blamed. If a report contains words related to criticizing or condemning government agencies, such as criticizing, questioning, evading responsibility, lying, concealing truth, and misconduct, then the report is classified as having a negative view or attributing responsibility to the government agency. If a news article just describes the facts, fairly reports how the incident was handled, or presents some suggested remedial actions, then the report will be classified as not attributing responsibility to an agency.

This study next classifies the BA strategies into nine categories based on the staged model proposed by Hood et al. (2009), as mentioned in the above section, to examine the BA strategies adopted by agencies at different time points. The BA strategies used in the news articles and official press releases were identified, classified, and assigned codes by two of the authors. Cohen’s kappa coefficient (Cohen, 1960), which measures the agreement between two raters, is 71%, which is acceptable. If the coding results of the two coders were inconsistent, then we generated the final codes through a discussion. A figure was produced to show the number of critical news reports and the BA strategies adopted for each agency. Next, the BA strategies of the agencies were qualitatively analyzed, because Hood et al. (2009, p. 712) suggest that it is more appropriate to use a qualitative approach for the examination of changes in the exercise and time sequence of BA strategies.

Research Findings

True Causes of the Waste Oil Scandal in Taiwan

The research results show that the cause of the waste oil crisis can be traced to the suspension of the “Green County Application and Promotion Plan” (GCAPP) of the Ministry of Economic Affairs (MEA) in May 2014.

To promote energy diversification, MEA implemented the GCAPP in July 2007, so that waste oil could be made into biodiesel, and then blended into diesel at a ratio of 1% to produce B1 diesel. In response to this policy, TEPA established an oil recycling system to collect and control the waste oil produced by big chain stores and food processing plants. The big chain stores are defined as those with a capital of more than 50 million New Taiwan Dollars (NTD) (about US$1.6 million) and the big food processing plants are those with a capital of more than 5 million NTDs (about US$0.16 million). TEPA defined the waste oil generated by these big chain stores and food plants as “industrial waste” and included it in the scope of the Trash Clearance Act. The flow of the above waste had to be reported by enterprises and transported to legitimate waste oil recycling plants. However, waste oil from households, schools, food stalls, and small catering businesses fell outside this control, creating the first loophole in TEPA’s policies.

Under the success of the B1 biodiesel policy, MEA elevated the ratio of biodiesel in diesel in July 2010 from 1% to 2%for B2 diesel. However, B2 diesel caused clogging in vehicle filters and shutdowns, so MEA suspended the B2 biodiesel policy in May 2014.

After GCAPP was suspended, MEA did not collaborate with TEPA to deal with the flow of waste oil. Consequently, TEPA terminated the waste oil recycling program. Waste oil from the big chain stores and food processing plants had nowhere to go, and its market value fell dramatically. In addition, the big chain stores and food processing plants were no longer required to declare how they disposed of their waste oil, so corrupt businesses took advantage of this loophole to collect waste oil from this market. This was the second loophole of TEPA.

In this case, the underground oil factory was accused and convicted of guilt, because it collected and processed waste oil and sold oil products to the market at a low price. Of the government agencies, TEPA was the one to have more actual responsibility, because of two loopholes in its regulations. In addition, TEPA was reluctant to clarify the division of responsibilities in gray areas and tried to avoid its obligatory responsibility, because the cost for it to manage waste oil produced by food vendors, night markets, and small catering operators was very high. With the persistent coordination of the upper level of government, the Executive Yuan (EY), TEPA began to enlarge the scope of its duties by establishing a new management system to bring all waste oil under control. This shows that TEPA took up most of the responsibility for the waste oil scandal.

The responsibility that the TFDA had was relatively small for two main reasons. First, it was difficult for the TFDA to determine whether the ingredients were legal through tests. Because of the sophisticated processing technology of the underground factory, products produced from waste oil can pass the test standards set by the government. In other words, it was impossible at that time for the TFDA to prevent waste oil from entering the oil production chain without the assistance of TEPA. Second, the TFDA was not the agency that caused the crisis. The loopholes in TEPA are the causes of the crisis. Although the TFDA was responsible for checking the safety and quality of oil products, if TEPA had controlled the flow of waste oil at the source, the incident would not have occurred.

