# Abstract

We contrast the laissez-faire regime with the regime of strict producer liability and draw the implications for competition policy in a setting where oligopolistic firms cannot differentiate themselves from rivals but rather are bound by a common industry reputation for product safety. We show that, first, unlike in the traditional products liability model, firms’ incentives to invest in precaution depend on market structure. Second, depending on the magnitude of expected damages awarded by the courts, laissez-faire can welfare dominate strict producer liability. Third, the relationship between social welfare and industry size, and hence the role for competition policy, depends on the institutional regime governing the industry. Under some circumstances, restricting industry size is unambiguously welfare-enhancing.

# Acknowledgments

For helpful comments and discussion, we thank Yao-Yu Chih, Valentina Dimitrova-Grajzl, Brett Frischmann, Keisuke Hattori, Jeroen Hinloopen, Wolfgang Kerber, Elodie Rouviere, Tinni Sen, Raphael Soubeyran, Spencer Waller, participants at the annual meetings of the Midwest Economics Association, the International Society on New Institutional Economics, the European Association for Research in Industrial Economics, the European Association of Law and Economics, seminar participants at Washington and Lee University, and two anonymous referees.

## Appendix A

**Derivation of expressions [4] and [5]**: Differentiating eq. [3] with respect to *a _{i}* and

*q*for all

_{i}*i*∈{1,…

*n*} gives

for all *i*∈{1,…*n*}. Letting *q _{i}*>0 for all

*i*∈{1,…

*n*}, eq. [18] immediately implies eq. [4]. Using eq. [19] in turn implies eq. [5]. □

**Derivation of expressions [8]–[10]**: Taking *a _{j}* and

*q*for all firms

_{j}*j*≠

*i*as given, firm

*i*chooses

*a*and

_{i}*q*to maximize eq. [7]. The resulting first-order conditions (FOC) for an interior solution are

_{i}After imposing symmetry (*Q*=*nq*), eq. [20] simplifies to

which implies expression [8]. Eq. [21] then simplifies to

Eq. [23] can be solved for *q* to obtain expression [9] and, when multiplying the resulting expression by *n*, expression [10]. □

**Derivation of expressions [13]–[15]**: Taking *a _{j}* and

*q*for all firms

_{j}*j*≠

*i*as given, firm

*i*chooses

*a*and

_{i}*q*to maximize eq. [12]. The resulting FOC for an interior solution are

_{i}Imposing symmetry (*Q*=*nq*), eq. [24] simplifies to

which implies expression [13]. Eq. [25] then simplifies to

Eq. [23] can be solved for *q* to obtain expression [14] and, when multiplying the resulting expression by *n*, expression [15]. □

**Lemma A1**: *The following statements hold*:

*The expression la+C(a) is decreasing for a<a*.^{FB}, increasing for a>a^{FB}, and attains a minimum at a=a^{FB}*The expression –la–C(a) is increasing for a<a*.^{FB}, decreasing for a>a^{FB}, and attains a maximum at a=a^{FB}*Let k*._{1}+C′(a_{1})=0 and k_{2}+ C′(a_{2})=0. Then, a_{1}>a_{2}if and only if k_{1}<k_{2}

*C*(

*a*) is strictly convex in

*a*and, second,

*l*+

*C*'(

*a*)=0 at

*a*=

*a*

^{FB}; see eq. [4]. □

**Lemma A2**: *When firms choose a common accident probability a and output level q*, *total industry output equals Q*=*nq*, *and social welfare equals*

*Then*

*and*

Proof: Straightforward, thus omitted. □

**Proof of Proposition 1**: If *d*=*l*, it follows from eqs. [4] and [13] that *a ^{L}*=

*a*

^{FB}. The comparison of eqs. [4] and [8] reveals that

*a*

^{FB}<

*a*. This proves part 1 of the result.

^{N}When *d*<*l*, the following holds

Thus, from Lemma A1, part 3, *a*^{FB}<*a ^{L}*<

*a*. This proves part 2 of the result.

