Forthcoming in Journal of abor Research DOES INTERNATIONA OUTSOURCING REAY OWER WORKERS INCOE? Jan König and Erkki Koskela December 00 Abstract: We analyze the impact of international outsourcing on income, if the domestic labor market is imperfect, i.e. there is a bilateral bargaining beteen a firm and a labor union. In our analysis e distinguish beteen the cases here the parties negotiate over the age only and here they negotiate over both age and profit sharing. We find in the first case that outsourcing has an ambiguous effect on the orkers income, hile it increases the orkers income in the second case. For the optimal amount of international outsourcing, e find that, depending on the age effect of outsourcing, in a pure age bargaining system it can be higher or loer than the level here domestic and foreign marginal labor costs are the same. In contrast, in a age and profit share bargaining system, the amount of outsourcing lies belo this level. JE classification: E3, E4, J3, J33, J8 Keyords: strategic outsourcing, profit sharing, labor market imperfection School of Business & Economics. Freie Universität Berlin Boltzmannstr. 0 D- 495 Berlin GERANY e-mail: jan.koenig@fu-berlin.de Erkki Koskela Department of Economics University of Helsinki P.O. Box 7 (Arkadiankatu 7) FI-0004 e-mail: erkki.koskela@helsinki.fi We are grateful for useful comments by an anonymous referee and the participants in ISNE 00 to improve the paper. Koskela thanks the Academy of Finland (grant No. 76) for financial support and Freie Universität Berlin for good hospitality. König thanks the University of Helsinki for good hospitality.
I. Introduction Over the last years, as a result of the groing globalization, international outsourcing, hich is defined as the acquisition of production parts from an independent foreign supplier, has become an important managerial tool in reorganizing a firm s production process. Attended ith this fact, many people fear the ide consequences for the domestic labor market, especially for ordinary orkers. Due to the possibility of substitution, such consequences may be the loss of employment or a age reduction. In this situation, the labor market structure and the existence of a labor union ith the poer to avoid a age decrease and/or to bargain ith the firm over employment guarantees, play an important role. This paper presents a theoretical frameork to analyze the effects of committed international outsourcing on orkers income, if orkers are represented by a labor union. 3 Thus, e assume an imperfect domestic labor market, i.e. a firm and a labor union negotiate over orkers remuneration, hile e distinguish beteen to kinds of negotiation. In the first case, e follo the classical bargaining approach here the age alone is determined, hile in the second case e assume an alternative approach here the firm and the labor union bargain over both the age and a profit share. Due to the actuality and importance of this topic, there is a groing amount of literature relating to the effects of outsourcing or globalization on ages. From a theoretical point of vie, Danthine and Hunt (994) sho that the globalization intensifies the product market competition. As a consequence, loer profits occur, hich results in age moderation in unionized sectors. A similar finding is presented by Glass and Saggi (00). Opposed to that, Naylor (998, 999) finds that domestic unionized orkers may benefit from globalization in terms of higher ages and employment, since total production expands if ne markets can be served by the firm. ommerud et al. (009) sho that higher market integration favours outsourcing to lo cost countries and increases the domestic age due to less elastic labor demand. The reason is that the used inputs are complements and thus, for a given amount of outsourcing, the resulting loss of the labor union due to a higher age decreases. Hoever, there are also studies, such as Skaksen and Sorensen (00) or Koskela and 3 Empirical studies like Hummels et al. (998, 00) or Yeats (00) sho the increase of imported intermediate goods over the last 30 years. For an overvie concerning the debate on employment effects due to outsourcing see Freeman (995) and Bhagati et al. (004). In the committed case, outsourcing takes place before age bargaining. Thus, the external procurement is seen as a long-term contract or investment that fixes the amount of outsourcing. An overvie about the relationship beteen outsourcing and age bargaining is presented by Perry (997).