However, the TFDA is still partly to blame because at the time, it did not have a good system for tracking food products and raw materials completely and quickly. This represented a policy failure.

As for COA, it played a very small role in this crisis. During the incident, COA was responsible for tracing the production and sales chain of fodder oil. It was found that one of the customers of the Guo underground factory sold oil products to seven animal feed traders. COA stated that even if waste oil had been mixed into fodder oil, it was used as poultry feed, and the impact on human health would have been negligible.

Responsibility Attributed to Administrative Agencies by News Articles

Which agency was perceived to be the most responsible for the crisis? This study counted the total news articles of each outlet and the number of articles in which administrative agencies were criticized and held responsible. The results are shown in Table 2. Among the 1,295 news articles, 572 news featured attributions of responsibility, representing 44.2% of all reports. Among these, LT had the largest number of reports (603 articles) and the largest number of news articles holding administrative authorities responsible (243 articles), perhaps due to the fact that LT was the media that held the most critical position toward the ruling party.

Table 2.

News Groups and Responsibility Attribution.

News Outlet	With attribution of responsibility		Without attribution of responsibility		Total
News Outlet	Number of articles	%	Number of articles	%	Number of articles
UDN	146	51.6	137	48.4	283
CT	110	41.7	154	58.3	264
AD	73	50.3	72	49.7	145
LT	243	40.3	360	59.7	603
Total	572	44.2	723	55.8	1,295

Note. UDN = United Daily News; CT = China Times; AD = Apple Daily; LT = Liberty Times.

Source. Prepared by the authors.

Table 3 shows that the TFDA and the MOHW were the most frequently referenced agencies (522 articles) and the ones criticized the most (114 articles), perhaps because the TFDA was primarily in charge of handling the scandal. After the crisis erupted, it was obliged to test the ingredients in waste oil products, track the flow of inferior oil products, pull them off store racks, collect relevant evidence, and enforce punishment. Because these actions were closely related to people’s daily lives, the press followed the TFDA’s actions closely. In addition, the TFDA and MOHW had the largest number of critical news articles. This shows that the press or public regarded the TFDA as the major agency to be blamed for the scandal, perhaps because during the crisis management, the TFDA did not adopt a proper BA strategy (described in the following section). This exacerbated the public and media’s scathing judgments against the TFDA.

Table 3.

Public Attribution of Responsibility to Administrative Agencies in News Reports.

Admin body	Number of articles with reference to the admin body	Number of articles with a reproachful stance	Proportion of critical news
TFDA and MOHW	522	114	22%
TEPA	147	54	37%
COA	78	21	27%

Note. TFDA = Taiwan Food and Drug Administration; MOHW = Ministry of Health and Welfare; TEPA = Taiwan Environmental Protection Administration; COA = Council of Agriculture.

Source. Prepared by the authors.

Table 3 shows that the TEPA came second in terms of the number of times mentioned (147 articles) and criticism received (54 articles), indicating that it was difficult for the media and public to link the crisis to environmental protection authorities. However, as more information was disclosed, the number of reports criticizing TEPA increased. Closer examination of the number of critical news reports showed that although fewer reports attributed responsibility to TEPA rather than to TFDA, in terms of ratio, it had the highest proportion of critical news (37%). The COA had relatively fewer mentions (78 articles) and the fewest number of reproaches (21 articles).

Based on the above analysis of the waste oil incident, Table 4 shows the classification of the three main agencies involved and the private waste oil manufacturer.

Table 4.

Classification of agencies according to amount of real and perceived responsibility.

	Perceived responsibility
Real responsibility	High	Low
Real responsibility
High	Private waste oil manufacturers (culpable)	TEPA (fortunate)
Low	TFDA and MOHW (disadvantaged)	COA (peripheral)

Note. TFDA = Taiwan Food and Drug Administration; MOHW = Ministry of Health and Welfare; TEPA = Taiwan Environmental Protection Administration; COA = Council of Agriculture.

