Abstract
This paper discusses different possibilities of licensing between Cournot duopolists with a technology characterized by constant marginal cost and positive fixed-cost. It is shown that optimal fixed-fee and per-unit royalty can be negative. This implies that the licensor can subsidize the licensee in the equilibrium to license its technology even if they compete in the output market in the post-licensing stage. Moreover superiority of up-front fee as an advance on royalties is also identified. The firm having lower marginal cost and higher fixed cost prefers royalty licensing than licensing by fixed-fee, while the other firm having lower fixed cost and higher marginal cost sometimes prefers fixed-fee licensing. Interestingly, a firm that earns less profit in the no-licensing stage may become the licensor of its technology.
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs10842-015-0194-8/MediaObjects/10842_2015_194_Fig1_HTML.gif)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs10842-015-0194-8/MediaObjects/10842_2015_194_Fig2_HTML.gif)
Similar content being viewed by others
Notes
See Zachary (1999)
See Reference for Business, Encyclopaedia of Business, 2nd ed.
Published November 4, 2013 by Nokia - Press Release.
Published December 3, 2007 by Toshiba - Press Release.
Microsoft News Centre https://www.microsoft.com/en-us/news/press/2008/jan08/01-15msjvcpr.aspx
Sen and Stamatopoulos (2009) consider the licensing by an outside innovator with technology characterized by zero fixed cost and identical linear marginal cost function. Beladi and Mukherjee (2013) consider technologies exhibiting significant decreasing returns in absence of any fixed-cost. Mukherjee (2013) shows licensing of a new product to all potential licensees in the presence of convex costs, which occur under decreasing returns to scale technologies.
As reported by Korea Technology Transfer Center (KTTC), see http://www.wipo.int/export/sites/www/uipc/en/documents/pdf/tmc-korea.pdf
According to Anderson (2013) a negative royalty rate might not be an abnormality if the patentee can offer other services that reduce the marginal costs of the licensee in addition to the innovation itself. It argues that if the innovation involves some kind of new machinery, the contract may also commit the patentee to provide maintenance of the equipment, thereby reducing the marginal costs further than the licensee could have managed on its own.
COMMISSION REGULATION (EU) No 316/2014 of 21 March 2014 on the application of Article 101(3) of the Treaty on the Functioning of the European Union to categories of technology transfer agreements.
It can be assumed that the firms competing like Cournot duopolists are catering a specific part of the market as the government is restricting the rest of the market to a competitive fringe.
The author wishes to thank an anonymous referee for pointing out this fact.
See Appendix 3 for the general demand framework.
This has been relaxed for two-part tariff licensing.
In the pre-licensing stage output of firm i is \(q_{i}=\frac {a-2c_{i}+c_{j}}{3}\), i,j = 1,2 and i≠j. However, in the post licensing stage \(q_{1}^{T_{2}}=q_{2}^{T_{2}}=\frac {a-c_{2}}{3}\), where \(q_{1}^{T_{2}}\) and \(q_{2}^{T_{2}}\) are the output of firm 1 and firm 2 respectively.
As \(0<L(c_{1})<\bar {F_{2}}\) and \(F_{2}-F_{1}\in (0,\bar {F_{2}})\), this is possible.
In the pre-licensing stage output of firm i is \(q_{i}=\frac {a-2c_{i}+c_{j}}{3}\), i,j = 1,2 and i≠j. However, in the post licensing stage \(q_{1}^{T_{1}}=q_{2}^{T_{1}}=\frac {a-c_{1}}{3}\), where \(q_{1}^{T_{1}}\) and \(q_{2}^{T_{1}}\) are the output of firm 1 and firm 2 respectively.
As \(\frac {d{\Pi }_{2}^{r_{2}}}{dr_{2}}=\frac {5(a-c_{2}-2r_{2})}{9}\) must be positive as otherwise relation (3) will not be satisfied.
