The Cost of Quality Uncertainty

Table of Contents

List of Abbreviations

List of Tables

List of Figures

1 Introduction

2 Theoretical Foundation of and Hypotheses on the Effects of Quality Uncertainty
2.1 Roots and Pitfalls of Quality Uncertainty
2.1.1 Asymmetric Information
2.1.2 Adverse Selection
2.2 Counteracting Institutions
2.2.1 Governmental Institutions
2.2.2 Private Institutions
2.3 Behaviour of Market Actors
2.3.1 Assumptions
2.3.2 Sellers
2.3.3 Buyers
2.4 Size and Structure of Prices
2.4.1 Correlation between Quality and Price
2.4.2 Price Components

3 Empirical Study
3.1 Positioning of present study
3.2 Data Basis
3.2.1 Data Collection
3.2.2 Descriptive Statistics
3.3 Data Analysis
3.4 Results

4 Lessons from “Lemons” - Insights and Implications
4.1 The Cost of Quality Uncertainty
4.1.1 Relative and Absolute Size
4.1.2 Structure
4.2 Implications for Market Actors
4.2.1 Government
4.2.2 Sellers
4.2.3 Buyers
4.3 Transferability according to main study features
4.3.1 Legal Framework
4.3.2 Platform
4.3.3 Product
4.4 Limitations and Avenues for Further Research
4.4.1 Limitations
4.4.2 Avenues for Further Research

5 Summary

Appendix

References

Web Sites

List of Abbreviations

Abbildung in dieser Leseprobe nicht enthalten

List of Tables

Table 1: Categories to assess Product Quality

Table 2: Dimensions of Product Quality and their Assessment Criteria

Table 3: Sociological Assumptions on Human Behaviour

Table 4: Basic Variables to Assess Used Car Prices

Table 5: Means to Assess the Quality of Used Cars

Table 6: Basic features of examined cars

Table 7: Insignificance of platform exemplified by Opel Corsa

Table 8: Descriptive Statistics for the Opel Corsa

Table 9: Descriptive Statistics for the VW Golf IV

Table 10: Descriptive Statistics for the BMW 316i

Table 11: Descriptive Statistics for the Mercedes E220

Table 12: Empirical Results for testing on H1

Table 13: Empirical Results for testing on H2

Table 14: Insignificance of platform exemplified by VW Golf IV

Table 15: Insignificance of platform exemplified by BMW 316i

Table 16: Insignificance of platform exemplified by Mercedes E220

List of Figures

Figure 1: Relevance of Used Car Market in Germany in the year 2002

Figure 2: Segments of potential car buyers and their preferences

Figure 3: Q-Q-Plot for Opel Corsa grouped by sellers

Figure 4: Q-Q-Plot for VW Golf IV grouped by sellers

Figure 5: Q-Q-Plot for BMW 316i grouped by sellers

Figure 6: Q-Q-Plot for Mercedes E220 grouped by sellers

Figure 7: Boxplot for the Opel Corsa

Figure 8: Boxplot for the VW Golf IV

Figure 9: Boxplot for the BMW 316i

Figure 10: Boxplot for the Mercedes E220

Figure 11: Relation between basic value and certainty premium

Figure 12: Absolute Cost of Quality Uncertainty in €

Figure 13: Relative Cost of Quality Uncertainty in %

Figure 14: Uncertainty premium for Opel Corsa and VW Golf IV

Figure 15: Uncertainty premium for BMW 316i and Mercedes E220

1 Introduction

In all markets trade takes place only if bid and offer agree on prices. Both sides have to value their received output higher than their wasted input in terms of materialistic, financial or virtual resources. The agreed price is hereby dependent on the traded product’s characteristics which are generally assumed to be obvious in equal measure for both buyer and seller. If one market actor exceeds the other in terms of relevant information about the product, trade becomes subordinate to his willingness to cooperate. The individually assigned prices are different and reflect the imparity of price determining knowledge. On stock exchanges this divergence of buying and selling price for the same stock or currency transaction is called bid/offer spread. For instance, the price might be €650 per Porsche share. A Porsche shareholder assumes this price to be reasonable meanwhile a Porsche manager, possessing a larger amount of price relevant information on the firm, rates the same share for only €600. After purchase, the Porsche shareholder will discover that the stock’s fair value was €600 rather than €650. This spread, resulting from the inability to objectively assess the product’s quality, often ends in an inefficient risk or uncertainty premium being paid before the information asymmetry is equalised in the long run.