Stenbacka (009), hich sho that the domestic age effect of foreign direct investments or outsourcing is a priori ambiguous. In Skaksen and Sorensen (00), the degree of substitution beteen the activities in the home country and abroad is decisive for the age effect. If the activities are good substitutes, a loer age results and the domestic orkforce loses, hile a higher age results from complementary activities and thus, the domestic orkforce gains. In Koskela and Stenbacka (009), the age effect of outsourcing depends on the labor union s relative bargaining poer, here it loers (increases) the age if the labor union is sufficiently strong (eak). Empirical studies also analyze the age effect of international outsourcing. In their study, Feenstra and Hanson (999) sho the age reducing effect for lo-skilled orkers in the United States over the period 979-990. Senses (00), also using U.S. data, provides empirical evidence of an increasing age elasticity and thus for a age moderating effect of outsourcing. 4 Focusing on German data, Geishecker and Görg (008) identify inners and losers from international outsourcing depending on the skills of the orkers. Although the German labor market is characterized by relatively rigid ages, there may be a age-moderating effect of outsourcing, if it improves the outside option of the firm. The authors find that this is true for lo-skilled orkers, ho receive a loer age ith higher outsourcing, hile the high-skilled age increases. The mentioned theoretical studies focus on pure age effects of outsourcing, by assuming that only the age is determined by bargaining beteen the firm and the labor union. Hoever, additional components such as bonus payments or profit sharing can be the result of such bargaining as ell. 5 The idea behind the incorporation of profit sharing in a compensation scheme is to stimulate the orkers motivation and identification ith the firm and thus to increase their productivity. 6 In this paper, e extend the literature by implementing profit sharing as a part of the compensation scheme. 7 The distinction beteen the case in hich the union and the firm negotiate over the age only, and the case in hich both the age and a profit share are determined simultaneously, allos to discern beteen a age and income effect. Thus, 4 5 6 7 Similar findings are shon in earlier studies by Slaughter (00) and Hasan et al. (007). Empirical studies as Pendleton et al. (00) sho that profit sharing is an often used compensation scheme in many OECD countries. For further evidence regarding the incidence of profit sharing see also Estrin et al. (997) and Conyon and Freeman (004). Hoever, empirical studies sho that the productivity effect is ambiguous. For an increasing effect on productivity see Cable and Fitzroy (980), hile Jones and Pliskin (99) and Kruse (99) demonstrate a negative productivity effect of profit sharing. There are some studies that analyze the implementation of profit sharing in collective bargaining, e.g. Holmlund (990) and Jerger and ichaelis (999). Concerning the efficiency property, Pohjola (987) and Anderson and Devereux (989) sho that also ithout an employment determination the outcome of a collective bargaining is efficient by introducing bargaining over ages and profit sharing. Hoever, all studies abstract from outsourcing. 3
our central research question is: Is the fear of income loss for unionized orkers justified? In our analysis e find that in the case in hich the firm and the labor union bargain over the age only, outsourcing has an ambiguous effect on age and thus on the orkers income. In contrast, if the labor union and the firm bargain simultaneously over the age and a profit share, outsourcing ill increase the orkers income, if the marginal costs of outsourcing are loer than the domestic outside option. Knoing the income and age effects, based on comparative statics, e analyze the optimal amount of outsourcing under the different remuneration schemes. Here, e find that depending on the age effect, the outsourcing demand under a pure age bargaining system can become higher or loer than the outsourcing level here domestic and foreign marginal labor costs are the same, hile under a simultaneous age and profit share bargaining system the amount of outsourcing is loer than the level here domestic and foreign marginal labor costs are the same. We proceed as follos. Section II presents the basic frameork. Section III investigates the model in terms of labor demand, the structure of the bargaining process and the optimal amount of strategic outsourcing. A conclusion and a brief discussion of extensions are presented in Section IV. II. Basic Frameork We assume that in our economy a representative firm produces the final good by using to input goods. The combination of these inputs for producing the final good is represented by the Cobb-Douglas production function F = X Y ith 0 < <, () here X and Y characterize the amount of to inputs. Furthermore, e assume that the X -input production must take place in-house, hereas the Y -input can be produced in-house or can be outsourced. For simplicity, e model a linear technology in every input production, here for one unit of the input good, one unit of labor respectively outsourced input is needed. Thus, the production functions for the input goods are X X = Y = Y, 4
