真品平行輸入之經濟分析

行政院國家科學委員會專題研究計畫 成果報告

真品平行輸入之經濟分析

計畫類別： 個別型計畫 計畫編號： NSC90-2415-H-002-022-執行期間： 90 年 08 月 01 日至 91 年 12 月 31 日 執行單位： 國立臺灣大學財務金融學系暨研究所 計畫主持人： 陳業寧 報告類型： 精簡報告 報告附件： 國外研究心得報告 處理方式： 本計畫可公開查詢

中

華

民

國 92 年 6 月 27 日

行政院國家科學委員會補助專題研究計畫成果報告

真品平行輸入之經濟分析

The Economic Analysis of Parallel Imports

計畫類別：個別型計畫

計畫編號：NSC

90-2415-H-002-022

執行期間：2001 年 8 月 1 日至 2002 年 12 月 31 日

計畫主持人：陳業寧

共同主持人：

計畫參與人員： 王衍智、林岳祥

成果報告類型(依經費核定清單規定繳交)：精簡報告

本成果報告包括以下應繳交之附件：

赴國外出差或研習心得報告一份

執行單位：台灣大學財務金融學系

中 華 民 國 92 年 6 月

27

日

中英文摘要

1. 中文摘要 本研究計畫探討與真品平行輸入有關的兩個重要議題。一是真品平行輸入對於合 法代理商訂價及行銷策略之影響，二是政府對於真品平行輸入之管制如何影響社 會福利。本計畫發現，當政府改變對真品平行輸入的管制時，合法代理商行銷支 出的變動會受其均衡策略之影響。若均衡時代理商的策略並不杜絕平行輸入之商 品，則此時政府放寬對真品平行輸入的管制會使代理商增加其行銷支出。相對 的，若均衡時代理商採用限價策略杜絕平行商品的輸入，則政府放寬對真品平行 輸入的管制會使代理商減少其行銷支出。在福利政策方面，本計畫發現一般來 說，政府放寬對真品平行輸入的管制可能增加也可能減少社會福利。計畫中並討 論在哪些情況下放寬對真品平行輸入之管制可以提高社會福利。 關鍵詞：訂價策略、行銷策略、真品平行輸入、福利分析 2. 英文摘要

This project studies two important issues about parallel imports. First, how parallel imports affect pricing and advertising strategies of authorized dealers who sell

legitimate items. Second, how regulation on parallel imports will affect social welfare. We find that an authorized dealer may either increase or decrease advertising

expenses when the law becomes more favorable to parallel imports; he will increase advertising expenses if there is a gray market in equilibrium, and he will reduce advertising expenses if he engages in limit pricing in equilibrium. We also find that loosening restrictions on parallel imports may either increase or decrease social welfare. Conditions under which loosening restrictions on parallel imports can raise welfare are identified.

一、報告內容: 前言、研究目的與文獻整理請參考 Introduction 一節。研究方法 請參考 The Model 一節。結果分析與討論請參考 The Analysis of the Model、 Welfare Analysis、An Example 與 Concluding Remarks 等四節。

1. Intr oduction

Parallel imports, or gray market products, are not counterfeit. They are

legitimately-produced items intended by the manufacturer for another territory. This paper studies two important issues about parallel imports. First, how parallel imports affect pricing and advertising strategies of authorized dealers who sell legitimate products. Authorized dealers often claim that parallel imports reduce their incentives to advertise. They also want the public to believe that they will not lower the prices of the legitimate items as a response to the competition of gray items. For example, in 1998, Australia relaxed its laws against parallel imports. Reebok responded by

warning that it would reduce its advertising and sports sponsorship. It said: “We're not going to continue to market a brand for someone else to benefit from.” At the same time, the recorded music industry in Australia warned: “the price of retail CDs in Australia will not fall if parallel imports are permitted”. We will verify whether these claims hold under a general setting.