In Akerlof’s work,¹ the problem is exemplified with reference to quality in the used car market. He claims that sellers and buyers possess asymmetric information about the used car’s quality. The seller has an intimate knowledge of the quality of the car, while the buyer can only observe the average quality on the market for used cars. According to this observation, the buyer’s willingness to pay is restricted to the price of a used car of average quality. Sellers with an above-average quality expect to receive an above- average price for their car, but will not be able to sell and hence take their car off the market. This leads to a decrease in the average quality of traded used cars and thereupon to a further increase regarding the withdrawal of above-average quality cars. The effect results in only low quality cars, often referred to as “lemons”, being offered on a small market. This process of crowding out is called adverse selection. To overcome this type of market failure, various governmental regulations and private signals assuring the quality of a used car have been forthcoming. Whether these are sufficiently effective to fully counter quality uncertainty has yet to be answered.

This paper aims to generate insights into the remaining costs of quality uncertainty on the market for “lemons”. The goal is to reveal the factors explaining the size and structure of the price and to deduce implications and explanations based on these results. An attempt is made to quantify and analyse the costs of quality uncertainty in the market for used cars, and an investigation into which fractions of the cost of quality uncertainty can be traced through additional services is undertaken. The size and structure of the non-traceable part is of central interest for this research. The purpose is to extract valuable insights into the market for “lemons” based on the statistically confirmed findings. The contribution of quality signals to reduce quality uncertainty is thereby documented. To conclude, this study will also attempt to provide explanations as well as an assessment regarding the transferability of the findings to other economic fields.

To achieve these research goals, a profound analysis of the existent quality uncertainty will be undertaken. The present work on the cost of quality uncertainty on the market for “lemons” splits into three central chapters, providing the reader with theoretical knowledge in chapter 2, empirical insights in chapter 3 and related implications in chapter 4.

2 Theoretical Foundation of and Hypotheses on the Effects of Quality Uncertainty

This chapter yields theoretical groundwork in order to explain the derivation of the main hypotheses. Sections 2.1 and 2.2 set the theoretical framework in which the hypotheses are then elaborated in section 2.3 and 2.4 respectively.

2.1 Roots and Pitfalls of Quality Uncertainty

This section displays how quality uncertainty, being anchored in asymmetric information, can result in adverse selection.

2.1.1 Asymmetric Information

The roots of quality uncertainty lie in the existence of asymmetric information. In economics, information asymmetry occurs when one party of a transaction has more or better information than the other.² Typically it is the seller who knows more about the product than the buyer, however, it is possible for the reverse to be true.³ This definition is based on a crucial assumption in the field of new institutional economics: bounded rationality. Depending on the scope of explanation and counteraction of asymmetric information, New Institutional Economics distinguishes between Property-Rights- Theory, Transaction-Cost-Theory and Principal-Agent-Theory.⁴ While bounded rationality is assumed in all three approaches, Principal-Agent-Theory explicitly focuses on its implications.⁵ Applied to human behaviour, the theory postulates the need to align the agent’s interests through incentives and monitoring.⁶ If market actors already have sufficient incentives to play fair, asymmetric information on a product can be neglected. According to previous assumptions, market actors possessed all the relevant information about their environment, trading partners and products - necessary for their decisions. However, this form of efficient trading is generally not found to be possible.⁷

Asymmetric information can be assessed in two ways:⁸ either according to its point in time, taking into account whether the information asymmetry exists and functions before or after contracting; or according to its subject, which can be a good, an agent’s intention or his action. It is generally differentiated between the following three types of asymmetric information. The agent’s hidden intention to exploit the principal’s dependency after contracting is known as hold-up and exists ex ante and interim. The principal can be harmed by the agent if they engage in idiosyncratic⁹ investment. The case where the principal cannot observe the agent’s action occurring ex post is called hidden action. The problem consists in the agent’s ability to commit a moral hazard.¹⁰ The central division prevails between incomplete and imperfect information.¹¹