X Y here or presents employment in the specific input production and the used amount of outsourcing. We further assume that labor in both input productions is homogenous and that the X Y overall orkforce = is represented by a labor union. This assumption ensures that no age discrimination beteen the input productions can be realized by the firm. The structure of actions can be interpreted as sequential decisions on three stages. Due to the investment of the external procurement, on the first stage, the representative firm commits to the amount of outsourcing. After the firm has decided about outsourcing, the firm and the labor union bargain over i) the age level or ii) the base age and profit sharing. Since the firm has the right-to-manage, it determines employment according to its labor demand after knoing the bargaining results. The timing sequence of the decisions is summarized in Figure. Figure : time sequence of decisions stage stage stage 3 outsourcing bargaining process labor demand (age or age and profit sharing τ ) The decisions at each stage are analyzed by using backard induction. It should be pointed out that our assumption of committed outsourcing implies that the firm undertakes irreversible investments and that the outsourced inputs are specifically designed for the final good producer. Of course, if the outsourced inputs are standard components, this assumption can be reversed by assuming that the outsourcing decision takes place after the domestic age determination. 8 According to Figure e also neglect the possibility of ex post renegotiations of the outsourcing contract. One may assume that the special requirements can be easily stipulated in an ex ante contract. Hoever, problems often arise ith regard to the verification of the agreement by a third party and if the economic environment is too complex and unpredictable. In that case, it can be impossible to design a contract that accounts for all outcomes and thus the contract becomes incomplete, hich creates the so-called hold-up -problem. 9 If not 8 9 See Skaksen (004) and Braun and Scheffel (007) for this strand of the literature, i.e. the case of flexible outsourcing. A survey concerning outsourcing and incomplete contracts is presented by Spencer (005). The hold-up -problem describes the opportunistic behavior by the input producer, if special investments are needed. 5
all properties can be clearly verified ex ante, the final producer evaluates the match of the external procured inputs ith the requirements after the outsourcing decision has taken place and additional costs for monitoring or quality control arise. In case of a mismatch beteen the needs and the effective quality, negotiations concerning the price beteen the outsourcing partners may be resumed. For our time sequences this means that, after domestic age formation, additional negotiations beteen the parties or additional monitoring costs of outsourcing may be introduced. Hoever, for keeping the analysis simple, e neglect the hold-up -problem in the relationship beteen the parties. Therefore, e abstract from costs for supervising the quality of the intermediate good or ex post renegotiations. III. Solving the odel In the next parts e solve the model according to the presented timing structure. We first derive at the third stage the labor demand in both input productions for given outsourcing. After that, at the second stage, e focus on the bargaining process by distinguishing the to mentioned approaches. While in the first approach the firm and the labor union bargain over the age only, in the second approach both parties negotiate the age and a profit share. Finally, at the first stage, e solve for the optimal amount of strategic outsourcing. III.. 3 rd stage: Domestic abor Demand The firm decides on domestic labor to maximize the profit function X Y ; X Y X Y ( ) ( ) ( ) f ( ) max π =, () taking, the amount of outsourcing, as given. For the cost of outsourcing, f ( ), e assume that there are additional costs associated ith outsourcing other than the price of the intermediate goods. Such costs could be costs for transport, hich exponentially increase ith higher outsourcing. To allo for an exponential cost increase, e model a quadratic cost function, i.e. f ( ) = c, ith c > 0, f '( ) > 0 and f ''( ) > 0. 6
Solving problem () leads to the standard result that employment is set here marginal productivity equals the age rate. From the first-order conditions e obtain as the labor demand for given outsourcing in the different input productions 0 X =, (3a) Y =. (3b) Thus, the overall domestic labor demand is = X Y = ( ). (4) Equation (4) shos that domestic labor demand is a negative function of both the age and the amount of outsourcing, here the substitutability of lo-skilled labor and international outsourcing is consistent ith empirical evidence, e.g. presented by Görg and Hanley (005). The labor demand reaction to age changes can be expressed by the age elasticity of labor demand, hich can be ritten as η = = >. (5) According to equation (5), the age elasticity depends on the age level and the amount of outsourcing. These effects can be determined by the first derivatives η η = > 0 and η η = > 0. Therefore, ith a higher domestic age or higher outsourcing, domestic labor demand becomes more elastic. In the absence of outsourcing, the age elasticity η = is constant and smaller than in the = 0 presence of outsourcing, hich is in line ith empirical evidence as shon by Senses (00). 0 Notice, that also in the presence of a bargained profit share, here the profit of the firm is ( τ ) π, e obtain the same labor demand reactions, since profit sharing orks as a profit tax. Due to the neutrality of this kind of tax, also in the case of a bargained profit share the domestic labor demand does not depend on profit sharing. For notational convenience e use in the next calculations the subscript as a characterization of the first derivative, i.e. η = η /. 7