The second issue we investigate is how parallel imports should be regulated. It seems that different countries have different regulation on parallel imports. In 1998, the U.S. Supreme Court held that, under the first sale doctrine, parallel imports could not be restrained under copyright law. It is argued that, by the first sale doctrine, a manufacturer that sells an item cannot restrict the buyer from re-selling it except in a few specific ways. However, in the same year, the European Court of Justice held that parallel imports could be restrained under the European Trade Mark Directive. To see which kinds of policy can result in higher welfare, this paper studies the welfare effects of the regulation on parallel imports.

As to the literature, most papers that discuss parallel imports considered pricing and advertising decisions separately. One stream of research focused on the conditions under which price discrimination through parallel imports would raise the

manufacturer's profit and social welfare (Malueg and Schwartz 1994; Papadopoulos 2000; and Ahmadi and Yang 2000). The other stream of research considered the free-riding by parallel imports on the authorized dealer's advertising and how the authorized dealer should respond (Tan et.al. 1997; and Lee et.al. 2000). Kek and Yong (1999) did consider both the issues of price and advertising. They showed that, in deciding the legal status of parallel imports, society should trade off decreased advertising by the authorized dealer against lower prices arising from increased competition. However, in their model, the authorized dealer's advertising does not

benefit consumers of the parallel imports, hence the free-rider problem does not really arise.

In this paper, we analyze a model of parallel imports where an authorized dealer competes against gray market products. The gray item provides proportionately less benefit than the legitimate product. The authorized dealer’s decision variables are the price of the legitimate product and advertising expenses. His advertising benefits consumers of both legitimate and gray items. Let ë denote the probability that a gray

item can be successfully sold in the market. Obviously, the stricter the law is against parallel imports, the smaller ë is. We show the following results.

First, the relationship between the authorized dealer’s advertising expenses and ë is not monotonic. In case the dealer engages in limit pricing in equilibrium, an increase in ë will reduce his advertising expenses. On the other hand, in case there is a gray market in equilibrium, the dealer will increase his advertising expenses as ë increases. Second, an increase in ë may either increase or decrease social welfare. We will document factors that contribute to the welfare effects of ë.

Third, we conjecture that, when the size of the gray market is relatively large compared with the market for the legitimate product, increasing ë is likely to raise

welfare.

Finally, if ë is determined by the government, then the optimal ë may be affected by

the effectiveness of law enforcement. We conjecture that, if the government is not very effective in law enforcement, it should lift restrictions on parallel imports.

The rest of the paper is organized as follows. Section 2 states the model. Section 3 analyzes the model and demonstrates comparative static. Section 4 discusses the welfare effects of the regulation on parallel imports. Section 5 provides an example that illustrates some interesting policy implications. Section 6 concludes.

2. The Model

Consider an item that is legitimately marketed by a single profit-maximizing

authorized dealer in competition with a parallel import available in a gray market. The authorized dealer sets a price, pl, and advertising, á, for the legitimate product. The

variable á could represent any element of marketing strategy, including advertising

and promotion, that increases the benefit a potential consumer derives from both the legitimate and gray versions of the product. The advertising cost, K(á), is increasing,

convex, and satisfies K(0) = 0 and K'(0) = 0. Let w denote the marginal cost of

providing a unit of legitimate item; w may be the wholesale price set by the global

manufacturer on a worldwide basis.

The price of the gray item is pi. Each potential consumer chooses among buying the

legitimate product provides a benefit, á v. The gray item provides a benefit, (1 - â) á v,

which is proportionately lower by the quality differential â owing to reasons that

possibly include differences in packaging, labeling, and warranty service. Note that, under this setting, retailers of the gray items also benefit from the authorized dealer’s advertising expenses. Therefore, a free-riding problem exists.

For simplicity, assume that a consumer chooses to buy the legitimate product when she is indifferent between buying the legitimate and gray versions; moreover, she chooses to buy when she is indifferent between buying the legitimate or gray versions and not consuming the product. All potential consumers are risk neutral. The benefit v is assumed to be uniformly distributed between 0 and 1, which means the

demand curve is linear.

The price of the gray item, pi, depends on the law on parallel imports. Suppose

that a gray item will be sold successfully in the market with probability ë, and will be

blocked and destroyed by the government with probability 1 – ë. Intuitively, ë will be

high if the law favors parallel imports, and will be low if the law is hostile toward parallel imports.