Source. Prepared by the authors.

BA Strategies of Multiple Administrative Agencies

Figure 1 summarizes day by day the number of press articles holding administrative agencies responsible and the BA strategies adopted by the three major agencies. For each agency, there is one figure. The vertical axis on the left and the rectangles in Figure 1 signify the blame level, represented by the number of news reports attributing responsibility (criticizing) to an agency on a specific day, and the vertical axis on the right and the black spots represent the agency’s BA strategies. The BA strategies of each agency are detailed in the following section.

Figure 1.

Number of critical news reports and BA strategies used by different agencies over approximately 30 days.

The TFDA

The rectangle in the upper part of Figure 1 shows that after the crisis erupted on September 4, the TFDA faced a vigorous clamor of blame. This peaked on September 7 with 17 reports. At the outset of the incident, the TFDA adopted three types of BA strategies: naming it a small problem (A2), shifting responsibility to private oil manufacturers (B2), and describing the business without taking responsibility (C1), as shown by the black spots in Figure 1. It shows that the TFDA initially used multiple BA strategies simultaneously under the pressure of time. Except for A2, the BA strategies used were as expected. These strategies are analyzed below.

First, the TFDA adopted a typical A2 strategy at the beginning of the incident as described below,

Yeh Ming-gung, head of the TFDA, said yesterday that the oil products were only on sale for six months and not everyone goes to the night market to eat every day; therefore, “that was not as harmful as expected.” (Reported by UDN 9/6)

MOHW emphasized that after the investigation, the consumption of waste oil was very low and would not cause much harm to the human body. (MOHW press release, 9/7)

The purpose of the TFDA statements was to dispel the public’s concerns. However, this strategy backfired. The public found it difficult to accept this explanation because no one wants to eat food made from waste oil. This BA strategy aroused public anger and made the TFDA a target of public criticism. As a result, reports criticizing TFDA peaked on September 7. Accordingly, the A2 strategy was no longer used after September 8.

Second, in terms of the BA strategy to shift responsibility, TFDA initially tried to deflect the blame to the private manufacturers. Although the underground factory was blamed the most by the public, the B2 strategy was not effective because the public thought that at least one administrative agency would be held responsible for the crisis. Only on September 10 did the TFDA first blame MEA and TEPA. Chiang Yu-mei, the then Deputy Director of the TFDA, stated,

In 2007, the MEA promoted the biodiesel project, and the TEPA implemented the “Compulsory Recycling Regulations for Waste Oil.” The waste cooking oil of big chain stores and food processing plants must be recycled and declared . . . But after the project was suspended, these companies were not required to recycle and report waste oil any more. Jiang Yumei said that unscrupulous companies added the illegal substances to the food products . . . (the TFDA press release on 9/10)

As a result, the media started attributing responsibility to TEPA. The number of accusatory reports rallied against TEPA began increasing from September 10, and peaked on September 12. The TFDA’s strategy of shifting responsibility to TEPA was effective. In fact, the number of news reports accusing the TFDA started decreasing from September 10. The second peak on September 12 and 13 happened, because the head of MOHW offered some explanation to avoid personal responsibility on September 11. However, even with the second peak, the number of news reports criticizing TFDA declined dramatically after September 10.

Third, at the beginning of the incident, the TFDA adopted a C1 strategy on September 5, namely to provide explanatory responses and illustrations. However, it did not acknowledge the responsibility. Its main argument was that the ingredients in waste oil could not be found through tests. Therefore, the TFDA could not know what substances the manufacturers used:

The Deputy Director of the Center of Regional Administration of TFDA, Xu Jing-feng, said that . . . the ingredients in waste oil were yet to be identified, . . . Even with tests, it is difficult to confirm the ingredients. (Reported by AD, 9/5)

The use of the C1 strategy by the TFDA to avoid blame seems to be ineffective because the media and public do not have enough scientific knowledge to make a judgment, and it seems to them that the TFDA tried to avoid its responsibility.