III.. nd stage: Bargaining Process At this stage, the firm and a labor union bargain over i) the age level or ii) the age and profit sharing. We distinguish beteen these scenarios since both are possible in observed bargaining rounds. Hoever, in our frameork the determination of the bargaining regime is exogenous, hich is driven by the fact that sometimes the bargained variables are given by politics or la. The outcome of the bargaining process is assumed to be determined by the Nash- Bargaining-Solution, here the Nash-Product is defined as ( ) ( π π ) Ω = U U 0 0. In the above notation U 0 and π 0 are the disagreement payoffs for the union and the firm. In case of disagreement there is no production, implying that every union member gets the exogenous outside option, i.e. U 0 = N b, here b captures the available minimum income for the labor union members N. On the other side, the firm π = f, hich means that the firm has an loses its investment in outsourcing, i.e. ( ) incentive to reach an agreement. 0 III.. Bargaining over Wages only Assuming that only the age ill be determined, e can rite the bargaining problem as max Ω = ( ) U U ( π π ) 0 0. To describe the preferences of the labor union, e model a utilitarian union utility function U = u( ) u( b) ( N ) in case of an agreement, here the individual utility u () is linear in income, i.e. u ( ) = and u ( b) = b. Combining this ith the union s outside option U 0, e can express the union rent as U = U U0 = ( b). The bargaining rent of the firm, π = π π 0, can also be expressed explicitly. Since the profit in case of an agreement is π = F f ( ) and the disagreement profit is π = f, e obtain a rent of π = F. 0 ( ) While in most European countries as Germany or Finland the age is the central determinant in a bargaining beteen the union and the firm, in France there exists an obligatory profit share system for firms ith more than 50 orkers. Hoever, in the bargaining round the firm and the labor union determine the details such as the calculation formula or the duration. oreover, in Section IV e briefly discuss the endogenous choice of the regime by the bargaining parties. 8
Ω aximizing the Nash-Product concerning the age, the first-order condition is U π = 0 = ( ), here U π and U U = π = π = η, (6a) b b ( ) /. (6b) Using these expressions as ell as the age elasticity of labor demand, as the solution of the first-order condition e obtain (, ) b = A,, (7) hich is the standard result that the age consists of the outside option and a mark-up. η [ η ( ) ( )] ( )( ) In our model the mark-up A = > ( η ) [ η ( ) ( )] ( ) ( ) depends on the relative bargaining poer of the labor union, the amount of outsourcing and the age. Therefore, equation (7) is an implicit formulation. Knoing the negotiated age e can distinguish the extreme cases of a monopoly labor union, hich sets the age unilaterally, and the absence of a labor union, here the firm sets the age independently. The case of a monopoly labor union is characterized by =, here the age becomes η = ( ) b = η, hile in the absence of a labor union, i.e. = 0, the age is = b. 3 =0 To anser our research question and thus to characterize the income effect of outsourcing, e no turn to a detailed analysis. After the implicit differentiation of (7) ith respect to outsourcing and substituting b = / A, e can characterize the impact of international outsourcing on the bargained age as d d A = A A A, (8) 3 Since η >, it is obvious that the relative bargaining poer of the labor union ill have a positive effect on the mark-up in the general case of 0 < <, i.e. A > 0. 9
A here > 0 (see Appendix A). A The outsourcing effect on the mark-up, A / A, is a priori ambiguous and depends on the relationship beteen the relative bargaining poer of the labor union and the outsourcing-labor ratio. 4 For the impact of outsourcing on the mark-up e find > A = 0. (9) < Thus, outsourcing has a priori an ambiguous effect on the domestic age. For a better understanding e can also identify the age effect of outsourcing in the extreme cases. In the case of unilateral age setting by the firm, the age ill be at the loest possible level, hich is the constant and exogenous alternative income. Thus, d outsourcing has no age effect, i.e. = 0. On the other hand, in the case of a d = 0 monopoly labor union e obtain < d A 0, hich yields < 0. = d = Concerning the more general case, in hich both parties are endoed ith a positive bargaining poer, i.e. 0 < <, e can summarize as Proposition : If the firm and the labor union bargain over the age only, outsourcing has an ambiguous effect on the orkers income. To explain this ambiguous effect e can identify to opposite mechanisms. First, ith higher outsourcing the labor demand (5) becomes more elastic. Due to a more elastic labor demand, a higher age increases the union s utility loss of less employment. Consequently, this mechanism induces age restraint and makes the labor union less aggressive, hich results in a loer age mark-up. Second, as outsourcing and domestic labor are substitutes, ith higher outsourcing the firm s profit is less affected by domestic labor costs. Thus, outsourcing moderates the profit-reducing effect 4 A similar result is obtained by Koskela and Stenbacka (009). Hoever, they use a model here only our Y -production characterizes the production technology and focus on the unemployment effects of outsourcing in a general equilibrium model, hile e concentrate on the comparison of different bargaining regimes concerning the income effects of outsourcing in a partial analysis. Thus, e sho hether this unclear result in the classical approach also holds for a more realistic description of the production technology ith more than one production chain. Additionally, e can anser if this unclear result depends on the bargaining regime. For that reason, the detailed presentation of a knon result is used for a better understanding and a complete analysis. 0