In the pre-licensing stage output of firm i is \(q_{i}=\frac {a-2c_{i}+c_{j}}{3}\), i,j = 1,2 and i≠j. However, in the post licensing stage \(q_{1}^{r_{2}}=\frac {a-2(c_{2}+\bar {r_{2}})+c_{2}}{3}\) and \(q_{2}^{r_{2}}=\frac {a-2c_{2}+(c_{2}+\bar {r_{2}})}{3}\), where \(q_{1}^{r_{2}}\) and \(q_{2}^{r_{2}}\) are the output of firm 1 and firm 2 respectively.
As \(\frac {d{\Pi }_{1}^{r_{1}}}{dr_{1}}=\frac {5(a-c_{1}-2r_{1})}{9}\) must be positive as otherwise relation (7) will not be satisfied.
As \(0<L(c_{1})<\bar {F_{2}}\) and \(F_{2}-F_{1}\in (0,\bar {F_{2}})\), this is possible.
In the pre-licensing stage output of firm i is \(q_{i}=\frac {a-2c_{i}+c_{j}}{3}\), i,j = 1,2 and i≠j. However, in the post licensing stage \(q_{2}^{r_{1}}=\frac {a-2(c_{1}+\bar {r_{1}})+c_{1}}{3}\) and \(q_{1}^{r_{1}}=\frac {a-2c_{1}+(c_{1}+\bar {r_{1}})}{3}\), where \(q_{2}^{r_{1}}\) and \(q_{1}^{r_{1}}\) are the output of firm 2 and firm 1 respectively.
\(q_{I}^{r_{2}}>q_{I}\) as \(c_{1}>c_{2}+\bar {r_{2}}\) which follows from equation (5).
It can be shown further that as firm 2 and firm 1 will employ royalty licensing and fixed-fee licensing (when fixed-fee is negative) respectively two-way transfer is possible as Y(c 1) is not always greater than X r(c 1). Two-way transfer is possible when Y(c 1)≤F 2−F 1≤X r(c 1).
As in the main text (Lemma 3), it can be argued that firm 2 will use the licensed technology even if T 1 < 0.
It is discussed in Appendix C.1, that if one of the firm’s unit cost increases then the industry output falls.
References
Anderson F (2013) Licensing to a More Efficient Rival. The Manchester School. doi:10.1111/manc.12036
Beggs AW (1992) The licensing of patents under asymmetric information. Int J Ind Organ 10:171–191
Beladi H, Mukherjee A (2013) Optimal licensing contract under convex costs. Mimeo
Caballero-Sanz F, Moner-Colonques R, Sempere-Monerris JJ (2002) Optimal Licensing in a Spatial Model. Annales d’Economie et de Statistique, ENSAE 66:257–279
Carvalho NP (2010) The TRIPS Regime of Patent Rights. Kluwer Law International
Eswaran M (1994) Cross-Licensing of Competing Patents as a Facilitating Device. Can J Econ 27:689–708
Fauli-Oller R, Sandonis J (2002) Welfare reducing licensing. Games Econ Behav 41:192–205
Fershtman C, Kamien MI (1992) Cross licensing of complementary technologies. Int J Ind Organ 10:329–348
Gallini NT, Wright BD (1990) Technology transfer under asymmetric information. Rand J Econ 21:147–160
Kamien MI, Tauman Y (1986) Fees Versus Royalties and the Private Value of a Patent. Q J Econ 101:471–492
Liao C, Sen D (2005) Subsidy in Licensing: Optimality and Welfare Implications. Manch Sch 73:281–299
Macho-Stadler I, Perez-Castrillo D (1991) Contrats de licences et asymtrie d’information. Annales d’Economie et de Statistique, ENSAE 24:189–208
Marjit S (1990) On a non-cooperative theory of technology transfer. Econ Lett 33:293–298
Mukherjee A (2002) Subsidy and Entry: the Role of Licensing. Oxf Econ Pap 54:160–171
Mukherjee A (2013) Licensing under convex costs. J Econ. doi:10.1007/s00712-013-0333-9
Muto S (1993) On licensing policies in Bertrand competition. Games Econ Behav 5:257–267
Ottoz E, Cugno F (2009) Hybrid Licensing of Product Innovations. Rev Law Econ 5:579–594
Rostoker MD (1984) A Survey of Corporate Licensing. IDEA: J Law Technol 24:59–92