^{N}Finally, when *d*>*l*, the following holds:

Thus, from Lemma A1, part 3, *a ^{L}*<

*a*

^{FB}<

*a*. This proves part 3 of the result. □

^{N}**Proof of Proposition 2**: From eqs. [10] and [15], note that for *r*∈{*N,L*} we have

Therefore, from Lemma A1, part 2,

Part 1 of the result then follows immediately from eq. [34] and Proposition 1, parts 1 and 2.

To prove part 2 of the result, note that *a ^{L}*<

*a*

^{FB}<

*a*when

^{N}*d*>

*l*(see Proposition 1). Thus, from eq. [33],

*Q*<

^{N}*Q*if and only if

^{L}When *d*=*l*, *a ^{L}*=

*a*

^{FB}<

*a*(see Proposition 1) and, thus, by Lemma A1, part 2, expression [35] holds. Note that the right-hand side of expression [35] is a continuous function of

^{N}*d*, whereas the left-hand side of expression [35] does not depend on

*d*. Thus, for

*d*>

*l*and

*d*close to

*l*, the inequality in expression [35] holds as well and, therefore,

*Q*<

^{N}*Q*. When

^{L}*d*increases,

*a*decreases (see eq. [13]) and, hence,

^{L}*C*(

*a*) increases. Therefore, for

^{L}*d*large enough, by continuity, the inequality sign in expression [35] is eventually reversed if

*α*–

*la*–

^{N}*C*(

*a*)>

^{N}*α*–

*C*(0), that is, if

*C*(0)>

*la*+

^{N}*C*(

*a*), or, equivalently, if [

^{N}*C*(0)–

*C*(

*a*)]/

^{N}*a*>

^{N}*l*. □

**Lemma A3**: *Let W ^{r} and Q^{r}*,

*respectively, denote welfare and total industry output under regime r*∈{

*N,L*}.

*Then*,

Proof: Using eqs. [10] and [15] we can write

for *r*∈{*N,L*}. By eq. [28]

Eq. [37] can be rewritten as

Substituting eq. [39] into eq. [38] and collecting terms give eq. [36].

**Proof of Proposition 3**: Follows immediately from Lemma A3 and Proposition 2.

**Proof of Proposition 4**: From Lemma A3,

for *r*∈{*N,L*}. Then, the direct effect of *n* on *W ^{r}* (viewing

*W*as a continuous function of

^{r}*n*and abusing notation somewhat) is

and

From eq. [37], the direct effect of *n* on *Q ^{r}* (again abusing notation somewhat) equals

where the last equality follows from eq. [39]. The competition effect thus equals

which is positive for all *r*∈{*N,L*}. This proves part 1 of the result.

To sign the industry reputation effect, first note that eq. [42] is positive for all *r*∈{*N,L*}. Thus, the sign of the industry reputation effect is determined by the sign of

Consider, first, the first derivative in the above product. We have

Thus, by Lemma A1,

Consider, next, the second derivative in eq. [45]. Accident probability *a ^{r}* is defined by

*γ*(

_{r}*n*) +

*C*′(

*a*) = 0, where the function

^{r}*γ*(

_{r}*n*) varies across regimes

*r*∈{

*N,L*} as follows:

Applying the implicit function theorem,

and since *C*(·) is strictly convex, thus,

Therefore,

To prove part 2(a) of the result, note that from eq. [48], sign[–*γ′ _{N}*(

*n*)]=1. Because

*a*>

^{N}*a*

^{FB}(see Proposition 1), sign(

*a*

^{FB}–

*a*)=–1. Therefore, by eq. [51], the industry reputation effect under

^{N}*r*=

*N*is negative.

To prove part 2(b) of the result, observe that from eq. [48], sign[–*γ′ _{L}*(

*n*)]=sign(

*l*–

*d*). From Proposition 1, for

*d*=

*l*we have

*a*=

^{L}*a*

^{FB}; for

*d*<

*l*we have

*a*>

^{L}*a*

^{FB}; and for

*d*>

*l*we have

*a*<

^{L}*a*

^{FB}. Thus, by eq. [51], the industry reputation effect under

*r*=

*L*equals zero for

*d*=

*l*and is negative for

*d*≠

*l*.