π /π of a age increase, hich promotes a higher age mark-up. In line ith this explanation, one may also argue that higher outsourcing increases the firms costs in case of disagreement and thus the firm has a stronger incentive to reach an agreement. Since no the firm faces a eaker bargaining position, the labor union is able to achieve a higher age mark-up. As one can see from the mark-up equation above, the interplay of the relative bargaining poer, the production technology and the reaction of the labor demand elasticity determine, hich of the to opposing effects dominate. The effect of outsourcing on the domestic age is also analyzed in ommerud et al. (009). They find an unambiguous age increasing effect of outsourcing, since the labor demand becomes less elastic, hich is driven by the assumption of complementary inputs. Additionally, they see the fixed costs of outsourcing as sunk costs and thus there ill be no negative impact for the firm in the bargaining. Hoever, in our model e follo a different assumption concerning the production technology and the investment costs. Thus, the different assumptions beteen the to analyses, yielding different effects on the labor demand elasticity and the firm s outside option, explain the possible bargaining outcome differences. As mentioned in the introduction, the orkers also fear the substitution of domestic jobs by outsourcing and thus the loss of employment. Knoing the age effect, the overall impact of outsourcing on domestic labor demand can be derived. Using equation (4) e can determine employment effect as d d d =. (0) d Equation (0) shos the to orking channels of outsourcing on the domestic labor demand. The first one is the substitution of employment in the Y -sector and the second one is the age effect due to the reaction of domestic labor costs. While the substitution effect decreases labor demand, the a priori unclear age effect can reinforce or offset this effect or does not affect the labor demand. Thus, e can determine an unambiguous employment effect if outsourcing has a age increasing effect only, since in that case both the substitution effect and age effect lead to less employment. Therefore, in this case outsourcing increases the domestic income, but feer employees ill receive a higher age. If outsourcing does not affect the negotiated age level, the labor demand reducing substitution effect is still orking and feer orkers receive the same income. Also in the case of a age moderating effect there is the negative substitution effect. Hoever, no the positive employment effect,
hich results from the loer labor costs, can offset the substitution effect. Thus, there could be higher or loer orkforce ith loer income due to higher outsourcing. III.. Simultaneous Bargaining over Wage and Profit Sharing As mentioned in the introduction, there are several studies concerning the simultaneous negotiation about the age and profit sharing. Hoever, these studies abstract from strategic outsourcing. Before e formally analyze this bargaining regime in the presence of outsourcing, e have to modify the objective functions of the labor union and the firm. Since e assume that the individual utility is linear in income, the overall remuneration of an π employed orker can be ritten as ω = τ, here τ characterizes the share of profit, hich is distributed to the orkforce. 5 On the other hand, the income of an unemployed orker is still characterized by the exogenous minimum income b. Folloing the assumption of a utilitarian union utility, e can rite the union rent in a bargaining regime ith profit sharing as U = τ π. Of course, the profit of the firm s oner ill change, too. In case of agreement he no earns ( τ ) π. Since the value of disagreement is the same as in section III.., the rent is no represented by π = τ π f. ( ) ( ) Due to the simultaneous negotiation of the age and the profit share, in this setting, the Nash-Product has to be maximized concerning both parameters, i.e. max Ω = U π. As the first-order conditions, e obtain, τ U π Ω = 0 = ( ) and (a) U π Ω τ Uτ = 0 = U πτ ( ) π. (b) Using U τ = π and π τ = π, (b) can be reritten to =. Inserting this U π expression in (a) yields 0 = U π, here U = ( b) ( τ ) and ( ) π = τ. Implementing these results, e obtain a negotiated base age of = b, () 5 The idea behind this is that the orker are assumed as a team, here the hole team gets the profit share τ π, hich is then distributed equally among the members.
so that the age is equal to the exogenous outside option, hich corresponds to the ell knon results of Weitzman (987). Comparing the negotiated ages (7) and () shos that the age stipulated in a simultaneous age and profit share bargaining is smaller than (equal to) the age in the case ithout any profit share negotiations, if the labor union has a positive (zero) relative bargaining poer. The intuition is relatively simple, since the parties actually bargain over the distribution of the rent realized by the production. If there is no labor union, the hole rent ill be earned by the firm. Since the rent is influenced by employment and the highest rent is realized ith the highest employment level, the firm reduces the age to the loest possible level, hich is the outside option b. If there is bargaining, due to a higher age, the rent decreases. Hoever, no the union realizes a part of it. In the presence of simultaneous age and a profit share bargaining, a similar mechanism leads to the derived result (). Both parties maximize the rent and fix the age on the loest level, i.e. the outside option, hile the distribution of the created rent beteen the parties ill be determined by the negotiated profit share level. Inserting () in the reritten first-order condition concerning the profit share, =, and using the labor demand (3a) and (3b), e obtain for the bargained profit U π share τ = b b ( ) b ( ) b f ( ). (3) From (3) e deduce that in the absence of outsourcing the profit share corresponds to the relative bargaining poer of the labor union, i.e. τ = 6, hile =0 in the presence of outsourcing the negotiated profit share is higher than the relative bargaining poer of the labor union, i.e. τ >. >0 As mentioned above, the profit share determines ho the created rent is distributed beteen the to parties. Thus, one ould expect that the share of the rent for every party equals its relative bargaining poer. Hoever, as shon in equation (3), this does not hold in our frameork. The economic intuition for this result is the folloing: Since the amount of outsourcing is determined before bargaining takes place, the firm has an incentive to reach an agreement and to avoid the negative profit, being the costs associated ith the outsourcing commitment, in case of a disagreement. 6 For this standard result see also Holmlund (990). 3