This study will focus on imperfect information and related consequences of hidden characteristics. Imperfect information refers to an ex ante state of knowledge about a good or service, describing the unequal distribution between seller and buyer. The principal discovers the constant and crucial properties related to the agent’s good, service or character when contracting has already taken place, and bears the costs of the agent’s private information after contracting. Beforehand it is impossible to evaluate the quality of products and services or an acceptable price for exchanging them. As the information influences both traders’ valuations, the first-best solution is no longer feasible.¹² This paper focuses on the asymmetric information associated with the hidden characteristics of a used car,¹³ whereby the behaviour of market actors is assumed to be self-interested rather than altruistic. In such a bilateral trading situation with extreme ownership settings, where only the seller has private information, second-best mechanisms are likely to occur and to end in adverse selection.

2.1.2 Adverse Selection

The process due to asymmetric information in general and quality uncertainty in particular is referred to as adverse selection. A key role is assigned to the point in time at which the quality becomes apparent. Table 1 displays the terminological distinction between various product categories in terms of quality valuation.¹⁴

Table 1: Categories to assess Product Quality

Abbildung in dieser Leseprobe nicht enthalten

Source: Own description according to Nelson (1970) and Darby / Karni (1973).

Per se, the product “used-car” contains experience-quality. This prevailing market uncertainty, unlike event uncertainty, is rooted in the endogenous variables (e.g. the product quality) of the economic system.¹⁵ With the growing complexity of products, differences between the quality of these products also increase. To define a common ground, the quality can be assessed in three basic dimensions¹⁶ - displayed in Table 2.

Table 2: Dimensions of Product Quality and their Assessment Criteria

Abbildung in dieser Leseprobe nicht enthalten

Source: Own description according to Abbott (1955), p.145.

By comparing equal cases, we prevent the vertical and the innovational dimension from containing quality uncertainty. It is one of many criteria and therefore only potentially relevant for the buyer. Assuming that his economic actions - regarding an experience- quality product - are non-trivial, quality uncertainty will at least unconsciously be integrated into the buyer’s decision-making process. A familiar paradigm of a market with adverse selection is implied by a set of owners who wish to sell their used cars to a set of potential buyers. While sellers differ in the quality of their owned cars, buyers differ in the value they attach to cars of the same quality. Each seller knows the quality of his own car; each buyer, however, can observe only the average quality of the cars sold at each price. The important phenomenon of signalling in this context is pervasive. In its absence the only variable that agents may use to distinguish quality is the price, which has substantial effects on the allocation of goods.¹⁷ The conventional nature of the market equilibrium presumes that if trade takes place at all, it must take place at a single price. A distribution of prices is only possible if there is some other observable characteristic correlated with quality that can serve as a signal. Furthermore, the mechanism by which prices are formed does not typically play a role in the analysis. An equilibrium is simply defined as a single price which equates supply and demand.¹⁸ This single price equals the value of the average quality of the cars sold as buyers are not willing to pay more. Sellers with low quality cars can exploit the buyer’s correlation of expected average quality and willingness to pay. Sellers with above-average quality cars will not accept such exploitation and will remove their cars from the market. Thus, as only low quality cars are offered, the expected quality will systematically deteriorate and potentially end with the collapse of the entire market. As the agent (seller) has the incentive to deceive the principal (buyer) regarding the quality of his offered products, the resulting mechanism is known as adverse selection and an example of market failure.¹⁹

2.2 Counteracting Institutions

The following two sections convey approaches, undertaken in the governmental and the private sector, to annul the roots and associated pitfalls of quality uncertainty. The specific governmental and private institutions presented will be based on Germany’s institutional environment for the examined used car market. A similar combination of these signalling tools exists in nearly all countries.

2.2.1 Governmental Institutions

The government aims to create an institutional environment which fosters fair competition and efficient trade. Although there is agreement regarding the advantages of governmental action to resolve market failure,²⁰ its optimal extent is not always obvious. Its monopoly of power is used to impose threats of sanctioning in order to channel human behaviour and therefore reduce asymmetric information in general and quality uncertainty in particular. In the vehicle market, the governmental regulations consist of legal warranties, norms, standardisation and informational duty.