Parts 3 and 4 of Proposition 4 follow immediately from expression [17] and parts 1 and 2, which we prove above.

**Proof of Proposition 5**: From eq. [36], ∂*W ^{r}*/∂

*n*=–

*β*(

*Q*)

^{r}^{2}

*n*

^{–2}and ∂

*W*/∂

^{r}*Q*=

^{r}*β*(

*n*+2)

*Q*

^{r}n^{–1}. From eq. [37], ∂

*Q*/∂

^{r}*n*=

*Q*[

^{r}*n*(

*n*+1)]

^{–1}. Then, the competition effect (see eq. [17]) equals

and the industry reputation effect equals

Thus, we can write the total effect (see eq. [17]), as

The right-hand side of eq. [54] can be expressed as *T*_{1}* ^{r}*· [

*T*

_{2}

*+*

^{r}*T*

_{3}

*], where*

^{r}with *T*_{1}* ^{r}*>0,

*T*

_{2}

*>0 for all*

^{r}*r*∈{

*N,L*}, and

*T*

_{3}

*<0 for*

^{r}*r*=

*N*and for

*r*=

*L*and

*d*≠

*l*(see Proposition 4).

Consider, first, eq. [55]. Using eq. [37], we can write

Let *n* → ∞. Then, if *r*= *N*, *α*–*la ^{N}* –

*C*(

*a*) →

^{N}*α*–

*l*–

*C*(1), where the limit of

*a*is 1. If

^{N}*r*=

*L*and

*d*≠

*l*,

*α*–

*la*–

^{l}*C*(

*a*) →

^{L}*α*–

*lκ*–

^{L}*C*(

*κ*), where

^{L}*κ*∈(0,1) is the limit of

^{L}*a*defined by –

^{L}*C'*(

*κ*) =

^{L}*d*. Thus, from eq. [58] it follows that, because the term [

*n*(

*n*+ 1)]

^{–1}→ 0 as

*n*→ ∞,

*T*

_{1}

*→ 0 for*

^{r}*r*=

*N*and for

*r*=

*L*when

*d*≠

*l*.

Consider, next, eq. [56]. Using eq. [37] again, we can write

From the argument above, the limit of *α*–*la ^{r}* –

*C*(

*a*) as

^{r}*n*→ ∞ is finite for

*r*=

*N*and for

*r*=

*L*when

*d*≠

*l*. Because the term

*n*[

*β*(

*n*+ 1)]

^{–2}→ 0 as

*n*→ ∞, thus,

*T*

_{2}

*→ 0 for*

^{r}*r*=

*N*and for

*r*=

*L*when

*d*≠

*l*.

Finally, consider eq. [57]. Using eqs. [46] and [56], we can write eq. [57] as

Suppose, first, that *r*=*N*. Then, from eq. [48], it follows that

Thus, *β*^{–1} · [*l* + *C'*(1)] · [(–*l*)/*C"*(1)] = *β*^{–1} · (–*l*^{2}) · [*C"*(1)]^{–1}≡*τ ^{N} <*0 as

*n*→ ∞.

Suppose, second, that *r*= *L* and *d*≠*l*. Then,

and *T*_{3}* ^{L}* →

*β*

^{–1}· [

*l*–

*d*] · [(–(

*l*–

*d*))/

*C"*(

*κ*)] =

^{L}*β*

^{–1}· [(–(

*l*–

*d*)

^{2})/

*C"*(

*κ*)]≡

^{L}*τ*as

^{L}*n*→ ∞. Note that

*τ*0 both when

^{L}<*d*>

*l*and when

*d<l*.