Therefore, the firm faces a eaker position than in the case of an outside option ith zero profits, here only the relative bargaining poer is decisive for the distribution. In hat follos, the firm receives a loer share of the rent than its relative bargaining poer predicts. Since in the former analysis the age equals the income, the age effect and the income effect of outsourcing are the same. Hoever, in the case of a simultaneous bargaining over the age and profit sharing e have to income components. Thus, in contrast to the former analysis, e no discern beteen a age and an income effect. As equation () shos, the age is the constant exogenous outside option and not affected by outsourcing so that, in the alternative bargaining approach, there is no age effect of outsourcing. Hoever, outsourcing affects the profit in the absence of an agreement. This provides an incentive for the firm to reach an agreement and affect the orkers profit income via the negotiated profit share. To determine this effect, e have to sho the effect of outsourcing on the negotiated profit share. Here e find that (see Appendix B) c V b τ = ( V b f ( )) > here V = b ( ) > 0 0, (4), so that the bargained profit share depends positively on the amount of outsourcing. Hoever, our research question focuses on the income effect of outsourcing, hich corresponds to the impact on the orkers profit income. Under this type of π compensation scheme, the income of an employed individual is ω = b τ, here the income effect of outsourcing can be formalized by ω τ π = τ π π, (5) τ here < 0 and > 0. To determine the outsourcing effect on profit e need the indirect profit function π. Using the derived results, e obtain π f ( ) ( ) b b f ( ) π = and thus = b As this formulation shos, under the assumption that the marginal costs of outsourcing,. 4
( ) f π, are loer than the domestic marginal costs of labor, b, e have > 0 and thus an unambiguous income effect of outsourcing. We can summarize our finding as Proposition : If the labor union and the firm bargain simultaneously over the age and the profit share, outsourcing increases the orkers income, if the marginal costs of outsourcing are loer than the domestic outside option. As one can see from (5) outsourcing affects orkers income in to ays. The first part shos the share-increasing effect, since every orker gets a higher share of the per capita profit. This effect results from the fact that higher outsourcing increases the loss of the firm if there is no agreement, hich makes the firm less aggressive and increases the negotiated profit share. The second mechanism is shon by the expression in brackets. On the one side, due to the substitution of domestic labor by outsourcing in the Y -activity, higher outsourcing increases the profit, if the marginal costs of outsourcing are loer than the domestic marginal costs of labor. On the other side, due to the decreased employment the per capita profit increases. Since the profit share, the overall and the per capita profit ill be positively affected by the external procurement, an employed orker ill unambiguously benefit from higher outsourcing. Similar to the above section e also determine the employment effect. Since the age equals the constant outside option, in the case of a simultaneous bargaining of a age and a profit share only the negative substitution effect affects the domestic d employment level, i.e. = < 0. Thus, in that case e find loer employment ith d higher income. III.3 st stage: Optimal Strategic Outsourcing So far e have restricted ourselves to a short-run analysis, here the amount of outsourcing is given, since the firm has committed itself. We no relax this point of vie by exploring the initial stage of the outsourcing decision and focusing on a longrun perspective, here the firm determines its investments into outsourced production. III.3. Optimal Outsourcing if Parties Bargain over Wages only Assuming a rational firm, there is a perfect forecast and thus, on this stage, the firm maximizes its profit subject to domestic labor demand (4) and age formation (7). Under the domestic labor demand, the indirect profit function is described by 5
( ) f ( ) π =. Thus, the firm s optimizing problem is characterized by max π s.t. = A(, ) b,. (6) Differentiating the indirect profit function yields the first-order condition π = c d d Y = 0. (7) As equation (7) points out, the level of outsourcing depends on the employment Y used in activity Y and on the age effect of outsourcing. As one can see under > 0 and d / d < 0 it follos that c >, hich means that the amount of outsourcing lies above the outsourcing level here domestic and foreign marginal costs are equal. Y Thus, e can conclude that under > 0 and in the presence of a age moderation effect of outsourcing, the firm can reap an additional benefit ith higher outsourcing. 