In accordance with §§ 474 ff. BGB, the legal warranty exists for one year after the purchase of a used good from either a private or a professional seller. This warranty is granted to every private buyer by law and cannot be restricted by professional sellers.²¹ In contrast, private sellers are allowed to exclude this warranty partially or completely from the contract. Effective since 01.01.2002 it imposes severe duties on the seller regarding defects which - noticed or not - were already in existence at the time the contracting took place. If the defect occurs within six months of purchase, the law automatically assumes that it existed at the time of contracting. Within this period the seller has either to prove that it did not exist beforehand or to repair the used car. If the seller is unable to repair a default covered by the legal warranty, the buyer has the right to lower the price afterwards or in some cases even to dissolve the contract. After the first six months, this so-called burden of proof²² shifts from the seller to the buyer for the remaining six months of warranty. The burden of proof is a crucial criterion when distinguishing between warranty and guarantee.

Norms and standardisations are a widespread tool to facilitate market interchanges. For new products, they are already inherent in the production process. They have to possess particular compulsory technical and safety characteristics²³ in order to be sold legally. Once purchased, the vehicle owner has to apply to a third independent party - in Germany called TÜV - for a certificate. This certificate assures that the car meets certain requirements regarding essential driving functions and emissions - namely HU and AU.²⁴ This test must be repeated every two years, with the exception of new cars for which it is not compulsory for the first three years.²⁵

The duty of information forces the seller to reveal all relevant information about the good²⁶ and thus increases the reliability of and trust in sellers. The better the seller and the buyer are informed, the more fair trade will occur. A mandatory disclosure law can hence be regarded as useful to prevent the practice of using higher prices as a device to signal higher quality.²⁷ Private sellers cannot exculpate themselves from this duty, but again the law as a basic principle assumes them to be unaware of existing defects.

Due to decreasing information asymmetry or information advantage, sellers’ private gains from trade also decrease.²⁸ Governmental institutions aim to narrow the gap between theoretical efficiency and practical reality, and to enlarge social welfare. These ends are also mirrored in private institutions.

2.2.2 Private Institutions

Taking the governmental institutions and remaining market imperfections as given, the market actors allocate their goods and resources in the most efficient way. The variety of private institutions employed is in parts very similar to governmental ones. The private institutions - namely guarantee, certificate, reputation or brand name, internet used as medium, and social pressure - are presented according to their relevance in reducing information asymmetry in general and quality uncertainty in particular.²⁹

A guarantee is in its implications very similar to a warranty.³⁰ It is based on a voluntary offer by professional sellers to signal quality through granting a certain standard. The seller is free to create the characteristics of the guarantee, such as the extent of the cover, its duration and examination intervals. The more a guarantee covers, the better it overcomes quality uncertainty.³¹ A potential side effect is the restriction on the buyer as examinations have to take place regularly and only the use of original parts is allowed for repair. Additionally, the cost of interpreting the terms of the guarantee might be prohibitively high for buyers. This is confirmed by evidence suggesting that buyers rarely read and compare alternative guarantees before reaching a purchase decision.³² Yet, from the buyer’s perspective, the guarantee itself can be seen as a competitive advantage compared to a warranty as the burden of proof³³ remains on the seller’s side. Non-compulsory technical certificates also entail a competitive advantage as they signal quality. Investments in certificates - similar to those in brand names or reputation, but contrary to those in guarantees - are ex ante devices to internalise the externalities from asymmetric information.³⁴ It is economically rational only for sellers of high quality goods to invest in a certificate for which certain requirements have to be fulfilled. The private certificates, just like the governmental ones, are available to both private and professional sellers from nationwide associations with a point-of-sale close by. Once a certificate is issued, its enforcement is immaterial. Examples are the Dekra’s³⁵ “Dekra seal” (for €68.50), and the TÜV Süd’s³⁶ “autocert” (for €63.80).³⁷

Advertisements or similar investments in signalling quality through a good reputation or brand name are rational in a dynamic model and therefore only recommended for professional dealers. The repeat-purchase mechanism contributes to resolving the adverse selection;³⁸ yet, no amount of advertising directly reveals the hidden product characteristics.³⁹ The positive effects on sales are only seen in the long term, as the premium for repeated trust is greater than that for one-time opportunism. It thus also serves the function of compensating sellers for their investment in reputation.⁴⁰ Buyers often use information surrogates or sources of non-market information.⁴¹ Many such surrogates are based on the idea of herding behaviour - prospective buyers join forces to use choices made by others as information to support their own.