Sen D, Tauman Y (2007) General Licensing Schemes for a Cost-reducing Innovation. Games Econ Behav 59:163–186
Sen D, Stamatopoulos G (2009) Technology transfer under returns to scale. Manch Sch 77:337–365
Sen N (2014) Unilateral technology licensing from an entrant to incumbent monopolist. Econ Bull 34:1028–1037
Shapiro C (1985) Patent Licensing R & D Rivalry. Am Econ Rev 75:25–30
Vishwasrao S (2007) Royalties vs. fees: How do firms pay for foreign technology? Int J Ind Organ 25:741–759
Vives X (2001) Oligopoly Pricing: Old ideas and new tools. MIT Press
Wang XH (1998) Fee versus royalty licensing in a Cournot duopoly model. Econ Lett 60:55–62
Zachary GP (1999) Many Industries Are Congealing Into Lineup of Few Dominant Giants. Wall Street J
Acknowledgments
I am indebted to Dr. Sukanta Bhattacharya, University of Calcutta and Mr. Rajit Biswas, Economic Research Unit, Indian Statistical Institute, Kolkata for their valuable comments and suggestions. The author is also thankful to Dr. Arijit Mukherjee, Nottingham University Business School and Dr. Debapriya Sen, Ryerson University for the discussions he had with them. I am indebted to the anonymous referees and the managing editor of this journal in upgrading this paper. The usual disclaimer applies.
Author information
Authors and Affiliations
Corresponding author
Appendices
The author wishes to thank an anonymous referee for suggesting this extension.
Appendix 1
From Eq. 5 of the main text we have
as firm 2 set the royalty rate as high as possible such that firm 1 is indifferent between licensing and not. From the above equation we get
and it must be positive because if \(a-c_{2}-2\bar {r_{2}}<0\) then \({\Pi }_{1}^{r_{2}}(\bar {r_{2}})<{\Pi }_{1}\) and licensing will never be possible. Hence, \(\bar {r_{2}}=\frac {1}{2}\left [(a-c_{2})-\sqrt {(a-2c_{1}+c_{2})^{2}+9(F_{2}-F_{1})}\right ]\) and substituting in \({\Pi }_{2}^{r_{2}}(\bar {r_{2}})\) (defined in the main text) gives
This is the net profit of firm 2 if licensing is undertaken. Moreover for licensing to be profitable for firm 2, \({\Pi }_{2}^{r_{2}}(\bar {r_{2}})\) must be greater than Π2. This is possible if
which is relation (6) of the main text, the necessary and sufficient condition for licensing with royalty from firm 2 to firm 1.
Appendix 2
As discussed in the main text firm 1 will charge the royalty rate as high as possible (\(\bar {r_{1}}\)) such that
At this royalty rate firm 2 is indifferent between licensing and not. Manipulating the above equation we get
Moreover \(a-c_{1}-2\bar {r_{1}}\) must be positive. As if \(a-c_{1}-2\bar {r_{1}}<0\) then \({\Pi }_{2}^{r_{1}}(\bar {r_{1}})<{\Pi }_{2}\) and licensing will never be possible as firm 2 will always reject the offer. From the above equation we thereby get \(\bar {r_{1}}=\frac {1}{2}[(a-c_{1})-\sqrt {(a-2c_{2}+c_{1})^{2}+9(F_{1}-F_{2})}]\). Further substituting \(\bar {r_{1}}\) in \({\Pi }_{1}^{r_{1}}(\bar {r_{1}})\) (defined in the main text) gives
Moreover \({\Pi }_{1}^{r_{1}}(\bar {r_{1}})\geq {\Pi }_{1}\) if
This is relation (9) of the main text which is the necessary and sufficient condition for licensing to be possible. As otherwise if the above inequality is not satisfied firm 1 will not be better off from licensing its technology and thereby not offer any royalty rate to firm 2.