In addition to the above three BA strategies, the TFDA subsequently adopted four other BA strategies, more than those of TEPA and COA. This shows that under the pressure of media attention, the TFDA tried to find ways to respond with different BA strategies by adopting a trial-and-error approach considering the reaction of the public and media responsibility the next day. This reaction aligns with the findings of Hood et al. (2016).

Finally, as the incident developed, the TFDA gradually shifted its BA strategies from C1 to C2. In other words, the TFDA began to admit responsibility, apologize, and take improvement measures. Finally, the head of MOHW resigned on October 3 and took political responsibility to quell media blame for the incident, which was in line with the finding of Hood et al. (2016), underlining the fact that the admission of personal responsibility is the most effective way to calm the media’s accusations.

TEPA

At the beginning of the incident, the media offered very little news concerning the blameworthiness of TEPA. TEPA stated on September 5^th that no unusual and suspicious activities had occurred in its existing management systems. From September 5^th to 9^th, the main BA strategy adopted by TEPA was to recognize some responsibility (B3) and to explain things, while trying to evade full responsibility (C1). For example,

Officials of the TEPD said yesterday that waste oil was flowing through the whole of Taiwan. Much of the kitchen waste was collected by illegal waste cleaning companies. How many “black operators” were there? What is the hidden amount of waste oil? The TEPA admitted that it was very difficult to grasp the scale of this black market. (Reported by CT on 9/7)

TEPA’s statements show the loopholes in the implementation of the recycling mechanism for waste oil. In addition, this statement may have aroused the attention of the TFDA. As a result, the TFDA shifted responsibility to TEPA on September 10, which increased the media coverage of TEPA, resulting in a number of critical news reports that peaked on September 12. This prompted TEPA to adopt a C2 strategy on September 12, namely to acknowledge responsibility and to propose reform activities, as follows:

TEPA yesterday invited the county and city environmental protection bureaus to hold a meeting and announced that waste oils from stalls, night markets, and home kitchens will be included in the waste oil recycling program of environmental protection agencies starting next month. The day after tomorrow, the county environmental protection authorities will report to TEPA the results of their daily inspection regarding the flow of waste oil from stalls and food sellers. (AD, 9/13)

This indicates that, before the outbreak of the crisis, the amount of waste oil controlled by TEPA was relatively small and that there was a need to improve the management program. After being criticized, TEPA quickly turned to C2 strategy, and the number of reports against it gradually decreased, proving that the strategy was effective. Thus, this strategy became the backbone of TEPA’s response. Other strategies such as C1 and B2 were only used sporadically.

Clearly, TEPA’s BA strategy changed very quickly. However, it was not without challenges because the waste oil recycling management system is very costly and labor-intensive. Because of strong intervention from upper levels of government, the EY, TEPA was charged to broaden its management of waste oil and to adopt related reform activities.

To summarize, at the beginning of the incident, TEPA adopted B3 and C1 strategies several times between September 5 and 9. TEPA tried to explain the problem, while denying responsibility (C1). However, these strategies did not achieve the anticipated effects. After the TFDA’s statement on September 10, the media gradually considered TEPA to be responsible for the flow of waste oil, forcing it to adopt strategy C2 and to make the corresponding reforms.

COA

In this incident, COA played a very small role. However, when the incident occurred, COA initially adopted A1 (pure denial) and B1 (open stance on investigation) strategies. For example, on September 4, COA explained that Jinwei Co., one of the customers of the Guo Lie-cheng underground factory, was not in the business of fodder trading, so it was not under the jurisdiction of COA. This was a strategy of outright denial (A1). In addition, COA had an open stance on investigation because it only had a small role or limited responsibility in this crisis.

As the incident developed, COA was required to trace the production and sales chain of fodder oil. The investigation found that Jinwei Co. (a client of Guo Lie-cheng underground factory) supplied seven animal feed traders. However, at a press conference on September 9, COA refused to publish the list of the seven fodder traders. This attitude irked the public and media, as it seemed that COA was hiding something from the public and press. As a result, news reports holding COA responsible began increasing. In response, COA published the list the next day (9/10). In addition, COA also adopted the A2 strategy and argued that even if waste oil had been mixed into fodder oil, it was used as poultry feed, and the impact on human health would be negligible.