Additionally, information asymmetry has recently been noted to be on the decline thanks to the internet used as a medium of information.⁴² The key characteristics of electronic markets - namely ubiquity, transparency, openness and reduced transaction costs - play a major role in the buyer’s procurement of information.⁴³ The internet generally reduces search costs as online agents scan all available information concerning the prices and quality attributes of internet offers.⁴⁴ It allows uninformed users to acquire heretofore unavailable information such as comparing the costs of used cars, insurance policies, etc.⁴⁵ Buyers may thus come close to the seller’s status of general knowledge about a car type, but remain uncertain about the specific car offered.

The theoretical institution of social pressure owes its lack of relevance to the absence of a credible threat to sanction deflection. A private seller may not be bound by social pressure but should at least be guided by his own moral concept or ethics. Financially, such a choice is as unlikely as it is irrational. Psychologically, it comes at high personal cost since it involves acting against one’s own moral judgement.⁴⁶ In the case of professional sellers social pressure will have some bearing where reputation is indirectly based on society’s opinion.

2.3 Behaviour of Market Actors

Having explained where quality uncertainty comes from (asymmetric information), where it leads to (market failure) and which means exist to counter it (governmental and private signalling institutions), we proceed to analyse and to postulate assumptions on the behaviour of market actors. As the study on the behaviour of market actors is by necessity - at least partly - a study of human behaviour, psychological aspects will also be integrated in the following section. It links the relevant assumptions on human behaviour with the individual decision making setting of the previous two sections in order to conclude with formulating the first main hypothesis.

2.3.1 Assumptions

This section will clarify the general assumptions - valid for all market actors - before the individual rationales are presented separately in the following sections. Firstly, all market actors maximize their own utility⁴⁷ with respect to the expected-utility-rule and game theory. The latter is a branch of mathematics that analyzes interactions with formalized incentive structures, allowing the study of predicted and actual behaviour.⁴⁸ Secondly, buyers especially are assumed to be risk-averse. To improve the decision making process, various means exist to gather information and thus minimize risk.⁴⁹ According to an empirical investigation 76% of durable goods buyers inform themselves “sufficiently”.⁵⁰ Information gathering is thus oriented by a psychological mechanism which persuades the buyer without the need for maximizing the state of information. The buyer is therefore assumed to acquire improved information within natural limits and to adapt his action to the state of this limited information.⁵¹ In accordance with the categories to assess product quality in section 2.1.2, the used car is a product characterised by experience-quality.⁵² While the price can be observed prior to purchase, the quality - at least from the buyer’s point of view - can be observed only after purchase and experience.⁵³ An extensively informed buyer is able to tell which of two products is a better fit for his needs. All buyers coincide in their ordinal ranking, although differences are not measurable on a cardinal scale. This assumption is unfeasible for complex products as buyers’ preferences may vary over different attributes and time. The complexity of the product “used car” will therefore be suppressed through the comparison of homogeneous models within each segment. The assumption that online prices are equal to trading prices is linked to the fact that the offer is not posted on auction platforms. Prices are therefore fixed and will not be subject to bargaining. The expected sociology of the market actors’ interplay is presented in Table 3.

Table 3: Sociological Assumptions on Human Behaviour

Abbildung in dieser Leseprobe nicht enthalten

Source: Own description according to Browne (1973), p.42.

