Asymmetric Information, Certification, and Labeling
Published on 1. December 2014
Book
X, 223 pages
978-3-86624-615-7 (ISBN)
Description
Since the path breaking contribution of Akerlof (1970), information asymmetry has been acknowledged as being a significant market failure. Informational asymmetries give rise to adverse selection. If buyers have less information on quality than sellers, the market underprovides high quality or, even worse, may completely break down. This "lemons" problem emerges in various markets as insurance markets, labor markets, financial markets, and product markets. The thesis at hand deals with adverse selection problems in two different settings. While the first part of this thesis considers imperfect information in a business-to-business relation (Chapters 2 and 3), the second part addresses asymmetric information between consumers and producers (Chapters 4 and 5).
In Chapters 2 and 3, a manufacturer buys an essential input from suppliers in order to sell his final product to consumers. Being responsible for potential product recalls, he is reliant upon the choice of a supplier whose production processes are of high quality. High process quality including precise documentation, traceability, and identity preservation allows the manufacturer to efficiently recall products and to reduce recall costs. A supplier's production process and its quality are, however, hidden to third parties.
In Chapters 4 and 5, consumers are willing to pay a premium for a product of higher quality such as an environmentally friendly, fairly traded, or organic product. This in line with empirical evidence (Lusk and Briggeman, 2009; Teisl et al., 2008; Wiser, 2007; Arnot et al., 2006), although environment protection or fair trade have public good characteristics. In the economic literature, this kind of altruism was introduced by Andreoni (1990). Consumers derive utility, the so called "warm glow" effect, by voluntarily providing a small amount of a public good. Yet, consumers have no means to verify producers' claims such as "green", "fair traded", or "organic".
The informational asymmetries in the two settings emerge because the quality characteristics described above are related to conditions of production, which are usually unobservable for buyers. Depending on how quality can be assessed, quality characteristics can be classified into different categories. While search attributes are known before purchase, experience attributes can only be determined after purchase (Nelson, 1970). With credence attributes, Darby and Karni (1973) added another quality dimension. Buyers cannot assess the quality even after purchase. They face two forms of asymmetric information. First, only an expert, mostly the seller, can determine a buyer's need. Secondly, only the seller knows the provided quality. The quality characteristics considered in this thesis are thus a simplified version of credence attributes (Roe and Sheldon, 2007). In the case of process attributes, buyers are assumed to know their preferences.
In his pioneering article, Akerlof (1970) states that "numerous institutions arise to counteract the effects of quality uncertainty" and that "this [quality uncertainty] may indeed explain many economic institutions [.]." During the past forty years, a vast part of microeconomic theory has thus focused on mechanisms which may be used to overcome informational asymmetries. One mechanism, recognized by Spence (1973), is signaling. Informed sellers signal their quality by investing in costly and observable actions. They can reveal information on unobservable quality by investing in reputation, spending money for advertising, or providing warranties (Shapiro, 1983; Milgrom and Roberts, 1986; Gal-Or, 1989). This thesis deals with another "economic institution" to cope with asymmetric information: certification by third party. According to the International Organization of Standardization (ISO), certification is the provision of a written assurance, the certificate, that the product, service, or system in question meets specific requirements by an independent third party.
In business-to-business relations as considered in the first part of this thesis, certifications according to management standards like ISO 9001 are prevalent. ISO 9001 provides a model how to set up and operate a management system and can be used by any organization in any activity. A management system is the framework of procedures to ensure that an organization can fulfill all tasks required in order to meet the organization's objectives as satisfying customers' quality requirements or complying with regulations.
The immense growth in the number of certifications according to ISO 9001 illustrates the popularity of this standard (see Figure 1.1). While in 1993, 46,571 organizations were certified worldwide, at the end of 2011, 1,111,698 certificates were issued. This increase in certification activity is observed, although certification is not a requirement of the standard and costs of a certification are significant (Anderson et al., 1999). Proponents of quality management systems according to ISO 9001 have pointed to their value as quality improvement tool. Quality improvements through greater quality awareness, increased efficiency, and enhanced productivity may translate in a higher operational performance. However, there is no clear relationship between the official act of certification and operational performance (Terlaak and King, 2006). Yet, why do so many firms put up with the additional difficulties and costs of certification in order to use ISO's official branding?