Thus, as *n* → ∞, *T*_{1}* ^{r}*· [

*T*

_{2}

*+*

^{r}*T*

_{3}

*] → 0 · [0 +*

^{r}*τ*] = 0, where

^{r}*τ*0 for

^{r}<*r*=

*N*and for

*r*=

*L*when

*d*≠

*l*. For

*r*=

*N*and for

*r*=

*L*when

*d*≠

*l*, by continuity, there, thus, exists an

*n*

_{0}

*<∞ such that*

^{r}*dW*/

^{r}*dn*<0 at

*n*=

*n*

_{0}

*and, for*

^{r}*n*>

*n*

_{0}

*,*

^{r}*dW*/

^{r}*dn*→ 0 as

*n*→ ∞. Hence, for

*r*=

*N*and for

*r*=

*L*when

*d*≠

*l*,

*W*attains maximum at

^{r}*n <*∞.

## Appendix B

This appendix demonstrates that, *in instances when the industry reputation effect is negative*, first, social welfare can decrease with the number of firms *n* and, second, the relationship between social welfare and the number of firms *n* can be either monotonic or non-monotonic.

Recall from Proposition 4 that the industry reputation effect is negative for *r*= *N* and for *r*= *L* when *d*≠*l*. Throughout this appendix, we assume that the per-unit cost of production equals

where *ρ*>0. Thus, *C*′(*a _{i}*)=–

*ρ*(1 –

*a*)<0 for

_{i}*a*∈(0,1) and

_{i}*C*"(

*a*)=

_{i}*ρ*>0. Accordingly, from eq. [8]

and from eq. [10]

Similarly, from eq. [13]

and from eq. [15]

Social welfare under laissez-faire (*W ^{N}*) is then obtained by substituting in eq. [65] into eq. [26]. Social welfare under strict producer liability (

*W*) is calculated by substituting in eq. [67] into eq. [26].

^{L}Using thus-obtained expressions for *W ^{N}* and

*W*, we plot social welfare under a given regime against

^{L}*n*for different values of the model’s parameters in Figures 1–6. Figures 1–6 use parameter values such that firm and industry output under any legal regime are strictly positive for all

*n*>1. Moreover, Figures 3–6 use parameter values such that the equilibrium accident probability under strict producer liability, as defined in eq. [66], is between 0 and 1 for any

*n*>1.

Figures 1–6 illustrate that the relationship between social welfare under a given regime for which the industry reputation effect is negative, and the number of firms *n*, varies with values of the model’s parameters. Specifically, in Figures 2, 4, and 6, the relationship between social welfare and *n* is monotonic. In contrast, in Figures 1, 3, and 5, the relationship between social welfare and *n* is non-monotonic. In all of the examples, however, social welfare is maximized for some *n*<∞ (see Proposition 5).

### Figure 1

### Figure 2

### Figure 3

### Figure 4

### Figure 5

### Figure 6

### References

Avraham, R. 2006. “Putting a Price on Pain-and-Suffering Damages: A Critique of the Current Approaches and a Preliminary Proposal for Change.” Northwestern University Law Review100(1):87–120.Search in Google Scholar

Baniak, A., and P.Grajzl. 2013. “Equilibrium and Welfare in a Model of Torts with Industry Reputation Effects.” Review of Law and Economics9(2):265–302.10.1515/rle-2013-0039Search in Google Scholar

Barnett, M. L., and A. J.Hoffman. 2008. “Beyond Corporate Reputation: Managing Reputational Interdependence.” Corporate Reputation Review11(1):1–9.10.1057/crr.2008.2Search in Google Scholar

Baumann, F., and T.Friehe. 2012. “Optimal Damages Multipliers in Oligopolistic Markets” University of Konstanz, Department of Economics Working Paper Series 2012–08.Search in Google Scholar

Carriquiry, M., and B. B.Babcock. 2007. “Reputations, Market Structure, and the Choice of Quality Assurance Systems in the Food Industry.” American Journal of Agricultural Economics89(1):12–23.10.1111/j.1467-8276.2007.00959.xSearch in Google Scholar