While cases A, B, C and D are universally valid, case E is applicable only to professional sellers. The assumptions provide an overview of the most probable interplay between action and reaction of human behaviour. Despite being formulated at a time in which communication between seller and buyer was personal, their validity can be transferred to contacts via phone, fax or e-mail. The seller offers a specific good knowing its exact quality and trying to signal it. The buyer’s reaction consists in scanning the relevant information, approving the signalled quality and basing his decision on his limited state of information. This succession provides the logic within the order of presenting the individual market actors and their decision making rationale.

2.3.2 Sellers

As indicated above, the distinction between professional and private sellers is crucial as it yields to impersonate the difference between certainty and uncertainty regarding quality. The price difference is expected to be equal to the quality certainty premium.

2.3.2.1 Professional

According to German Law, a professional seller is every individual of legal age or corporate entity whose legal transactions are executed in their commercial function or self-employed activity.⁵⁴ All professional sellers have to comply with the law in terms of granting legal warranties. Their differentiation according to warranties provided is thus neither possible nor necessary - and their distinction from private sellers is facilitated.⁵⁵ The regulation requires sellers to warrant the mechanical condition of each automobile sold. Since the regulation has been imposed only on professional and not on private sellers, its effectiveness is dependent on the magnitude of the induced shift of sales into the private market.⁵⁶ Professional sellers buy used cars from private sellers who evade any involvement with the details and work related to selling a car. Cars of inferior quality will either not be bought by professional sellers or will be purchased at comparatively low costs, having munificently discounted necessary reparations. Apart from mandatory warranties, professional sellers have a variety of means to capture a comparative advantage vis-à-vis competing professional and private sellers. Guarantees are a typical example - as professional sellers are familiar with the product, they can easily construct the guarantee’s coverage in their favour.⁵⁷ The buyer relies on this volunteer promise with the implicit assumption of having received a benefit free of charge. The professional seller can create a similar impression by including free accessories instead of lowering the price. Buyers assume to have proven their skills during the bargain, while professional sellers possess far more means to apply. Yet, professional sellers do not necessarily have to reach an agreement on the specific car. Alternative offers can be taken into account without additional costs. A buyer might come to inspect one car and then go home with another - be it cheaper or more expensive.⁵⁸ Professional sellers’ competitive advantage or distinctive asset can be linked to cost savings. Those internet sellers who are aiming at a large volume of sales, must take into account that buyers are prepared to travel long distances to secure a bargain.⁵⁹ Alternative strategies focus on one-to-one marketing, geographical proximity or similar intangible assets. Central among these is the reputation conveyed by the firm's name⁶⁰ - while reputation refers to an individual seller, with whom positive experiences have been made in the past, the brand indicates the good quality on the whole span of influence. The professional seller’s influence on the value chain of the product “used car” consist of his sincerity, seriousness and service. One such aspect of seriousness is thorough car cleaning and changing of oil and/or brake fluid if necessary. Another key advantage of the professional seller is his mere presence and the buyer’s opportunity to call in person.

2.3.2.2 Private

Contrasting professionals, private sellers primarily want to exchange their used car without necessarily achieving economic gains. This does not imply an indifference regarding losses as they will try to get the best price they can. Based on a lack of technical knowledge, private sellers are not required by law - in contrast to previous assumptions - to be aware of the technical defects of their car. Per se, the legal warranty originally applied to both professional and private sellers, but no granting is required by private ones. It is therefore difficult to enforce one’s claim legally based on either the legal warranty or the duty of information. Private sellers generally evade getting involved in selling their car - either completely by selling to a professional or partly by minimising their commitment before and after contracting. They ask a price that they have observed for a similar car on the market or which satisfies them. Logically, they try to get as much out of their used car as possible and do not care at whose expense they do so. Consequently, they generally restrict the legal warranty completely. In practice, this is easily done by a paragraph in the selling contract entailing phrases such as “bought as seen and test-driven” or “with exclusion of liability”, which professional sellers are not allowed to employ. Just like warranties, also guarantees are accordingly unheard of on the private seller’s side.

The incentives for using private institutions to signal quality are very limited. Private sellers do not want to invest in cars they are about to sell. The benefit, for which they would bear the costs is invisible or might be seen as being someone else’s. This rationale loses its motive if private sellers perceive that a related, higher willingness to pay will overcompensate for this investment. The impact of such a strategic investment is unknown to them and the costs of finding out - if possible in the first place - are substantially higher than the expected return. Similarly, private sellers are not willing to invest in their own reputation as the transaction will not be repeated in the near future and buyers are unlikely to remember their name. In accordance with this bounded rationality, private sellers will hence not invest in cost-intensive quality signals.