With the certification scheme, independent certification bodies issue ISO certificates with validity of three years based on quality inspections throughout the production process and annual surveillance audits. A firm's certification confirms the conform use of documented and standardized processes to ensure that products meet the buyer's requirements. One could conclude that an official certification may constitute an "economic institution", as mentioned by Akerlof (1970), to counteract informational asymmetries between a manufacturer and potential suppliers whose process quality is crucial but unobservable. Certification may satisfy the requirements for a credible signal of a supplier's hidden process quality.
Empirical research has indeed found some evidence that certification with a management standard may act as a signal of superior but unobservable organizational attributes such as process quality (Johnstone and Labonne, 2009; Terlaak and King, 2006). To the best of our knowledge, certification as a signaling device has never been examined theoretically in a business-to-business context. Therefore, the first part of this thesis aims to analyze a manufacturer's choice of a qualified supplier under asymmetric information. How can a supplier of high process quality signal his superior process quality to the manufacturer? And, under which conditions does certification act as a signal of unobservable process quality?
In Chapter 2, we answer these questions by using a signaling model with a manufacturer and two suppliers, who engage in price competition. The manufacturer's costs of potential product recalls are affected by suppliers' unobservable process quality. He knows that one supplier is of high process quality, whereas the other is of low process quality. Yet, he is unaware of who offers high process quality and thus induces lower recall costs. Chapter 2 presents two different scenarios. Either suppliers' production costs are quality-independent or quality-dependent. We first consider the pure price signaling model and show that price alone cannot signal superior process quality in a business-to-business relation. Rather, the supplier of high process quality needs an additional signal in a non-price dimension. Certification by third party acts as a signal if production costs are independent of suppliers' process qualities. If production costs are quality-dependent, we are unable to predict the full impact of certification on signaling activities. No (pure strategy) equilibrium exists. In Chapter 2, we neglect the possibility that suppliers have identical process qualities. This assumption of perfectly negatively correlated process qualities is rather restrictive and prevents certification to be a signal of superior process quality in the case of quality-dependent production costs.
Therefore, Chapter 3 captures an extended setting, in which process qualities are uncorrelated. Each supplier may have either low or high process quality. We restrict attention to quality-dependent production costs. Consistent with Chapter 2, pure price signaling offers no or only little information. At best, only the information that two suppliers of low process quality are in the market is revealed. With certification as an additional signal, fully separating equilibria exist if the difference in expected recall costs due to different process quality is small and marginal costs of certification are not too high for a supplier of high process quality.
Chapters 2 and 3 exclude the possibility that suppliers may enhance process quality in response to certification by third party. Neglecting the suppliers' reactions to certification allows us to focus on the certification's signaling function. Nevertheless, third-party certification may exhibit a quality improvement function. In product markets with asymmetric information between consumers and producers, as considered in the second part of this thesis, firms have no means to capitalize on consumers' willingness to pay for high product quality. As long as consumers cannot verify firms' claims, the latter have no incentive to offer fairly traded or organic products, or to use environmentally friendly production techniques. Third-party certification may constitute a remedy to mitigate these adverse selection problems and to improve product quality in these markets.
In product markets, third-party certification is referred to as "labeling" or "eco-labeling". With labeling, an independent third party verifies that a firm's product or production technique fulfills certain criteria or standards. Products or production techniques which are less harmful to environment, society, or health are awarded to use a label. Providing information about a product's quality, labels reduce the information gap between consumers and producers. Given an increased willingness to pay for ecological or ethical correct products, a market for superior quality emerges and firms have incentives to produce these products.
Similar to certifications according to management standards, the number of labels has grown substantially. A global directory of eco-labels, "Ecolabel Index", is currently tracking 437 eco-labels in 197 countries and 25 industry sectors. Well known eco-labels are the German "Blue Angel", the Nordic Council's "Nordic Swan", or the "EU Ecolabel". In 2000, the "EU Ecolabel" awarded 53 licences, whereas today more than 17,000 products carry this label. One could also observe a worldwide emergence of social labels such as "FairTrade".
The continuous proliferation of labels suggests that they are another manifestation of an "economic institution" to mitigate the consequences of quality uncertainty, as Akerlof (1970) predicted. While the first part of this thesis considers certification as a firm device to signal superior quality, the second part adopts a policy perspective. Chapters 4 and 5 deal with third-party certification, i.e. labeling, as a policy instrument.