Chen, Y., and X.Hua. 2012. “Ex Ante Investment, Ex Post Remedies, and Product Liability.” International Economic Review53(3):845–66.10.1111/j.1468-2354.2012.00703.xSearch in Google Scholar

Chiang, S-C. and R. T.Masson. 1988. “Domestic Industrial Structure and Export Quality.” International Economic Review29(2):261–70.10.2307/2526665Search in Google Scholar

Daughety, A. F., and J. F.Reinganum. 1995. “Product Safety: Liability, R&D, and Signaling.” American Economic Review85(5):1187–206.Search in Google Scholar

Daughety, A. F., and J. F.Reinganum. 1997. “Everybody Out Of the Pool: Products Liability, Punitive Damages, and Competition.” Journal of Law, Economics, and Organization13(2):410–32.10.1093/oxfordjournals.jleo.a023390Search in Google Scholar

Daughety, A. F., and J. F.Reinganum. 2006. “Markets, Torts, and Social Inefficiency.” RAND Journal of Economics37(2):300–23.10.1111/j.1756-2171.2006.tb00017.xSearch in Google Scholar

Daughety, A. F., and J. F.Reinganum. 2012. “Cumulative Harm and Resilient Liability Rules for Product Markets.” Journal of Law, Economics, and Organization. Forthcoming. doi:10.1093/jleo/ews045.Search in Google Scholar

Daughety, A. F., and J. F.Reinganum. 2014. “Economic Analysis of Products Liability: Theory.” In Research Handbook on the Economics of Torts, edited by J.Arlen, 69–96. Cheltenham: Edward Elgar.10.4337/9781781006177.00011Search in Google Scholar

De Geest, G., and G.Dari-Mattiacci. 2007. “Soft Regulators, Tough Judges.” Supreme Court Economic Review15(1):119–40.10.1086/656029Search in Google Scholar

Epple, D., and A.Raviv. 1978. “Product Safety: Liability Rules, Market Structure, and Imperfect Information.” American Economic Review68(1):80–95.Search in Google Scholar

Fleckinger, P. 2007. “Collective Reputation and Market Structure: Regulating the Quality vs. Quantity Trade-Off.” Ecole Polytechnique Cahier nº 2007–26.Search in Google Scholar

Geistfeld, M. A. 2009. “Products Liability.” In Tort Law and Economics, Encyclopedia of Law and Economics, Second Edition, edited by M.Faure, 287–340. Cheltenham: Edward Elgar.10.4337/9781848447301.00021Search in Google Scholar

Hattori, K., and T.Yoshikawa. 2013. “Free Entry and Social Inefficiency under Co-opetition.” MPRA Working Paper No. 44816.Search in Google Scholar

King, A. A., M. J.Lenox, and M. L.Barnett. 2002. “Strategic Responses to the Reputation Commons Problem.” In Organizations, Policy and the Natural Environment: Institutional and Strategic Perspectives, edited by A. J.Hoffman and M. J.Ventresca, 393–406. Stanford: Stanford University Press.Search in Google Scholar

Landes, W. M., and R. A.Posner. 1985. “A Positive Economic Analysis of Products Liability.” Journal of Legal Studies14(3):535–67.10.1086/467785Search in Google Scholar

Levin, J. 2009. “The Dynamics of Collective Reputation.” The B.E. Journal of Theoretical Economics9(1):Article 27.10.2202/1935-1704.1548Search in Google Scholar

Mankiw, N. G., and M. D.Whinston. 1986. “Free Entry and Social Efficiency.” RAND Journal of Economics17(1):48–58.10.2307/2555627Search in Google Scholar

Marette, S. 2007. “Minimum Safety Standards, Consumers’ Information and Competition.” Journal of Regulatory Economics32:259–85.10.1007/s11149-007-9036-xSearch in Google Scholar

Marino, A. M. 1988a. “Monopoly, Liability and Regulation.” Southern Economic Journal54(4):913–27.Search in Google Scholar