2.3.3 Buyers

According to German Law, a buyer is every individual whose legal transactions are executed for ends that are neither a commercial function nor a self-employed activity.⁶¹ This definition excludes professional buyers who aim to resell the car at a higher price than it was originally purchased for. According to the purpose of this paper, professional buyers are not presented separately. Different rules regarding the legal warranty prevent the central feature from signalling quality and professional buyers from receiving further attention.

Buyers are assumed to know exactly which car they want. They attempt to maximize the utility they expect of a used car. In their decision making process, they are either implicitly or explicitly aware of their restricted ability to gather information on the car’s quality. Even though their individually allocated value may vary, buyers can rank qualities if these are heterogeneous and observable.⁶² They adapt a decision making process that takes all relevant information into account, such as car type, age, motorisation, optical appearance, accessories, fuel consumption and the residual risk. As buyers’ preferences for complex products vary over different attributes and time, the analysis will exclude horizontal and innovational differences and focus explicitly on one key vertical difference: the remaining risk related to quality uncertainty. Buyers assess these risks through signals containing advanced information on the true quality of the vehicle. This crucial information is either at prohibitive cost or not at all accessible for buyers. Professional sellers are assumed to be - at least implicitly - aware of this phenomenon and its potential to generate unjustifiably higher resale prices. As the difference in quality uncertainty is best approached by the distinction between professional and private sellers, the first hypothesis is:

H1: On average, professional sellers offer used cars at significantly higher prices than private sellers.

H1 will be tested against the null hypothesis that there is no significant price difference between private and professional sellers.

2.4 Size and Structure of Prices

In order to formulate further hypotheses, the suspected price difference from the previous chapter will be taken as given and its fractions analysed. This section presents the price determining factors in order to quantify the individual basic value of a used car. The first part describes the assumed correlation between quality and price, while the second part assembles the price’s structure. It then resumes by formulating hypotheses on prevalence and price dependency of the cost of quality uncertainty.

2.4.1 Correlation between Quality and Price

In early days of trade it was assumed that products were sufficiently described when they were named (tea, wheat, coal etc.). Back then, economists were eager to explain exchange relations. Since the heterogeneity and complexity of products was increasing, the impact of quality on prices was growing immensely; meanwhile economists were still lacking a theory on quality.⁶³ In general, the correlation between quality and price is positive.⁶⁴ A related conclusion is that this also validates for the correlation between quality certainty and price. Quality certainty is negatively correlated with the expected measure of unanticipated costs.⁶⁵ Hence, in today’s markets it does not suffice any longer to simply name the product. The level of quality has to be ascertained as well.

[...]

---

¹ See Akerlof (1970).

² It has also been called asymmetrical information and markets with asymmetrical information.

³ This situation was first described by Arrow (1963) in a seminal article on health care.

⁴ See Williamson (2000), p.595.

⁵ See Macho-Stadler / Pérez-Castrillo (2001); Stiglitz (2000), p.1441ff.

⁶ See Jensen / Meckling (1976).

⁷ See Gresik (1991), p.41-63.

⁸ See Lupton (2005), p.399.

⁹ A highly specific investment cannot easily be reversed. Its value for second best use, its opportunity costs, is close to zero. Optimal pricing and quality differ hereby according to Pennings (2004), p.569.

¹⁰ In economics, moral hazard is the name given to immoral behaviour, and the related negative outcome. The person who caused the problem does not suffer the full or any consequences, or may benefit.

¹¹ For a precise definition and distinction see Rasmusen (1994), p.165ff.

¹² See Fieseler / Kittsteiner / Moldovanu (2003), p.223.

¹³ The effect on new goods markets, see Hendel / Lizzeri (1999), and the special class of markets where approximate efficiency is obtained despite asymmetric information, see Hendel / Lizzeri / Siniscalchi (2005), p.467, are neglected as the effects of quality uncertainty are largest in the second hand market.