Appendix 3 General Demand Structure
The author wishes to thank an anonymous referee for suggesting this extension.
1.1 Appendix 3.1 Basic Set-Up
The inverse market demand is P = f(q), where P is the price, q = q 1 + q 2 is the total output and q i , i = 1,2, is the output produced by the firm i. Assume same cost structure as in the main text.
Assumptions 1
-
i)
f ′(q)<0,
-
ii)
∃ q 0 > 0 such that f(q) > 0 for q < q 0 and f(q)≤0 for q≥q 0 and
-
iii)
f ′(q) + q i f ″(q)≤0 for i = 1,2.
For firm i the profit function is Π i = [f(q)−c i ]q i −F i , where c i and F i are the unit cost and fixed cost of firm i respectively. For profit maximization by firm i: F.O.C. is f(q)−c i + q i f ′(q)=0 and S.O.C. is 2f ′(q) + q i f ″(q)<0. The above assumptions are standard and guarantee the uniqueness and stability of Cournot equilibrium (see Vives 2001). These assumptions also ensure the uniqueness and stability of Cournot equilibrium is case of licensing. From the F.O.Cs. and S.O.Cs. it can be said that if the unit cost of any firm increases industry output will fall.
1.1.1 No Licensing
If there is no licensing, the profits of firm 1 and firm 2 are \({\Pi }_{1}=[f(q^{*})-c_{1}]q_{1}^{*}-F_{1}\) and \({\Pi }_{2}=[f(q^{*})-c_{2}]q_{2}^{*}-F_{2}\) respectively; \(q^{*}=q_{1}^{*}+q_{2}^{*}\), where \(q_{1}^{*}\) and \(q_{2}^{*}\) are the optimal output of the firm 1 and firm 2 respectively. As c 1 > c 2, therefore \(q_{1}^{*}<q_{2}^{*}\). These profits are assumed to be positive, which implies that \(F_{2}\in \left (0,\bar {F_{2}}=[f(q^{*})-c_{2}]q_{2}^{*}\right )\), \(F_{1}\in \left (0,\bar {F_{1}}=[f(q^{*})-c_{1}]q_{1}^{*}\right )\) and \(F_{2}-F_{1}\in (0,\bar {F_{2}})\) as F 2 > F 1.
1.2 Appendix 3.2 Fixed-Fee Licensing
Firm i offers a fixed-fee (T i ) to license its technology to firm j. The profit(net) of the firm i and firm j respectively are \({\Pi }_{i}^{T_{i}}\) and \({\Pi }_{j}^{T_{i}}\) for i,j = 1,2 when the offer of T i is accepted by firm j and technology is transferred.
1.2.1 Offer by Firm 2
Firm 2 offers T 2 such that \({\Pi }_{1}^{T_{2}}=[f(2q^{a})-c_{2}]q^{a}-F_{2}-T_{2}={\Pi }_{1}\), where q a is the output produced by each firm if firm 2 licenses its technology to firm 1. Therefore if \({\Pi }_{2}^{T_{2}}=[f(2q^{a})-c_{2}]q^{a}-F_{2}+T_{2}\geq {\Pi }_{2}\), such that licensing is profitable for firm 2 or
technology will be licensed at T 2 > 0. Moreover after licensing,
As the unit cost of firm 1 reduces after transfer the industry output also increases after transfer, or \(2q^{a}>q_{1}^{*}+q_{2}^{*}\).