Discussion

TFDA, as a disadvantaged agency, was expected to spend more resources on the investigation of the true causes of the crisis, move responsibility over to other agencies, and avoid saying the wrong things. Unfortunately, its inaccurate remarks on September 6 made it the object of rebuke. When it tried to shift responsibility to TEPA on September 10, it was too late, and the TFDA became the scapegoat in the scandal for three possible reasons.

First, at the beginning of the incident, the TFDA tried to deal with the incident by taking a professional and technical perspective. Accordingly, it spent much time examining whether the oil products did indeed contain substances harmful to human beings and tried to explain its findings by citing scientific evidence. This may be because the head of TFDA was Yeh Ming-gung, who was once a professor in the department of pharmacy at a university. His characteristics and background may have pushed him to try to solve problems in a professional and scientific way. This could have affected the BA strategies used by TFDA and had an impact on subsequent BA strategies. Scientific evidence might show that waste oil products are not too harmful to people’s health, but the A2 strategy is still inappropriate, because people may buy products made from waste oil, and waste oil is very dirty and smelly. Thus, the A2 strategy backfired, because TFDA failed to grasp that the core of the incident was the illegitimate source of the oil products, not the health effects of the products. In addition, it missed the opportunity to find the true causes of the crisis.

Second, the TFDA seldom adopted B1 (open stance on investigation) or B3 (admission of minor responsibility) strategies. At the beginning of the incident, the public and media did not know which agency is obliged to take responsibility for the crisis. TFDA could have adopted B1 and B3 strategies to show that it had a positive and responsible attitude, giving it more time to obtain more information and find the true causes of the crisis.

Third, it was too late for the TFDA to shift responsibility to other government agencies. From September 5 to 9, the TFDA only tried to shift responsibility to the private waste oil manufacturer and related businesses. The TFDA overlooked the fact that the public was infuriated and needed to channel that anger toward a government agency. Therefore, the TFDA is expected to have found other government agencies earlier to share the blame.

Regarding TEPA, this study contends that it was the government agency that was most responsible for the crisis. However, it was not blamed the most. TEPA’s BA strategy was effective for the following three reasons. First, the failure of the TFDA’s BA strategy was the main reason why TEPA avoided blame. The TFDA did not quickly shift responsibilities to TEPA, so TEPA can evade the attention of the media and the public for a period of time. Second, TEPA quickly adopted the C2 strategy after the media started criticizing it. The agency announced on September 12 that it would bring the disposal of the waste oil produced by food vendors, night markets, and small catering operators under control. On September 13, the head of TEPA inspected the related firms and proposed further amendments to the law. This effectively reduced TEPA’s share of the blame from the media. Third, over the course of the incident, TEPA did not adopt a strategy to deny the problem or shirk responsibility. B1 was used only once at the beginning, and B2 twice. The strategies mostly adopted were B3, C1, and C2, as this article expected. However, TEPA did not take the initiative to admit its responsibility. It adopted C2 only after the TFDA shifted blame and after the consequent increase in critical news reports. In sum, TEPA avoided becoming the center of media attention, and acknowledged the responsibility only when unavoidable. Its BA strategy was consistent with the expected strategy.

The role of the COA in the incident was relatively minor, and it received less blame from the media and public, making it a peripheral agency in the crisis. As a result, COA seldom made any responses. However, noteworthy is its adoption of a C2 strategy to alleviate public fury as soon as blame emerged from the media.

Past research shows that the characteristics of the crisis and environment as well as the agencies’ leadership style may affect the BA strategies adopted by them (Gilad et al., 2013; Hinterleitner, 2018; Hinterleitner & Sager, 2015, 2019; Hood, 2010; Hood et al., 2009). The above discussion shows at the beginning of the incident that the characteristics of the head of TFDA affected the BA strategies used by the agency and had an impact on subsequent BA strategies. In addition, because the crisis involved many manufacturers and because the public lost trust in the government, it caused great public concern and the BA strategy of PD (A1–A3) is not appropriate to use. When the crisis occurred, the government agencies received a lot of blame from the media and the public, and the media and public wanted to hold at least one government agency accountable. As a result, the interactions of the BA strategies adopted by the agencies became very fierce, because at least one agency was being targeted for sole responsibility.