Given the above assumptions, the expected profit of a retailer who sells gray items is ë pi– w. Since there is no entry barrier for selling gray items, a gray retailer’s

expected profit should be zero, therefore we have

λ w

pi = . … … (1)

3. The Analysis of the Model

In this section, we will first analyze consumers’ behavior. We will then study the authorized dealer’s pricing and advertising strategies. If a potential consumer with value v buys the legitimate product, she will receive a net benefit of á v – pl. On the

other hand, if she buys the gray item, her net benefit is (1 – â) á v – w/ë. Finally, if the

potential consumer does not consume the product, her net benefit will be 0.

Define αl p v0 ≡ , … … (2) λ α β) 1 ( 1 − ≡ w v , … … (3) α β λ w p v l − ≡ 2 . … … (4)

The following lemma documents a potential consumer’s choice.1

Lemma 1. A consumer with value v will buy the legitimate product if v max {v0, v2},

will buy the gray market product if v1 v < v2, and will not consume any product

otherwise.

From Lemma 1, a gray market exists if and only if v1 < v2, or equivalently,

λ β β (1 ) 1− = − > p w p i l . … … (5)

Intuitively, a gray market exists only when the price for the legitimate item is

relatively high compared with the price of the gray item. It can be verified that v2 > v0

> v1 if (5) holds, and v2 v0 v1 if (5) is violated. Therefore, if (5) holds, consumers

with v v2 buy the legitimate product, consumers with v1 v < v2 buy the gray

product, and consumers with v < v1 do not consume any product. On the other hand, if

(5) is violated, consumers with v v0 buy the legitimate product and the remaining

consumers do not consume.

We now turn to the authorized dealer’s behavior. From the discussions about consumers, the dealer’s pricing and marketing strategies must fit into one of the three cases.

(i) Gray Market. In this case, pl > w / [(1 - â) ë] and a gray market exists.

(ii) Limit Pricing. In this case, pl = w / [(1 - â) ë] and no consumer buys the gray item.

(iii) No Gray Market Threat. In this case, pl < w / [(1 - â) ë] and no consumer buys

the gray item.

Since the optimal pl and á in different cases are different, these three cases will

be analyzed separately.

3.1 Gr ay Mar ket

First consider the gray market case. The authorized dealer chooses pl and á to

maximize ) ( ) 1 ( ) (pl w v2 K α G ≡ − − − Π , … … (6) subject to (5) and the non-negative constraints for pl and á. The first-order conditions

are 0 ) 1 ( 2 = − − − = ∂ Π ∂ α β w p v p l l G _{. … … (7)} 0 ) ( ' ) ( ₂ − = − − = ∂ Π ∂ _α α β λ α K w p w p_l l G _{. … … (8)}

Proposition 1. Suppose a gray market exists in equilibrium.

(a) The advertising, á, is increasing in ë and â, and is decreasing in w.

(b) The price for the legitimate product, pl, is increasing in â. It is decreasing in ë if

and only if ) ( ) ( 1 ) ( " ₃ p w w w K > _l− − λ α β α . … … (9) It is increasing in w if and only if

λ λ λ α β α + − − > 1 1 ) ( ) ( " p ₃ w w K l _{. … … (10)}

Part (a) of Proposition 1 says that the authorized dealer will increase advertising expenses when the law becomes more favorable to parallel imports, when the quality difference between the legitimate and the gray products is greater, and when the marginal cost of a legitimate product is lower. The result that á is increasing in ë is

interesting because it contradicts Reebok Australia's claim that Reebok should reduce advertising and sports sponsorship upon the change in copyright law to allow parallel imports. The intuition of this result can be explained as follows. In this model, advertising not only raises the demand for both the legitimate product and parallel imports, it also enhances the differentiation between the two. Because of the latter effect, the authorized dealer should increase rather than reduce advertising and

promotion when the law becomes more favorable to parallel imports. Also, the higher the â, the more significant the latter effect is, so the authorized dealer has a stronger

incentive to advertise when â increases. By (8), an increase in w has a negative effect

on dΠG/dá, so will lead to a lower á.