1.2.2 Offer by firm 1
Firm 1 offers T 1 such that \({\Pi }_{2}^{T_{1}}=[f(2q^{b})-c_{1}]q^{b}-F_{1}-T_{1}={\Pi }_{2}\), where q b is the output produced by each firm if firm 1 licenses its technology to firm 2.Footnote 28 As the unit cost of firm 2 increases after transfer the industry output gets reduced after transfer, or \(2q^{b}<q_{1}^{*}+q_{2}^{*}\). Moreover, \({\Pi }_{1}^{T_{1}}=[f(2q^{b})-c_{1}]q^{b}-F_{1}+T_{1}\geq {\Pi }_{1}\) must hold, such that licensing is profitable for firm 1 or
If technology is licensed then
and T 1≥0 if \(L=[f(q^{*})-c_{2}]q_{2}^{*}-[f(2q^{b})-c_{1}]q^{b}\leq F_{2}-F_{1}\) (this is possible as \([f(2q^{b})-c_{1}]q^{b}>[f(q^{*})-c_{1}]q_{1}^{*}\)) and T 1 < 0 otherwise.
Moreover, assume Π2 > Π1 or \(J=[f(q^{*})-c_{2}]q_{2}^{*}-[f(q^{*})-c_{1}]q_{1}^{*}>F_{2}-F_{1}\) and Y < J (as \([f(2q^{b})-c_{1}]q^{b}>[f(q^{*})-c_{1}]q_{1}^{*}\)). Therefore if Y≤F 2−F 1 < J, firm 1 even though earning a lower profit in the no-licensing stage, will license its technology to firm 2.
1.3 Appendix 3.3 Royalty Licensing
Firm i offers a per-unit royalty (r i ) to license its technology to the other firm. Let the profit(net) of the firm i and firm j are respectively \({\Pi }_{i}^{r_{i}}\) and \({\Pi }_{j}^{r_{i}}\) for i,j = 1,2 when the offer of r i is accepted by firm j and technology is transferred.
1.3.1 Offer by Firm 2
Firm 2 offers a per-unit royalty (r 2) to firm 1 to license its technology. Let the profits of firm 1 and firm 2 respectively after the licensing be \({\Pi }_{1}^{r_{2}}=[f(q_{1}^{r_{2}}+q_{2}^{r_{2}})-c_{2}-r_{2}]q_{1}^{r_{2}}-F_{2}\) and \({\Pi }_{2}^{r_{2}}=[f(q_{1}^{r_{2}}+q_{2}^{r_{2}})-c_{2}]q_{2}^{r_{2}}-F_{2}+r_{2}q_{1}^{r_{2}}\), where \(q_{i}^{r_{2}}\) is the output of firm i.
The F.O.Cs for profit maximization of firm 1 and firm 2 respectively for given r 2 are
and the S.O.Cs are \(2f^{\prime }(q_{1}^{r_{2}}+q_{2}^{r_{2}})+q_{i}^{r_{2}}f^{\prime \prime }(q_{1}^{r_{2}}+q_{2}^{r_{2}})<0\) for i = 1,2.
Using the F.O.Cs and S.O.Cs it can be shown that
as \(\frac {\delta q_{1}^{r_{2}}}{\delta r_{2}}<0\). Therefore firm 2 will charge r 2 as high as possible such that \({\Pi }_{1}^{r_{2}}={\Pi }_{1}\) but c 2 + r 2 < c 1, as otherwise technology will never be licensed due to the increase in fixed-cost for firm 1. Finally technology is licensed to firm 1 if \({\Pi }_{2}^{r_{2}}\geq {\Pi }_{2}\) or
As, \(q_{1}^{r_{2}}+q_{2}^{r_{2}}>q_{1}^{*}+q_{2}^{*}\) and \(q_{2}^{r_{2}}<q_{2}^{*}\), r 2 must be positive. Therefore technology is licensed if
and
1.3.2 Offer by Firm 1
Firm 1 offers a per-unit royalty (r 1) to firm 2 to license its technology. Let the profits of firm 2 and firm 1 respectively after the licensing be \({\Pi }_{2}^{r_{1}}=[f(q_{1}^{r_{1}}+q_{2}^{r_{1}})-c_{1}-r_{1}]q_{2}^{r_{1}}-F_{1}\) and \({\Pi }_{1}^{r_{1}}=[f(q_{1}^{r_{1}}+q_{2}^{r_{1}})-c_{1}]q_{1}^{r_{1}}-F_{1}+r_{1}q_{2}^{r_{1}}\), where \(q_{i}^{r_{1}}\) is the output of firm i.