Marino, A. M. 1988b. “Products Liability and Scale Effects in a Long-Run Competitive Equilibrium.” International Review of Law and Economics8:97–107.10.1016/0144-8188(88)90018-XSearch in Google Scholar

Marino, A. M. 1991. “Market Share Liability and Economic Efficiency.” Southern Economic Journal57(3):667–75.10.2307/1059781Search in Google Scholar

McQuade, T., S. W.Salant, and J.Winfree. 2010. “Markets with Untraceable Goods of Unknown Quality: A Market Failure Exacerbated by Globalization.” Resources for the Future Discussion Paper 09–31.10.2139/ssrn.1588241Search in Google Scholar

McQuade, T., S. W.Salant, and J.Winfree. 2012. “Regulating an Experience Good Produced in the Formal Sector of a Developing Country When Consumers Cannot Identify Producers.” Review of Development Economics16(4):512–26.10.1111/rode.12001Search in Google Scholar

“Note: Deception as an Antitrust Violation.” 2012. Harvard Law Review125(5):1235–55.Search in Google Scholar

Polinsky, A. M., and W. P.Rogerson. 1983. “Products Liability, Consumer Misperceptions, and Market Power.” Bell Journal of Economics14(2):581–9.10.2307/3003659Search in Google Scholar

Polinsky, A. M., and S.Shavell. 2010. “The Uneasy Case for Product Liability.” Harvard Law Review123:1437–92.Search in Google Scholar

Pistor, K., and C.Xu. 2004. “Incomplete Law.” NYU Journal of International Law and Politics35:931–1013.Search in Google Scholar

Pouliot, S., and D. A.Sumner. 2010. “Traceability, Product Recalls, Industry Reputation and Food Safety.” Unpublished manuscript.Search in Google Scholar

Rouviere, E., and R.Soubeyran. 2011. “Competition vs. Quality in an Industry with Imperfect Traceability.” Economics Bulletin31(4):3052–67.Search in Google Scholar

Rubin, P. H. 2011. “Markets, Tort Law, and Regulation to Achieve Safety.” Cato Journal31(2):217–36.Search in Google Scholar

Segerson, K. 1999. “Mandatory Versus Voluntary Approaches to Food Safety.” Agribusiness15(1):53–70.10.1002/(SICI)1520-6297(199924)15:1<53::AID-AGR4>3.0.CO;2-GSearch in Google Scholar

Shavell, S. 1980. “Strict Liability Versus Negligence.” Journal of Legal Studies9(1):1–25.10.1086/467626Search in Google Scholar

Shavell, S. 1984. “Liability for Harm Versus Regulation of Safety.” Journal of Legal Studies13(2):357–74.10.1086/467745Search in Google Scholar

Shavell, S. 2004. Foundations of Economic Analysis of Law. Cambridge, MA: Belknap Press/Harvard University Press.10.4159/9780674043497Search in Google Scholar

Shavell, S. 2007. “Liability for Accidents.” In: Handbook of Law and Economics, Vol. I, edited by S.Shavell, and A.Mitchell Polinsky, 139–182. Amsterdam, The Netherlands: Elsevier.10.1016/S1574-0730(07)01002-XSearch in Google Scholar

Spence, M. 1977. “Consumer Misperceptions, Product Failure, and Producer Liability.” Review of Economic Studies44(3):561–72.10.2307/2296908Search in Google Scholar

Spulber, D. F. 1988. “Products Liability and Monopoly in a Contestable Market.” Economica55(219):333–41.10.2307/2554011Search in Google Scholar

Stucke, M. E. 2013. “Is Competition Always Good?” Journal of Antitrust Enforcement1(1):162–97.10.1093/jaenfo/jns008Search in Google Scholar

Tirole, J. 1996. “A Theory of Collective Reputations.” Review of Economic Studies63(1):1–22.10.2307/2298112Search in Google Scholar

Viscusi, W. K. 2012. “Does Product Liability Make Us Safer?” Regulation35(1):24–31.10.2139/ssrn.1770031Search in Google Scholar

Winfree, J. A., and J. J.McCluskey. 2005. “Collective Reputation and Quality.” American Journal of Agricultural Economics87(1):206–13.10.1111/j.0002-9092.2005.00712.xSearch in Google Scholar

- 1
Fleckinger (2007, 1), for example, describes the environments in which firms share a common reputation as “an intermediate situation between the perfect information and the asymmetric information setting”.