7 Y In contrast, in the case of a age increasing effect and > 0, no additional benefit can be realized by the external procurement and the amount of outsourcing lies belo the outsourcing level here domestic and foreign marginal costs are equal. We summarize our finding as Proposition 3: If the parties bargain over the age only, strategic outsourcing ill be higher (loer) than the level here domestic and foreign marginal labor costs are the same, if outsourcing decreases (increases) the domestic age. This result can be explained as follos. On the one hand, higher outsourcing increases total production costs, but on the other hand, it may also lead to a age moderating effect and becomes a strategic instrument for the firm as it reduces the age bill. As equation (7) shos, the optimal amount of outsourcing is given hen both effects are equalized. In contrast, if outsourcing leads to a age increase, it increases total production costs only. Thus, the amount of outsourcing is loer than the level here domestic and foreign marginal labor costs are the same. Hoever, if there is no Y employment in activity Y, = 0, or no age moderating effect, d / d = 0, e 7 For a graphical argumentation see Koskela and Schöb (00). 6
obtain the ell knon result that the firm chooses an amount of outsourcing here the marginal costs are the same. III.3. Optimal Outsourcing if Parties Bargain over Wage and Profit Sharing Since in this scenario the age is set to the constant outside option, the indirect profit is given by π = ( ) b b f ( ). Thus, the firm s problem is characterized by 8 max ( τ ) π s.t. τ = Φ. (8) Taking into account the formerly derived results, the first-order condition is [( τ ) π ] τ = ( τ ) ( b c ) π = 0. (9) τ Since π > 0 and 0 < τ <, e see from (9) that b > c. Under the assumption of a age moderating effect of outsourcing, this lies in contrast to the case here the parties bargain over the age only. So the firm no chooses an amount of outsourcing loer than the level at hich the marginal cost of outsourcing equals the domestic marginal cost. Comparing the optimal amount of outsourcing under the different bargaining approaches and the reasonable assumption that b <, e can conclude from the conditions b > c respectively c >, that the bargained profit share approach leads to a loer investment in the outsourced production. From (9) e can also determine the impact of labor market imperfections. As can be seen from (3) and (4), the labor union s bargaining poer affects the optimal profit share and the impact of outsourcing on the profit share. For a given amount of outsourcing a stronger labor union reaps a higher share of the profit, hich is shon in equation (3). Since outsourcing increases the profit, this provides an incentive to reduce outsourcing. This effect ill be reinforced, if the firm takes into account the impact of outsourcing on the profit share, because ith loer outsourcing the firm faces a stronger bargaining position, hich increases the firm oner s earned profit. Thus, both effects have the same direction and the firm oner faces a higher incentive to 8 According to (3), the profit share mark-up is π b Φ = π b f ( ). 7
reduce its outsourcing activities in the presence of a strong labor union in order to reap a higher share of the profit. 9 Proposition 4: If the parties bargain over the age and profit sharing, strategic outsourcing ill be loer than the level here domestic and foreign marginal labor costs are the same and the amount of outsourcing decreases ith the labor union s bargaining poer. Our analysis shos that the bargaining structure can be crucial for outsourcing demand, as the different regimes induce different effects on the firm s cost parameters. In the case of a bargained profit share, the age is the exogenous alternative income and is not affected by outsourcing. Thus, the relevant cost parameter for determining the amount of outsourcing is the profit share. Since higher outsourcing decreases the share of profit the firm oner earns, due to a higher loss in case of a disagreement, this provides - independent of the poer of the union - an incentive for less outsourcing. Hoever, this incentive ill be reinforced by a stronger labor union, since the profit share also reflects the bargaining poer, meaning that a higher union bargaining poer decreases the firm s profit share. As a consequence, the firm ill react ith less outsourcing the stronger the labor union becomes. In contrast, if the parties bargain over the age only, the age is the cost parameter. With that structure, the firm may only realize higher profits by higher outsourcing, if outsourcing has a age-moderating effect. Thus, age-moderation increases the incentive for higher outsourcing in order to reduce the labor costs. Folloing this argument, it is easy to see that the different bargaining structures may lead to different amounts of outsourcing for a given union s bargaining poer. Thus, bargaining over both the age and profit sharing leads to less outsourcing than in the classical age bargaining approach, if outsourcing has a age moderation effect. IV. Conclusions and Discussion The main goal of this paper as to sho the effect of outsourcing on orkers income in an imperfect domestic labor market hich as modeled by a bargaining round beteen a firm and a labor union. In our analysis e distinguished beteen to 9 Also ommerud et al. (009) have found a negative relationship of union s bargaining poer and the amount of outsourcing. Hoever, as mentioned earlier, their analysis differs ith respect to their assumptions concerning the production technology and they model only the classical age bargaining. Nevertheless, the argument for explaining the result is the same. 8