The F.O.Cs for profit maximization of firm 2 and firm 1 respectively given r 1 are
and the S.O.Cs are \(2f^{\prime }(q_{1}^{r_{1}}+q_{2}^{r_{1}})+q_{i}^{r_{1}}f^{\prime \prime }(q_{1}^{r_{1}}+q_{2}^{r_{1}})<0\) for i = 1,2.
As in the previous case, using the F.O.Cs and S.O.Cs it can be shown that
Therefore firm 1 will charge r 1 as high as possible such that \({\Pi }_{2}^{r_{1}}={\Pi }_{2}\). Finally firm 1 will license technology if \({\Pi }_{1}^{r_{1}}\geq {\Pi }_{1}\) or
Therefore technology is licensed if
and
Moreover, r 1≥0 if \(L=[f(q^{*})-c_{2}]q_{2}^{*}-[f(2q^{b})-c_{1}]q^{b}\leq F_{2}-F_{1}\) (this is possible as \([f(2q^{b})-c_{1}]q^{b}>[f(q^{*})-c_{1}]q_{1}^{*}\)) and r 1 < 0 otherwise.
Moreover, assume \(J=[f(q^{*})-c_{2}]q_{2}^{*}-[f(q^{*})-c_{1}]q_{1}^{*}>F_{2}-F_{1}\) such that Π2 > Π1. Here also Y r < J (as \([f(q_{1}^{r_{1}}+q_{2}^{r_{1}})-c_{1}]q_{1}^{r_{1}}>[f(q^{*})-c_{1}]q_{1}^{*}\)). Therefore if Y r≤F 2−F 1 < J, firm 1 even though earning a lower profit in the no-licensing stage, will license its technology to firm 2.
1.4 Appendix 3.4 Royalty vs Fixed-Fee
1.4.1 Firm 2’s Choice
Let us assume that Eqs. C.1 and C.7 are satisfied simultaneously such that firm 2 transfers its technology both in case of fixed-fee as well as royalty licensing. In case of fixed-fee licensing firm 2 gets (as in Eq. C.2)
and in case of royalty licensing it gets (as in Eq. C.8)
Let
where G 2 is the excess profit that firm 2 earns in case of royalty licensing than in case of fixed-fee.
If optimal r 2 = 0, then G 2 = 0 as \({\Pi }_{2}^{r_{2}}={\Pi }_{2}^{T_{2}}\) because \(q^{a}=q_{1}^{r_{2}}=q_{2}^{r_{2}}\). Firm 2 always sets optimal r 2 such that \({\Pi }_{1}^{r_{2}}={\Pi }_{1}\). When optimal r 2 = 0 then
In case of fixed-fee licensing also, firm 2 sets optimal T 2 such that \({\Pi }_{1}^{T_{2}}={\Pi }_{1}\). The above equation thus implies that if optimal r 2 = 0 then optimal fixed-fee T 2 = 0 as \({\Pi }_{1}^{T_{2}}=[f(2q^{a})-c_{2}]q^{a}-F_{2}-T_{2}=[f(q^{*})-c_{1}]q_{1}^{*}-F_{1}={\Pi }_{1}\). So it can be argued that if c 2 = c 1 and F 2 = F 1 then technology will transferred but r 2 = 0 and T 2 = 0.