From part (b) of Proposition 1, as â increases, the threat from the gray product

becomes less serious, so the authorized dealer will raise pl. It is also shown that pl

may be either increasing or decreasing in ë. An increase in ë has two effects on the

optimal pl. First, when ë increases, the authorized dealer faces more fierce

competition from gray items, so he should reduce pl as a result. On the other hand,

note that pl and á are complements (that is, 0

2 > ∂ ∂ Π ∂ l G p α ), and á is increasing in ë.

Therefore, an increase in ë might indirectly raise pl through increasing á. The sign of

d pl /dë would depend on the tradeoff of these two effects.2 The sign of d pl /dw can

be explained in a similar way.

3.2 Limit Pr icing

2

However, in all examples we construct, d pl /dë > 0, which is consistent with the impression that

Now turn to the case of limit pricing. In this case, the authorized dealer engages in limit pricing. It sets the price pl just low enough to choke off the gray market.

Therefore, we know that (5) will bind. Let pll denote the pl that makes (5) binding. We

have λ β) 1 ( − = w p_ll . … … (11) In this case, v1 = v2 = α λ β) 1 ( − w .

The authorized dealer’s maximization problem can be written as ) ( ) ) 1 ( 1 ( ) ( α λα β α K w w p Max _{lp ll} − − − − ≡ Π . … … (12)

By (12), the authorized dealer only determines á in this case. Note that in this case,

potential consumers with v w/[(1-â) ë á] will buy the legitimate item, while the

remaining consumers will not consume any product. This means there is no gray market in this case. It is easy to show that Ðlp is concave in á, so the optimal á is

determined by the following first-order condition.

0 ) ( ' ) 1 ( ) 1 ) 1 ( 1 ( ₂ 2 = − − − − = ∂ Π ∂ α λ α β λ β α K w lp . … … (13) Condition (13) equates the marginal benefit and marginal cost of advertising. The following proposition reports comparative static results.

Proposition 2. Suppose that the authorized dealer engages in limit pricing. The price

of the legitimate item, pl, is decreasing in ë, and is increasing in w and â. Also, the

advertising, á, is decreasing in ë, and is increasing in w and â.

A striking difference between the cases of a gray market and limit pricing is that, when the authorized dealer engages in limit pricing, an increase in ë leads to a

reduction in advertising. The reason is that, by (13), obviously ( pl/ á) is decreasing

in á. This means the marginal benefit of advertising is reduced as the legal status

become more favorable to parallel imports. Therefore, when ë increases, á will

decrease. On the other hand, ( pl/ á) is increasing in both â and w. This result is

intuitive. When the quality or cost differences between legitimate and gray items are more significant, advertising becomes more effective marginally, so the authorized dealer will increase advertising expenses. The comparative static results about pl

follow from (11) and the fact that pl = pll in the limit pricing case.

In this case, the authorized dealer sets plandá like a textbook monopoly. From the

previous discussion, the marginal consumer will receive benefitpl / á. The authorized

dealer chooses price and advertising to maximize:

) ( ) 1 ( ) ( α α K p w p l l N ≡ − − − Π . … … (14) The first-order conditions imply

2 α + = w pl . … … (15) ) ( ' ) ( ₂ α α K p w p l l− = . … … (16)

The following proposition documents the comparative static results.

Proposition 3. Suppose there is no gray market threat in equilibrium. The price of the

legitimate product, pl, is increasing in w if and only if

3 ) ( " α α p w K > l _{. … … (17)}

The advertising expenses, á, is decreasing in w. Moreover, both pl and á are

independent of ë and â.

3.4 Over all Profit Maximum

The above subsections discuss the three possible equilibrium pricing and advertising strategies that the authorized dealer may use. Now we briefly describe how the dealer chooses among allowing a gray market, limit pricing, or having no gray market. Let

pln and áln denote the pl and á that satisfy (15) and (16). If pln w/[(1 – â) ë] (so that

(5) is violated), then there is no gray market threat in equilibrium and pln and áln are

the optimal pl and á.