A main characteristic of eco-labels is their voluntariness of adoption. Eco-labeling is therefore a typical instrument of a new environmental policy, which was promoted in the last years and emphasizes a higher degree of market-orientation (Rubik et al., 2008). Eco-labeling is seen as a voluntary "market based instrument [.]. Its main purpose [among others] is [.] to promote the adoption of more environmentally sound production methods and technologies." (Lehtonen, 1997). Thus, eco-labeling seems to have the potential to regulate environmental externalities and to reduce emissions. Moreover, since many environmental problems are transnational or even global problems, the voluntary nature of this instrument is a decisive advantage.
Chapter 4 thus analyzes whether voluntary eco-labeling is a suitable instrument to enhance social welfare by reducing emissions. However, given the voluntary nature of eco-labels, the effectiveness in terms of reducing emissions crucially depends on firms' willingness to participate. Setting up a voluntary eco-labeling standard, policy makers must be aware of this aspect. Furthermore, they must take into account an eco-label's effect on a market's competitive structure. A labeling scheme allows firms to differentiate their products with respect to environmental quality, which reduces competition and increases firms' market power (Tirole, 1994). Accordingly, introducing an eco-label may have countervailing effects on social welfare. Therefore, Chapter 4 addresses the following questions. How must policy makers set up a welfare maximizing labeling standard, taking into account the voluntariness of eco-labeling as well as competition between firms? What are the labeling's effects on emissions and social welfare? And, how is the labeling's impact affected if firms dispose of different abatement technologies?
We answer these questions by using a duopoly model of vertical product differentiation with firms disposing of different abatement technologies. Since labeling is seen as another effective means of product differentiation for firms, this is a common approach in the literature and also employed in Chapter 5. We extend the model by introducing asymmetric information about environmental quality and a pollution externality as well as including an initial standard setting stage. A welfare maximizing regulator sets up a labeling standard, taking into account firms' labeling and price decisions.
An optimally designed voluntary eco-label always reduces emissions and enhances social welfare. However, while the impact of introducing an optimal labeling standard on producer surplus is unambiguously positive, the effect on consumer surplus is only positive if the social valuation of damages associated to emissions is high enough. Our analysis further reveals that an optimally designed eco-label may lead to a cost-inefficient and inefficiently low emission reduction if firms dispose of different abatement technologies.
Chapter 4 focuses on eco-labels which are based on a verification by an independent third party. Labels, however, are not always the result of third-party certification schemes. Some labels are developed by firms on their own to advertise some characteristics of their products. Labels and the underlying certification processes are issued and controlled by several organizations as government, non-governmental organizations (NGO), industry groups, and for-profit certifying firms. The multiplicity of labels and issuing organizations has been criticized for causing confusion and distrust on the consumers' side (Dendler, 2013; Bratt et al., 2011). Uncertainty over the source of a label is a key factor for consumers' confusion (Harbaugh et al., 2011). How can consumers asses the reliability of labels which differ in their source, the underlying labeling standards, and the labeling schemes' strictness? Self-declared labels are not verifiable, and as a consequence, a moral hazard problem arises. Firms can declare a certain level of quality through their labels while in fact they provide lower quality. Non-confused consumers recognize these labels as not reliable and link them to lower quality. However, if consumers are confused about labels' reliability, this may perturbs the intended welfare effects of an existing third-party label.
Chapter 5 thus answers the following questions. What are the welfare effects of a labeling scheme which intends to mitigate information asymmetry concerning product quality? And, how does consumers' confusion about labels' reliability influence its impact?
Though the focus is different, Chapter 5 employs the same model as Chapter 4. We consider a three-stage game of standard setting, labeling decisions, and price competition in a duopolistic set-up. For reasons of traceability, we refrain from modeling an externality and consider symmetric firms. In order to assess the impact of consumers' confusion about labels, one firm has the possibility to create its own label. In contrast to the regulator's label, which is based on a perfect certification process, the firm's own label is not monitored and thus not reliable. On the consumers' side, we add uncertainty about labels' reliability.
Consistent with Chapter 4, voluntary labeling by third party is an effective policy instrument to eliminate asymmetric information, to improve quality, and to enhance welfare. However, allowing firms to differentiate their products with respect to product quality, labeling alleviates price competition between firms. While firms are better off, consumers are worse off. Labeling is thus unable to attain the social optimum. The existence of an additional label leading to consumers' confusion perturbs the third-party labeling's impact if the consumers' trust in the third-party label is too low. In particular, social welfare is as low as without a labeling scheme. The problem of firms exerting market power is aggravated, and consumers are penalized even more.