Conclusion

This study examined the BA strategies that are expected to be adopted after a crisis arises from a policy failure caused by multiple agencies. Most previous studies have examined the BA strategies of a single agency and the factors affecting the choice thereof. The main contribution of this article is presenting a classification of agencies comprising two factors: the amount of actual responsibility an agency is obliged to undertake and the extent of its perceived responsibility. Although the ultimate goal of these agencies is the same—that is, to protect their reputation by reducing the news reports criticizing them—different types of agencies have their own BA tactical goals and specific strategies. For example, the fortunate agencies are expected to avoid becoming the focus of media attention, deter any investigation, admit some minor responsibilities, provide some explanations, make some corrections, and acknowledge their major responsibility only when unavoidable. The disadvantaged agencies are expected to stay calm, find the true causes of the crisis, shift responsibility to other agencies with more actual responsibility, admit some minor responsibilities, take some remedial action, and avoid saying the wrong things. The peripheral agencies are expected to avoid media attention, have an open stance on investigation, acknowledge their responsibility, and make corrections quickly.

The BA strategies of different types of agencies mentioned above may contradict each other. The gain of one agency may mean the loss of another agency. For example, the suggested BA strategy of disadvantaged agencies is to find the real causes of the crisis and transfer responsibility to other agencies. However, this is not an easy task, because the culpable or fortunate agencies will not accept it and will try to deter an investigation or shift focus to other agencies. Hence, the results of each agency’s BA behavior will be determined not only by its own BA strategy, but also by the BA strategies adopted by other agencies.

To test the expectations, this article conducts a case study on a crisis in Taiwan. The results show that there are three major administrative agencies related to the crisis. TEPA is classified as a fortunate agency, and the results show that its BA strategy is consistent with the expected strategy of avoiding becoming the center of media attention and acknowledging responsibility only for unavoidable circumstances. COA is classified as a peripheral agency, and its BA strategy is also consistent with the expected strategy—that is, there are few responses and corrections to be made immediately after being accused. TFDA is classified as a disadvantaged agency, and the results show that it fails to avoid blame and becomes the scapegoat, because it does not adopt the suggested BA strategy. For example, it failed to spend more resources to find the real cause of the incident and provided controversial statements. In addition, it was too late for TFDA to transfer responsibility to TEPA. Therefore, in this case study, the general patterns of the BA strategy adopted by two of the three related agencies and the interaction between them are consistent with the expectations developed in this article. Another agency, that is TFDA, did not follow the suggested BA strategy and therefore failed to avoid blame.

The findings of this study show that these agencies adopted staged BA strategies, as mentioned in Hood et al. (2009, 2016). These agencies used a number of BA strategies at a certain time point. Furthermore, there was a convergent trend toward admission of responsibility for each agency. However, the number and the types of BA strategies differ for various agencies. For example, it was found that the disadvantaged agency used the most BA strategies, while the peripheral agency used the least BA strategies and conducted remedial activities quickly.

As most crisis events are caused by multiple agencies with varying degrees of responsibility, the classification of agencies developed in this study has two important implications. First, it can help agencies understand the BA strategies used by other relevant agencies and enable them to use specific strategies to avoid blame. For example, if TFDA had commented correctly and adopted proper BA strategies, then TEPA would have faced greater pressure and TFDA would not have been the scapegoat.