On the other hand, suppose that pln > w/[(1 – â) ë] , then the gray item is a real

threat. Let plg and álg denote the pl and á that satisfy (7) and (8). In this case, if plg >

w/[(1 – â) ë] , then there is a gray market in equilibrium, and plg and álg are the optimal

pl and á. Otherwise, the authorized dealer will use the limit pricing strategy.

4. Welfar e Analysis

We define social welfare as the sum of consumer surpluses and seller profits within the country. This measure ignores the welfare of people in other countries. Since there are three possible kinds of optimal strategy for the authorized dealer, we analyze welfare effects in these three cases separately.

When there is a gray market, social welfare is

∫

∫

= = − − + − − ≡ 2 1 2 1 ) ( ] [ ] ) 1 [( v v v v v G dv v w dv K w v W α α λ α β . … … (18)

We will focus on how a change in ë affects welfare. Totally differentiate WG with

respect to ë, we have         + + − + − − + − =

∫

∫

= = 2 1 2 1 2 2 2 ) '( ) [(1 ) ] ( v v v v v G _dv d d v dv w d d v d d K d v d w w v d W d λ α λ λ α β λ α α λ λ α β λ . … … (19) It can be shown that dv2/dë > 0,3 which implies that some consumers shift from

legitimate to gray items as ë increases. Moreover, from Proposition 1 we know that dá/dë > 0. Equation (19) demonstrates the ways in which an increase in ë changes

welfare.

First, when ë increases, some consumers will shift from legitimate to gray items. Such a shift reduces welfare because (i) the gray product gives consumers less benefits than the legitimate product ((1 – â) á v rather than á v), and (ii) the marginal cost of the

gray product is higher than that of the legitimate product (w/ë rather than w). This

effect is reflected in the first item in the RHS (right hand side) of (19).

Second, because á is increasing in ë, the authorized dealer will spend more in

advertising when ë increases. This effect is reflected in the second item in the RHS of

(19).

Third, an increase in ë raises welfare in two ways. (i) When ë increases, á becomes

higher. As a result, consumers of both the legitimate and the gray products enjoy higher benefits. (ii) An increase in ë lowers the marginal cost of the gray product.

These two effects are captured in the third item in the RHS of (19).

Since the first two effects are negative and the last one is positive, dWG/dë may

be either positive or negative. We have worked out some numerical examples, and find that welfare is not always monotonic in ë when there is a gray market in

equilibrium. The following lemma presents some results in special cases.

Lemma 2. Suppose there is a gray market in equilibrium.

(a) When (9) holds, dWG/dë > 0 if the size of the gray market is greater than that of

the market for the legitimate product (that is, v2– v1 > 1 – v2).

(b) When ë = 1, a decrease in ë will reduce welfare if and only if v2– v1 > 0.5 (1 –

v2).

Lemma 2 states that an increase in ë is likely to raise welfare when the size of the gray market is large relative to the market for the legitimate product. This result is intuitive. By (19), when ë increases, consumers who buy gray items not only get more benefits from the product (because á increases), but also pay less for the product

(because w/ë decreases). Therefore, an increase in ë is welfare enhancing if the

fraction of consumers who buy gray items is not small.

4.2 Limit Pr icing

In the limit pricing case, social welfare is

∫

− = − − ≡ 1 ] ) 1 /[( ) ( ] [ αλ β α α w v L v w dv K W . … … (20)

Analyzing the welfare impact of a change in ë, we have

λ α α λ β λ α β λ β αλ β λ _βαλ d d K dv v w w d W d w v L         − + − − − + − − − =

∫

− = 1 ] ) 1 /[( 2 2 2 2 ) ( ' ) 1 ) 1 ( 1 ( ) 1 ( ) 1 ) 1 ( 1 ( ) 1 ( . … … (21) By (21), an increase in ë changes welfare in two ways. First, it lowers the price for the

legitimate product (w/[(1-â)ë] ), which will induce more consumers to buy the

legitimate product. This effect is reflected in the first item in the RHS of (21). Second, by Proposition 2, an increase in ë reduces á in the limit pricing case. A decrease in á

may reduce welfare by (i) inducing some consumers not to consume the product, and (ii) lowering the benefit of the legitimate product to consumers. On the other hand, a decrease in á reduces the authorized dealer’s advertising expenses, which will raise welfare. The welfare effects of ë through affecting á are reflected in the second item in the RHS of (21). Again, as in the gray market case, the welfare effects of ë in the

limit pricing case are ambiguous.