To close this chapter, this thesis with its four self-contained chapters contributes to our understanding of third-party certification as another "economic institution" to cope with asymmetric information, as Akerlof (1970) called it. The first part shows under which conditions certification as a firm device signals unobservable quality of production processes. The second part provides evidence that certification or labeling is a welfare enhancing policy instrument but may generate other market imperfections alleviating its positive impact.
In Chapters 2 and 3, a manufacturer buys an essential input from suppliers in order to sell his final product to consumers. Being responsible for potential product recalls, he is reliant upon the choice of a supplier whose production processes are of high quality. High process quality including precise documentation, traceability, and identity preservation allows the manufacturer to efficiently recall products and to reduce recall costs. A supplier's production process and its quality are, however, hidden to third parties.
In Chapters 4 and 5, consumers are willing to pay a premium for a product of higher quality such as an environmentally friendly, fairly traded, or organic product. This in line with empirical evidence (Lusk and Briggeman, 2009; Teisl et al., 2008; Wiser, 2007; Arnot et al., 2006), although environment protection or fair trade have public good characteristics. In the economic literature, this kind of altruism was introduced by Andreoni (1990). Consumers derive utility, the so called "warm glow" effect, by voluntarily providing a small amount of a public good. Yet, consumers have no means to verify producers' claims such as "green", "fair traded", or "organic".
The informational asymmetries in the two settings emerge because the quality characteristics described above are related to conditions of production, which are usually unobservable for buyers. Depending on how quality can be assessed, quality characteristics can be classified into different categories. While search attributes are known before purchase, experience attributes can only be determined after purchase (Nelson, 1970). With credence attributes, Darby and Karni (1973) added another quality dimension. Buyers cannot assess the quality even after purchase. They face two forms of asymmetric information. First, only an expert, mostly the seller, can determine a buyer's need. Secondly, only the seller knows the provided quality. The quality characteristics considered in this thesis are thus a simplified version of credence attributes (Roe and Sheldon, 2007). In the case of process attributes, buyers are assumed to know their preferences.
In his pioneering article, Akerlof (1970) states that "numerous institutions arise to counteract the effects of quality uncertainty" and that "this [quality uncertainty] may indeed explain many economic institutions [.]." During the past forty years, a vast part of microeconomic theory has thus focused on mechanisms which may be used to overcome informational asymmetries. One mechanism, recognized by Spence (1973), is signaling. Informed sellers signal their quality by investing in costly and observable actions. They can reveal information on unobservable quality by investing in reputation, spending money for advertising, or providing warranties (Shapiro, 1983; Milgrom and Roberts, 1986; Gal-Or, 1989). This thesis deals with another "economic institution" to cope with asymmetric information: certification by third party. According to the International Organization of Standardization (ISO), certification is the provision of a written assurance, the certificate, that the product, service, or system in question meets specific requirements by an independent third party.
In business-to-business relations as considered in the first part of this thesis, certifications according to management standards like ISO 9001 are prevalent. ISO 9001 provides a model how to set up and operate a management system and can be used by any organization in any activity. A management system is the framework of procedures to ensure that an organization can fulfill all tasks required in order to meet the organization's objectives as satisfying customers' quality requirements or complying with regulations.
The immense growth in the number of certifications according to ISO 9001 illustrates the popularity of this standard (see Figure 1.1). While in 1993, 46,571 organizations were certified worldwide, at the end of 2011, 1,111,698 certificates were issued. This increase in certification activity is observed, although certification is not a requirement of the standard and costs of a certification are significant (Anderson et al., 1999). Proponents of quality management systems according to ISO 9001 have pointed to their value as quality improvement tool. Quality improvements through greater quality awareness, increased efficiency, and enhanced productivity may translate in a higher operational performance. However, there is no clear relationship between the official act of certification and operational performance (Terlaak and King, 2006). Yet, why do so many firms put up with the additional difficulties and costs of certification in order to use ISO's official branding?
With the certification scheme, independent certification bodies issue ISO certificates with validity of three years based on quality inspections throughout the production process and annual surveillance audits. A firm's certification confirms the conform use of documented and standardized processes to ensure that products meet the buyer's requirements. One could conclude that an official certification may constitute an "economic institution", as mentioned by Akerlof (1970), to counteract informational asymmetries between a manufacturer and potential suppliers whose process quality is crucial but unobservable. Certification may satisfy the requirements for a credible signal of a supplier's hidden process quality.