¹⁴ See Nelson (1970), p.327; Darby / Karni (1973), p.67-88. Under the consideration of time and substitutability their boundaries are not clear cut.

¹⁵ See Hirshleifer / Riley (1992), p.2-3.

¹⁶ See Abbott (1955), p.145-152.

¹⁷ See Wilson (1979), p.314.

¹⁸ See Wilson (1980), p.108.

¹⁹ See Akerlof (1970).

²⁰ Often adduced advantages exist in cases of a natural monopoly or the internalisation of external effects.

²¹ In German Law called “Haftungsausschluss”, see § 444 BGB.

²² In German Civil Law this is called “Beweislastumkehr”, see § 476 BGB.

²³ Example: Third rear breaking light compulsory for new cars, see 76/756/EWG i.V.m. ECE-R48;

43.AusnahmeVO.

²⁴ Haupt-Untersuchung (HU) and Abgas-Untersuchung (AU), see § 29 StVZO and § 47a StVZO.

²⁵ See § 29 StVZO.

²⁶ See § 442 BGB.

²⁷ See Simó (2003), p.233.

²⁸ See Levin (2001), p.657.

²⁹ We neglect leasing as firstly, it is primarily a tool to overcome the burden of financing capability or to avoid taxes and hence is not a subject for this study; secondly, a bank is likely to offer credit at a lower cost due to its economies of scale; and thirdly, despite the fact that lessors can structure contracts to prevent adverse selection, it is not in their interest; see Hendel / Lizzeri (2002), p.135.

³⁰ See § 443 BGB.

³¹ See Rapold (1988), p.86.

³² See Crocker (1986), p.147.

³³ For a more detailed explanation see section 2.2.1 on page 6.

³⁴ See Ortmann (1991), p.57-58.

³⁵ See Dekra [ed.] (2006), Used cars and the DEKRA SEAL.

³⁶ See TÜV Süd [ed.] (2006a), Was kostet’s?; certificate costs €75.40 in the unlikely case that the compulsory HU has been conducted somewhere else.

³⁷ TÜV Nord “Vertrauenscheck” for €34.80 is not mentioned here, because it is not a signal for a certain level of quality, but only a written documentation of the vehicle’s current condition. See TÜV Nord [ed.] (2005), Preise in Euro.

³⁸ See Klein / Leffler (1981), p.615; Lott (1988), p.165.

³⁹ See Overgaard (1991), p.14-20.

⁴⁰ See Shapiro (1983), p.659.

⁴¹ See Choi / Kim (1996), p.53.

⁴² See Dudenhöffer / Koster (2003), p.1.

⁴³ See Langer (2003), p.166-167.

⁴⁴ See Fabel / Lehmann (2002), p.175-176.

⁴⁵ See Huston / Spencer (2002), p.50.

⁴⁶ See Akerlof (2003), p.36.

⁴⁷ By doing so, they do not regard the maximization of social welfare as a primary objective.

⁴⁸ See Von Neumann / Morgenstern (1947), p.15-31.

⁴⁹ See Wiggins / Lane (1983), p.892.

⁵⁰ See Kupsch (1978), p.122.

⁵¹ See Hirshleifer / Riley (1992), p.2.

⁵² See Abbott (1955), p.145-152.

⁵³ See Hey / McKenna (1981), p.54.

⁵⁴ See § 14 BGB and read in conjunction with § 1 HGB.

⁵⁵ See § 6 TDG.

⁵⁶ See Metzger (1983), p.129.

⁵⁷ See section 2.2.2 on page 8.

⁵⁸ See Royal (2004), p.131.

⁵⁹ See O’Neill / Jones (2002), p.108-110; Dudenhöffer / Koster (2003), p.1-2.

⁶⁰ See Tadelis (1997), p.549.

⁶¹ See § 13 BGB.

⁶² See Shapiro (1982), p.20.

⁶³ See Abbott (1955), p.22-23.

⁶⁴ See Caves / Greene (1996), p.29; Ordòñez (1998), p.258.

⁶⁵ Such costs occur after contracting and result from value losses through hidden defects.