4.3 No Gr ay Mar ket Threat

When the gray product is not a threat to the authorized dealer, social welfare is

∫

= − − ≡ 1 / ) ( ] [ α α α w v N v w dv K W . … … (22) Obviously, by (22), a change in ë will not affect welfare in this case.

4.4 The Optimal Policy

If the government knows the demand function and the cost function of the authorized dealer, it should choose a ë to maximize social welfare. However, from the above

analysis, the welfare effects of ë are quite complicated. This complexity arises

because ë affects not only the price for the gray product, but also the authorized dealer’s pricing and advertising strategies. As a result, the welfare effects can be ambiguous. The contribution of this paper is to point out the ways that the legal status of parallel imports may change welfare. It suggests that, when the government designs policies related to parallel imports, it should properly evaluate all the benefits and costs of the policies.

5. An Example

In this section, we develop a simple example to show how ë affects the authorized

dealer’s strategies and social welfare. For simplicity, we assume that á is a constant

and á > 1. Although this assumption is quite strong, it greatly simplifies analysis.

Such an example provides the foundation for analyzing a more general model where á is not a constant. For this example, we get the following results.

Proposition 4. Suppose that á is a constant and á 1. Denote ë1 = ) )( 1 ( 2 α β + − w w . … … (23) ë2 = ) )( 1 ( ) 1 ( αβ β β + − + w w . … … (24) (a) If ë > ë2, then there is a gray market in equilibrium. In this case, the optimal pl is

2 /λ

βα +w+w

, and the authorized dealer’s profit is ( ) 4 ) / ( 2 _α αβ αβ λ K w w − + ₋ . (b) Ifë1 ë ë2, the authorized dealer will engage in limit pricing. In this case, the

optimal pl is β λ − 1 / w

, and the authorized dealer’s profit is

) ( ) 1 ( )] 1 ( / [ ] / ) 1 ( [ 2 α β α β α λ λ β α K w w ₋ − − − − − .

(c) If ë < ë1, then the gray product is not a threat to the authorized dealer. In this case,

the optimal pl is

2

w + α

, and the authorized dealer’s profit is ( ) 4 ) ( 2 α α α K w ₋ − . (d) Social welfare is independent of ë if ë < ë1, is increasing in ë if ë1 ë ë2, and is

convex in ë ifë > ë2. This implies that the optimal ë is either ë1 or ë2.

Proposition 4 demonstrates that there is a gray market if ë is large, the authorized

dealer will engage in limit pricing ifë is in the medium range, and the gray product is

not a threat ifë is small. This result is quite intuitive and general.4 Proposition 4 also

4 _{This result holds in all other examples we construct. In these examples,}

sheds light on the optimal policy on parallel imports. When á is a constant, a change

in ë does not affect the authorized dealer’s advertising expenses. Therefore, by (21), social welfare is increasing in ë in the limit pricing case. This result implies that the

government should never set ë so low that the gray product is not a threat to the

authorized dealer. The result that welfare is convex in ë when there is a gray market

also has an interesting policy implication, which is documented in the following corollary.

Corollar y. Suppose á is a constant and á 1. Moreover, dWG/dë > 0 when ë = 1. In

this case, there exists a λ∈(0,1) such that the optimal ë = 1 if the government

cannot choose any ë smaller than λ .

The above corollary means that, if dWG/dë > 0 when ë = 1 and the government

cannot enforce strict restrictions on parallel imports (that is, all the feasible ës are greater than λ ), then it should not regulate parallel imports at all. For many countries, law enforcement is poor. The corollary suggests that these countries should not

impose any restriction on gray products.