Second, this classification has some implications for policy outcomes and interagency cooperation and coordination. The results show that related agencies will adopt various BA strategies to protect their reputation. However, this may have several shortcomings. For example, because the battle between agencies is fierce, the related agencies may prioritize reputation protection and thus, have less time and resources to deal with other tasks (Hinterleitner & Sager, 2019). The interactions between related agencies may weaken the mutual trust between agencies, hurt other agencies’ reputations, and decrease agencies’ incentive to cooperate (Ansell & Gash, 2008; Bryson et al., 2015; Emerson et al., 2012; Hinterleitner & Sager, 2019). In other words, cooperation can be risky, and this makes interagency cooperation and coordination more difficult (Busuioc, 2016; Moynihan, 2012). In addition, the results indicate that all related governmental agencies will try to shirk their responsibilities, sway each other, blur the truth, and avoid responsibility through politicized discourse or manipulation (Moynihan, 2012). In extreme cases, no agency will take responsibility and no one is willing to solve the problem. So, from the perspective of the public, this will make the public dissatisfied with the government and lose confidence in the government. Thus, BA strategies may have negative impacts on policy effectiveness and may hurt government operations (Hinterleitner & Sager, 2019). Therefore, BA must be temperate, and one or some agencies must take responsibility and save the overall image of the government. In this case study, the upper level of the government, the EY, played the role of a driver or leader that initiated and required the coordination and cooperation between agencies.

This study only examined the BA strategy used by three government agencies and did not conduct in-depth analysis of private manufacturers in Taiwan’s waste oil scandal. In fact, these related upstream and downstream companies had similar BA behaviors. Thus, future studies are suggested to study the BA strategies adopted by all agencies and private entities in a policy failure involved by more agencies, especially when at least one agency is unjustifiably blamed. In addition, the context of the event, the characteristics of the public and the media, and the responsibilities of the agencies involved are limited by the choice of the case. Therefore, different types of policy failure cases in different regions, under different institutions, and in different political cultures can be studied to test the usefulness of the classification and to deepen our understanding of the BA strategy adopted by multiple agencies. For these future studies, three different data sets are needed to clarify the degree of actual responsibility of the related agencies, the degree of perceived responsibility of related agencies, and the BA strategies adopted by these agencies during the crisis. In addition, the interaction between related agencies has to be investigated to reveal the whole picture of the BA strategy adopted by them. Because much of the data and information are qualitative, it is appropriate to test the propositions of this article through in-depth case studies.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This article is part of a project that has received funding from the Center for Public Policy and Law, National Taiwan University.

ORCID iDs

Hongwung Wang

Yi-Ching Tsai

Notes

Author Biographies

Hongwung Wang is an associate professor in the Department of Political Science and the Graduate Institute of Public Affairs at National Taiwan University. His fields of research include public policy and public management.

Yi-Ching Tsai receives his master degree from the Graduate Institute of Public Affairs at National Taiwan University. He is an Officer at the Marine Bureau, Kaohsiung City Government.

References

Ansell

Gash

(2008). Collaborative governance in theory and practice. Journal of Public Administration Research and Theory, 18(4), 543–571.

Boin

Hart

P. T.

McConnell

Preston

(2010). Leadership style, crisis response and blame management: The case of Hurricane Katrina. Public Administration, 88, 706–723.

Brändström

Kuipers

(2003). From “normal incidents” to political crises: Understanding the selective politicization of policy failures 1. Government and Opposition, 38, 279–305.

Bryson

J. M.

Crosby

B. C.

Stone

M. M.

(2015). Designing and implementing cross-sector collaborations: Needed and challenging. Public Administration Review, 75(5), 647–663.

Busuioc

E. M.

(2016). Friend or foe? Inter-agency cooperation, organizational reputation, and turf. Public Administration, 94(1), 40–56.

Busuioc

E. M.

Lodge

(2016). The reputational basis of public accountability. Governance, 29, 247–263.

Carpenter

D. P.

(2002). Groups, the media, agency waiting costs, and FDA drug approval. American Journal of Political Science, 490–505.

Carpenter

D. P.

Krause

G. A.

(2012). Reputation and public administration. Public Administration Review, 72, 26–32.

Cohen

(1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 37–46.

10.

Djerf-Pierre

Ekström

Johansson

(2013). Policy failure or moral scandal? Political accountability, journalism and new public management. Media, Culture & Society, 35, 960–976.

11.

Emerson

Nabatchi

Balogh

(2012). An integrative framework for collaborative governance. Journal of Public Administration Research and Theory, 22(1), 1–29.