Empirical research has indeed found some evidence that certification with a management standard may act as a signal of superior but unobservable organizational attributes such as process quality (Johnstone and Labonne, 2009; Terlaak and King, 2006). To the best of our knowledge, certification as a signaling device has never been examined theoretically in a business-to-business context. Therefore, the first part of this thesis aims to analyze a manufacturer's choice of a qualified supplier under asymmetric information. How can a supplier of high process quality signal his superior process quality to the manufacturer? And, under which conditions does certification act as a signal of unobservable process quality?
In Chapter 2, we answer these questions by using a signaling model with a manufacturer and two suppliers, who engage in price competition. The manufacturer's costs of potential product recalls are affected by suppliers' unobservable process quality. He knows that one supplier is of high process quality, whereas the other is of low process quality. Yet, he is unaware of who offers high process quality and thus induces lower recall costs. Chapter 2 presents two different scenarios. Either suppliers' production costs are quality-independent or quality-dependent. We first consider the pure price signaling model and show that price alone cannot signal superior process quality in a business-to-business relation. Rather, the supplier of high process quality needs an additional signal in a non-price dimension. Certification by third party acts as a signal if production costs are independent of suppliers' process qualities. If production costs are quality-dependent, we are unable to predict the full impact of certification on signaling activities. No (pure strategy) equilibrium exists. In Chapter 2, we neglect the possibility that suppliers have identical process qualities. This assumption of perfectly negatively correlated process qualities is rather restrictive and prevents certification to be a signal of superior process quality in the case of quality-dependent production costs.
Therefore, Chapter 3 captures an extended setting, in which process qualities are uncorrelated. Each supplier may have either low or high process quality. We restrict attention to quality-dependent production costs. Consistent with Chapter 2, pure price signaling offers no or only little information. At best, only the information that two suppliers of low process quality are in the market is revealed. With certification as an additional signal, fully separating equilibria exist if the difference in expected recall costs due to different process quality is small and marginal costs of certification are not too high for a supplier of high process quality.
Chapters 2 and 3 exclude the possibility that suppliers may enhance process quality in response to certification by third party. Neglecting the suppliers' reactions to certification allows us to focus on the certification's signaling function. Nevertheless, third-party certification may exhibit a quality improvement function. In product markets with asymmetric information between consumers and producers, as considered in the second part of this thesis, firms have no means to capitalize on consumers' willingness to pay for high product quality. As long as consumers cannot verify firms' claims, the latter have no incentive to offer fairly traded or organic products, or to use environmentally friendly production techniques. Third-party certification may constitute a remedy to mitigate these adverse selection problems and to improve product quality in these markets.
In product markets, third-party certification is referred to as "labeling" or "eco-labeling". With labeling, an independent third party verifies that a firm's product or production technique fulfills certain criteria or standards. Products or production techniques which are less harmful to environment, society, or health are awarded to use a label. Providing information about a product's quality, labels reduce the information gap between consumers and producers. Given an increased willingness to pay for ecological or ethical correct products, a market for superior quality emerges and firms have incentives to produce these products.
Similar to certifications according to management standards, the number of labels has grown substantially. A global directory of eco-labels, "Ecolabel Index", is currently tracking 437 eco-labels in 197 countries and 25 industry sectors. Well known eco-labels are the German "Blue Angel", the Nordic Council's "Nordic Swan", or the "EU Ecolabel". In 2000, the "EU Ecolabel" awarded 53 licences, whereas today more than 17,000 products carry this label. One could also observe a worldwide emergence of social labels such as "FairTrade".
The continuous proliferation of labels suggests that they are another manifestation of an "economic institution" to mitigate the consequences of quality uncertainty, as Akerlof (1970) predicted. While the first part of this thesis considers certification as a firm device to signal superior quality, the second part adopts a policy perspective. Chapters 4 and 5 deal with third-party certification, i.e. labeling, as a policy instrument.
A main characteristic of eco-labels is their voluntariness of adoption. Eco-labeling is therefore a typical instrument of a new environmental policy, which was promoted in the last years and emphasizes a higher degree of market-orientation (Rubik et al., 2008). Eco-labeling is seen as a voluntary "market based instrument [.]. Its main purpose [among others] is [.] to promote the adoption of more environmentally sound production methods and technologies." (Lehtonen, 1997). Thus, eco-labeling seems to have the potential to regulate environmental externalities and to reduce emissions. Moreover, since many environmental problems are transnational or even global problems, the voluntary nature of this instrument is a decisive advantage.