- 2
For recent surveys of the voluminous literature on products liability, see Daughety and Reinganum (2014), Geistfeld (2009), and Shavell (2004, 2007).

- 3
Baniak and Grajzl (2013) study the interaction of strict producer liability and industry reputation effects in a model of torts where harm occurs in a non-market setting.

- 4
See Hattori and Yoshikawa (2013) for a model with endogenous firm investments in a common property resource that could be interpreted as representing firms’ common reputation.

- 5
For discussion about the relationship between tort law and antitrust law, see also “Note: Deception as an Antitrust Violation” (2012).

- 6
The traditional products liability model summarized by Daughety and Reinganum (2014) in fact implies that profit-maximizing firms choose the socially optimal level of precaution regardless of the liability rule in place. Firms choose socially suboptimal precaution when, for example, consumers systematically misperceive the risk of harm (Spence 1977; Polinsky and Rogerson 1983; Marino 1988a). However, the level of precaution selected by firms in models where consumers misperceive risk, but at the same time all other assumptions of the traditional products liability model continue to hold, is still independent of market structure.

- 7
For analyses allowing for “scale effects” in expected harm, see Marino (1988a, 1988b) and Spulber (1988). For models featuring fixed costs of safety, see Daughety and Reinganum (2006) and Baumann and Friehe (2012).

- 8
Chiang and Masson (1988) do not provide a fully fledged welfare analysis. However, they do show that in the presence of a common industry reputation, when firms have no market power, labor is the only input, and returns to scale are constant, firm consolidation leads to higher wages.

- 9
The literature, of course, suggests reasons other than the existence of a common industry reputation for why greater intra-industry competition, as captured by the number of firms in an industry, need not increase social efficiency. One prominent reason is the existence of fixed set-up costs that firms must incur upon entry; see, e.g. Mankiw and Whinston (1986) and references therein. Another is the application of the strict market share liability rule; see Marino (1991). Stucke (2013) provides an illuminating discussion of when competition leads to suboptimal results.

- 10
Chen and Hua (2012) do not study common industry reputation effects. Instead, they focus on the impact of full producer liability, partial producer liability, and punitive damages on monopolist’s incentives to increase product safety through ex ante investment when the firm can also take ex post (i.e. after sales) remedial measures concerning product quality.

- 11
Assuming risk aversion rather than risk neutrality would render the model intractable, but would, we anticipate, not change the key qualitative insights.

- 12
In the related products liability literature that does not study implications of firms sharing a common industry reputation, linear demand is directly assumed (as opposed to derived from an underlying utility function) for example in Polinsky and Rogerson (1983) and Daughety and Reinganum (1995). In contrast, in the industrial organization literature on common industry reputation, Fleckinger (2007), for example, starts from a setup with heterogeneous consumers and develops a multiplicative inverse demand.

- 13
Another scenario that gives rise to free-riding in firms’ choice of precaution is the application of the market share liability rule, under which firms are held strictly liable for their market share of the total damages caused by the industry; see Marino (1991).

- 14
When consumers’ knowledge of risk of failure is perfect, the precise allocation of liability for losses from defective products does not matter for firms’ investments in precaution under the assumptions of the traditional products liability model (see, e.g. Shavell 1980; Landes and Posner 1985). Laissez-faire performs just as well as strict producer liability. Given administrative and litigation costs associated with products liability, laissez-faire is in fact the preferred regime.

**Published Online:**2014-5-3

**Published in Print:**2014-10-1

©2014 by De Gruyter