Hence, if c 1 > c 2 and F 1 < F 2 (as assumed in the model), then technology if transferred will be at r 2 > 0 (via royalty) or T 2 > 0 (via fixed-fee) as discussed earlier in the appendix. However (using Eq. C.5)
as \(q_{1}^{r_{2}}>0,f^{\prime }(.)<0\) and \([\frac {dq_{2}^{r_{2}}}{d{r_{2}}}+\frac {dq_{2}^{r_{2}}}{d{r_{2}}}]<0\). Therefore \(G_{2}={\Pi }_{2}^{r_{2}}-{\Pi }_{2}^{T_{2}}>0\) always as optimal r 2 > 0, if firm 2 licenses its technology by per-unit royalty. Hence, firm 2 will prefer royalty licensing than licensing by fixed-fee.
Alternative proof From the Eq. C.13 \([f(q_{1}^{r_{2}}+q_{2}^{r_{2}})-c_{2}](q_{1}^{r_{2}} +q_{2}^{r_{2}})\) and 2[f(2q a)−c 2]q a can be considered as the industry profit, when the unit cost of production is c 2 (for both the firms) in absence of fixed cost, if the industry output is \(q_{1}^{r_{2}}+ q_{2}^{r_{2}}\) and 2q a respectively. Moreover, \(q^{m}<q_{1}^{r_{2}}+q_{2}^{r_{2}}<2q^{a}\) (where q m is the monopoly output produced by a firm using similar technology). As under royalty licensing (r 2 > 0) the effective unit cost of firm 1 is r 2 + c 2 for which the industry output is lower in case of royalty licensing than in case of fixed-fee.Footnote 29 If the industry output is above q m (which maximizes the industry profit), the industry profit falls if output increases. Hence, \([f(q_{1}^{r_{2}}+q_{2}^{r_{2}})-c_{2}](q_{1}^{r_{2}}+q_{2}^{r_{2}})\) must be greater than 2[f(2q a)−c 2]q a. Therefore \(G_{2}={\Pi }_{2}^{r_{2}}-{\Pi }_{2}^{T_{2}}>0\) always as optimal r 2 > 0 (and T 2 > 0). Hence, firm 2 will prefer royalty licensing than licensing by fixed-fee.
1.4.2 Firm 1’s choice
Let us assume that Eqs. C.3 and C.11 are satisfied simultaneously such that firm 1 transfers its technology both in case of fixed-fee as well as royalty licensing. In case of fixed-fee licensing firm 1 gets (as in Eq. C.4)
and in case of royalty licensing it gets (as in Eq. C.12)
Let
where G 1 is the excess profit that firm 1 earns in case of royalty licensing than in case of fixed-fee.
If optimal r 1 = 0, then G 1 = 0 as \({\Pi }_{1}^{r_{1}}={\Pi }_{1}^{T_{1}}\) because \(q^{b}=q_{1}^{r_{1}}=q_{2}^{r_{1}}\). Firm 1 always sets optimal r 1 such that \({\Pi }_{2}^{r_{1}}={\Pi }_{2}\). When optimal r 1 = 0 then
In case of fixed-fee licensing also, firm 1 sets optimal T 1 such that \({\Pi }_{2}^{T_{1}}={\Pi }_{2}\). The above equation thus implies that if optimal r 1 = 0 then optimal fixed-fee T 1 = 0 as \({\Pi }_{2}^{T_{1}}=[f(2q^{b})-c_{1}]q^{b}-F_{1}-T_{1}=[f(q^{*})-c_{2}]q_{2}^{*}-F_{2}={\Pi }_{2}\).