12.

Gilad

(2012). Attention and reputation: Linking regulators’ internal and external worlds. In Lodge

Wegrich

(Eds.), Executive politics in times of crisis (pp. 157–175). Palgrave Macmillan.

13.

Gilad

Maor

Bloom

P. B. N.

(2013). Organizational reputation, the content of public allegations, and regulatory communication. Journal of Public Administration Research and Theory, 25, 451–478.

14.

Hinterleitner

(2017). Reconciling perspectives on blame avoidance behaviour. Political Studies Review, 15, 243–254.

15.

Hinterleitner

(2018). Policy failures, blame games and changes to policy practice. Journal of Public Policy, 38, 221–242.

16.

Hinterleitner

Sager

(2015). Avoiding blame—A comprehensive framework and the Australian home insulation program fiasco. Policy Studies Journal, 43(1), 139–161.

17.

Hinterleitner

Sager

(2017). Anticipatory and reactive forms of blame avoidance: Of foxes and lions. European Political Science Review, 9(4), 587–606.

18.

Hinterleitner

Sager

(2019). Blame, reputation, and organizational responses to a politicized climate. In The blind spots of public bureaucracy and the politics of non-coordination (pp. 133–150). Palgrave Macmillan.

19.

Hood

(2002). The risk game and the blame game. Government and Opposition, 37(1), 15–37.

20.

Hood

(2007). What happens when transparency meets blame-avoidance? Public Management Review, 9, 191–210.

21.

Hood

(2010). The blame game: Spin, bureaucracy, and self-preservation in government. Princeton University Press.

22.

Hood

Jennings

Copeland

(2016). Blame avoidance in comparative perspective: Reactivity, staged retreat and efficacy. Public Administration, 94, 542–562.

23.

Hood

Jennings

Dixon

Hogwood

Beeston

(2009). Testing times: Exploring staged responses and the impact of blame management strategies in two examination fiasco cases. European Journal of Political Research, 48, 695–722.

24.

Hood

Rothstein

(2001). Risk regulation under pressure: Problem solving or blame shifting? Administration & society, 33(1), 21–53.

25.

Liu

B. F.

Horsley

J. S.

Yang

(2012). Overcoming negative media coverage: Does government communication matter? Journal of Public Administration Research and Theory, 22, 597–621.

26.

Maor

(2010). Organizational reputation and jurisdictional claims: The case of the US Food and Drug Administration. Governance, 23(1), 133–159.

27.

Maor

Gilad

Bloom

P. B. N.

(2012). Organizational reputation, regulatory talk, and strategic silence. Journal of Public Administration Research and Theory, 23, 581–608.

28.

McConnell

(2010). Policy success, policy failure and grey areas in-between. Journal of Public Policy, 30, 345–362.

29.

McGraw

K. M.

(1991). Managing blame: An experimental test of the effects of political accounts. American Political Science Review, 85, 1133–1157.

30.

Mortensen

P. B.

(2013). Public sector reform and blame avoidance effects. Journal of Public Policy, 33, 229–253.

31.

Moynihan

D. P.

(2012). Extra-Network organizational reputation and blame avoidance in networks: The Hurricane Katrina example. Governance, 25, 567–588.

32.

Schillemans

(2016). Fighting or fumbling with the beast? The mediatisation of public sector agencies in Australia and The Netherlands. Policy & Politics, 44(1), 79–96.

33.

TFDA & MOHW. (2014). Black-hearted oil incident: Management measures and presentation on the progression of policy adoption, black-hearted oil section. http://www.fda.gov.tw/TC/siteContent.aspx?sid=4159

34.

Weaver

R. K.

(1986). The politics of blame avoidance. Journal of Public Policy, 6, 371–398.

35.

Weaver

R. K.

(2013). Policy leadership and the blame trap: Seven strategies for avoiding policy stalemate (pp. 1–9). Governance Studies, Brookings Institution.

36.

Yin

R. K.

(2009). Case study research: Design and methods (4th ed.). SAGE.