Chapter 4 thus analyzes whether voluntary eco-labeling is a suitable instrument to enhance social welfare by reducing emissions. However, given the voluntary nature of eco-labels, the effectiveness in terms of reducing emissions crucially depends on firms' willingness to participate. Setting up a voluntary eco-labeling standard, policy makers must be aware of this aspect. Furthermore, they must take into account an eco-label's effect on a market's competitive structure. A labeling scheme allows firms to differentiate their products with respect to environmental quality, which reduces competition and increases firms' market power (Tirole, 1994). Accordingly, introducing an eco-label may have countervailing effects on social welfare. Therefore, Chapter 4 addresses the following questions. How must policy makers set up a welfare maximizing labeling standard, taking into account the voluntariness of eco-labeling as well as competition between firms? What are the labeling's effects on emissions and social welfare? And, how is the labeling's impact affected if firms dispose of different abatement technologies?
We answer these questions by using a duopoly model of vertical product differentiation with firms disposing of different abatement technologies. Since labeling is seen as another effective means of product differentiation for firms, this is a common approach in the literature and also employed in Chapter 5. We extend the model by introducing asymmetric information about environmental quality and a pollution externality as well as including an initial standard setting stage. A welfare maximizing regulator sets up a labeling standard, taking into account firms' labeling and price decisions.
An optimally designed voluntary eco-label always reduces emissions and enhances social welfare. However, while the impact of introducing an optimal labeling standard on producer surplus is unambiguously positive, the effect on consumer surplus is only positive if the social valuation of damages associated to emissions is high enough. Our analysis further reveals that an optimally designed eco-label may lead to a cost-inefficient and inefficiently low emission reduction if firms dispose of different abatement technologies.
Chapter 4 focuses on eco-labels which are based on a verification by an independent third party. Labels, however, are not always the result of third-party certification schemes. Some labels are developed by firms on their own to advertise some characteristics of their products. Labels and the underlying certification processes are issued and controlled by several organizations as government, non-governmental organizations (NGO), industry groups, and for-profit certifying firms. The multiplicity of labels and issuing organizations has been criticized for causing confusion and distrust on the consumers' side (Dendler, 2013; Bratt et al., 2011). Uncertainty over the source of a label is a key factor for consumers' confusion (Harbaugh et al., 2011). How can consumers asses the reliability of labels which differ in their source, the underlying labeling standards, and the labeling schemes' strictness? Self-declared labels are not verifiable, and as a consequence, a moral hazard problem arises. Firms can declare a certain level of quality through their labels while in fact they provide lower quality. Non-confused consumers recognize these labels as not reliable and link them to lower quality. However, if consumers are confused about labels' reliability, this may perturbs the intended welfare effects of an existing third-party label.
Chapter 5 thus answers the following questions. What are the welfare effects of a labeling scheme which intends to mitigate information asymmetry concerning product quality? And, how does consumers' confusion about labels' reliability influence its impact?
Though the focus is different, Chapter 5 employs the same model as Chapter 4. We consider a three-stage game of standard setting, labeling decisions, and price competition in a duopolistic set-up. For reasons of traceability, we refrain from modeling an externality and consider symmetric firms. In order to assess the impact of consumers' confusion about labels, one firm has the possibility to create its own label. In contrast to the regulator's label, which is based on a perfect certification process, the firm's own label is not monitored and thus not reliable. On the consumers' side, we add uncertainty about labels' reliability.
Consistent with Chapter 4, voluntary labeling by third party is an effective policy instrument to eliminate asymmetric information, to improve quality, and to enhance welfare. However, allowing firms to differentiate their products with respect to product quality, labeling alleviates price competition between firms. While firms are better off, consumers are worse off. Labeling is thus unable to attain the social optimum. The existence of an additional label leading to consumers' confusion perturbs the third-party labeling's impact if the consumers' trust in the third-party label is too low. In particular, social welfare is as low as without a labeling scheme. The problem of firms exerting market power is aggravated, and consumers are penalized even more.
To close this chapter, this thesis with its four self-contained chapters contributes to our understanding of third-party certification as another "economic institution" to cope with asymmetric information, as Akerlof (1970) called it. The first part shows under which conditions certification as a firm device signals unobservable quality of production processes. The second part provides evidence that certification or labeling is a welfare enhancing policy instrument but may generate other market imperfections alleviating its positive impact.