6. Concluding Remar ks

In this paper, we develop a model to study issues related to parallel imports. We analyze the pricing and advertising strategies of authorized dealers who sell legitimate products. We also investigate welfare effects of the regulation on parallel imports. We find that the authorized dealers may increase rather than decrease their advertising expenses when the law becomes more favorable to parallel imports. In addition, we find that the welfare effects of parallel import regulation may be ambiguous. We identify several conditions under which loosening restrictions on parallel imports will increase social welfare.

Two assumptions in our model deserve discussions. First, we implicitly assume that the costs of supplying the legitimate item and the parallel import are identical. Realistically, the costs may differ. For instance, the authorized dealer may incur costs to provide warranty service for the legitimate product. On the other hand, parallel importers may be too small to benefit from economies of scale. These cost differences will affect the welfare analysis in an obvious way.

Second, we assume that the seller sets the same price to all potential consumers. However, in the real world authorized dealers can effect price discrimination. An obvious response to the gray market is for the authorized dealer to target a lower price at the individuals who are more likely to buy the gray version, that is, those who receive relatively low benefit from the product. Price discrimination will allow

authorized dealers to increase their profits and to reduce the threat of gray items. Taking the possibility of price discrimination into account will be a promising way to extend our model.

二、參考文獻

Ahmadi, Reza, and B. Rachel Yang, “Parallel Imports: Challenges from Unauthorized Distribution Channels,” Marketing Science 19 (Summer 2000), 279-294.

Band, Jonathan, “The Digital Millenium Copyright Act”, Memorandum, Morrision & Foerster LLP, Washington, D.C., October 1998.

(http://www.arl.org/info/frn/copy/band.html).

Duhan, Dale, and Mary Jane Sheffet, “Gray markets and the legal status of parallel importation”, Journal of Marketing 52 (1988), 75-83.

Gallini, Nancy T., and Aidan Hollis, “A Contractual Approach to the Gray Market”, International Review of Law and Economics 19 (March 1999), 1-21.

Gopal, Ram D., and G. Lawrence Sanders, “International Software Piracy: Analysis of Key Issues and Impacts”, Information Systems Research 9 (December 1998), 380-397.

Ghosh, Shubha, “An Economic Analysis of the Common Control Exception to Gray Market Exclusion”, University of Pennsylvania Journal of International Business Law 15 (1994), 373- 390.

Grossman, Gene M., and Carl Shapiro, “Counterfeit-Product Trade”, American Economic Review.

Grossman, Gene M., and Carl Shapiro, “Foreign Counterfeiting of Status Goods”, Quarterly Journal of Economics (1988b), 79--100.

Iyer, Ganesh, and V. Padmanabhan, “Dynamic Uncertainty and Supply Chain Contracting”,

Working Paper, Olin School, Washington University, St Louis, 2001.

Kek, Gina P.L., and Jong-Say Yong, “Assessing the Welfare Effects of Parallel Importing: A Vertical Differentiation Model with Investment”, Working Paper, Department of Economics, National University of Singapore, August 1999. Lee, Khai S., Guan H. Lim, and Soo J. Tan, “Strategic Management of Grey Marketing”,

Journal of Strategic Marketing 8 (March 2000), 67-88.

International Price Discrimination”, Journal of International Economics 37 (November 1994), 41-57.

Purohit, Devavrat, “Marketing Channels and the Durable Goods Monopolist: Renting versus Selling Reconsidered”, Journal of Economics and

Management-Strategy 4 (Spring 1995), 69-84.

Tan, Soo J., Guan H. Lim, and Khai S. Lee, “Strategic responses to parallel importing”

Papadopoulos, Theo, “Copyright, Parallel Imports and National Welfare:

The Australian Market for Sound Recordings”, Australian Economic Review 33 (December 2000),

Tom, Gail, Barbara Garibaldi, Yvette Zeng, and Julie Pilcher, “Consumer demand for counterfeit goods”, Psychology & Marketing 15 (August 1998), 405-421.

三、記畫成果自評

This project has achieved the goals it plans to accomplish. In this project, a theoretical model is built to investigate issues related to parallel imports. The model analyzes the authorized dealers’ behavior when they face the threat of parallel imports. It also discusses the welfare effects of loosening restrictions on parallel imports. This project enhances our understanding on parallel imports. It also has interesting policy