So it can be argued that given c 1,F 2 a n d F 1 there exists a unique c 2 say \(c_{2}^{*}\) such that technology is transferred but r 1 = 0 and T 1 = 0. Hence, if \(c_{2}>c_{2}^{*}\), then technology if transferred will be at r 1 > 0 (via royalty) or T 1 > 0 (via fixed-fee) as Π2 falls if c 2 increases. On the other hand if \(c_{2}<c_{2}^{*}\), then technology may be transferred at r 1 < 0 (via royalty) or T 1 < 0 (via fixed-fee) as Π2 rises if c 2 falls. However (using Eq. C.9)
as \(q_{2}^{r_{1}}>0,f^{\prime }(.)<0\) and \([\frac {dq_{2}^{r_{1}}}{d{r_{1}}}+\frac {dq_{1}^{r_{1}}}{d{r_{1}}}]<0\). Therefore if firm 1 sets r 1 > 0, then \(G_{1}={\Pi }_{1}^{r_{1}}-{\Pi }_{1}^{T_{1}}>0\) and firm 1 will prefer royalty licensing than licensing by fixed-fee. However, if firm 1 sets r 1 < 0, then \(G_{1}={\Pi }_{1}^{r_{1}}-{\Pi }_{1}^{T_{1}}<0\) and firm 1 will prefer fixed-fee licensing than licensing by royalty. As firm 1 subsidizes (negative fixed-fee/negative royalty) if \(L=[f(q^{*})-c_{2}]q_{2}^{*}-[f(2q^{b})-c_{1}]q^{b}> F_{2}-F_{1}\), hence if L > F 2−F 1 firm 1 will prefer fixed-fee licensing and royalty licensing otherwise.
Alternative proof From the Eq. C.16 \([f(q_{1}^{r_{1}}+q_{2}^{r_{1}})-c_{1}](q_{1}^{r_{1}} +q_{2}^{r_{1}})\) and 2[f(2q b)−c 1]q b can be considered as the industry profit, when the unit cost of production is c 1 (for both the firms) in absence of fixed cost, if the industry output is \(q_{1}^{r_{1}}+q_{2}^{r_{1}}\) and 2q b respectively. Moreover, \(q^{m}<q_{1}^{r_{1}}+q_{2}^{r_{1}}<2q^{b}\) (where q m is the monopoly output produced by a firm using similar technology) because under royalty licensing if r 1 > 0 the effective unit cost of firm 1 is r 1 + c 1 for which the industry output is lower in case of royalty licensing than in case of fixed-fee. If the industry output is greater than q m (which maximizes the industry profit) the industry profit falls if output increases. Hence, \([f(q_{1}^{r_{1}}+q_{2}^{r_{1}})-c_{1}](q_{1}^{r_{1}}+q_{2}^{r_{1}})\) must be greater than 2[f(2q b)−c 1]q b. Therefore \(G_{1}={\Pi }_{1}^{r_{1}}-{\Pi }_{1}^{T_{1}}>0\) if optimal r 1 > 0 (and T 1 > 0). Hence, firm 1 will prefer royalty licensing than licensing by fixed-fee if optimal r 1 > 0 and T 1 > 0.
However, \(q^{m}<2q^{b}<q_{1}^{r_{1}}+q_{2}^{r_{1}}\) if r 1 < 0 as the effective unit cost of firm 1 is r 1 + c 1 (< c 1) for which the industry output is greater in case of royalty licensing than in case of fixed-fee. If the industry output is above q m (which maximizes the industry profit) the industry profit falls if output increases. Hence, \([f(q_{1}^{r_{1}}+q_{2}^{r_{1}})-c_{1}](q_{1}^{r_{1}}+q_{2}^{r_{1}})\) must be lower than 2[f(2q b)−c 1]q b. Therefore \(G_{1}={\Pi }_{1}^{r_{1}}-{\Pi }_{1}^{T_{1}}<0\) if optimal r 1 < 0 (and T 1 < 0). Hence, firm 1 will prefer fixed-fee licensing than licensing by royalty if optimal r 1 < 0 and T 1 < 0.
Rights and permissions
About this article
Cite this article
Sen, N. Technology Transfer in Oligopoly in Presence of Fixed-Cost in Production. J Ind Compet Trade 15, 383–409 (2015). https://doi.org/10.1007/s10842-015-0194-8
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10842-015-0194-8