bargaining approaches, one here the firm and the union negotiate over the age only, and a second, as discussed in the political debate, here the parties bargain over both the age and a profit share. For the first case, e found that outsourcing has an ambiguous effect on the orkers income, hile in the second case, outsourcing to lo cost countries ill unambiguously increase the orkers income. Thus, the bargaining regime may be crucial for the income effect. While under a pure age bargaining the fear of losing income entailed by higher outsourcing can be justified under certain circumstances, in a age and profit share bargaining the fear of losing income is unjustified. Concerning the amount of strategic outsourcing e find different results as ell, since outsourcing has different effects on the firm s cost parameters. If the parties bargain over the age only, the age is the cost parameter. Since the impact of outsourcing on the age is a priori ambiguous, the result for the optimal amount of outsourcing is also a priori unclear. Here e find that in case of age-moderation (increase) outsourcing becomes higher (loer) than the level here domestic and foreign marginal labor costs are the same. If the age and a profit share are determined in the bargaining, the age equals the exogenous alternative income, meaning that only the profit share characterizes the firm s cost parameter. If the inputs are outsourced to a lo-cost country, meaning that the domestic outside option is bigger than the marginal costs of outsourcing, the resulting amount of outsourcing is loer than the outsourcing level here domestic and foreign marginal labor costs are the same. Furthermore, in our model e find that - as the profit share increases ith the poer of the labor union and the level of outsourcing - a stronger labor union reinforces the incentive for less outsourcing. Based on this knoledge, e are able to compare the optimal amount of outsourcing and the employment level under the different bargaining regimes. Here e find that for an equally strong labor union, under the assumption of a age-moderating effect in the classical approach, the firm s optimal investment in outsourced products is loer, hile the employment level is higher, in the case of a simultaneously negotiated age and profit share. Thus, in order to reduce the fear of substitution and less income of domestic employment, the union has an incentive to adopt profit sharing as a part of the bargaining round and compensation package. It ould have been interesting to analyze the endogenous choice of the regime by the bargaining parties in our model by comparing the profit the firms oner receives and the union utility under the different approaches on an initial stage prior to our stage. Hoever, despite the fact that e formalize a relatively simple frameork, e are not able to explicitly solve the model for the amount of outsourcing and, due to the 9
0 complexity of the model on this stage, a comparison of the parties outcomes cannot be realized. Another research question could be the implementation of different labor types by assuming that the X -input production uses high-skilled labor, hile in the Y -input production lo-skilled labor is used. An analysis ith this frameork under the classical bargaining approach is done by Koskela and Stenbacka (00). They find that if the labor union represents both types of labor, the high-skilled age increases, hile the lo-skilled age decreases ith higher outsourcing. Hoever, this increasing age dispersion effect of outsourcing can be reduced by a stronger union preference for age solidarity. Of course, also in the case of a bargained age and profit share one could distinguish beteen different types of labor. In the case here the high-skills participate on the firm performance via profit sharing, one ould expect that the age for the highskilled orker equals their outside-option and is unaffected by outsourcing, hile the age for the lo-skilled orker, due to the substitutability of the inputs, is negatively correlated ith outsourcing. Hoever, to be more precise, this question should be analyzed explicitly. Thus, the analysis of income and employment effects for different types of labor if there is bargaining over age and profit sharing is an interesting topic for further research. Appendix A: Derivation of the Wage Effects As the mark-up e have ( ) ( ) [ ] ( )( ) ( ) ( ) ( ) [ ] ( )( ) Z T A = = η η η η, hich depends on the age and the amount of outsourcing. The age impact is shon by Z Z T Z T A =, here ( ) ( ) T / η η = and T Z =. Using this, e obtain ( ) ( )( ) = η η Z A. (A) Since 0 > = η η, e have 0 < A and thus 0 > A A. The impact of outsourcing can be analyzed in a similar ay. Here e have Z Z T Z T A =, here ( ) T = η η and T Z =. Using these expressions e find that
A From (A) e obtain yielding equation (9). [ [ η ( ) ( )] ( )( )( )] η = Z. (A) < A = 0, (A3) > Appendix B: Relationship beteen Profit Sharing and Outsourcing V b Equation (3) can be reformulated to τ =, here V b f ( ) ( ) f ( ) b f '( ) V f '( ) ( V b f ( )) V = b. Thus, the effect of outsourcing on bargained τ b profit sharing is =. Using f ( ) = c, e can rite the effect of outsourcing on the profit share as c V b τ =, (B) ( V b f ( )) hich equals equation (4). References Anderson, S. and. Devereux (989): Profit Sharing and Optimal abor Contracts, Canadian Journal of Economics 89, 45-433. Bhagati, J., A. Panagariya and T. N. Srinivasan (004): The uddles over Outsourcing, Journal of Economic Perspectives 8, 93-4. Braun, F. D. and J. Scheffel (007): A Note on the Effect of Outsourcing on Union Wages, Humboldt Universität Berlin, SFB 649 Discussion Paper, No. 34. Cable, J. R. and F. R. FitzRoy (980): Cooperation and Productivity: Some Evidence from West Germany s Experience, Economic Analysis and Workers anagement 4, 63-90. Conyon,. and R. Freeman (004): Shared odes of Compensation and Firm Performance: U.K. Evidence, in D. Card, R. Blundell and R. B. Freeman (eds): Seeking a Premier Economy: The Economic Effects of British Economy Reforms 980-000, 09-46.
Danthine, J.-P. and J. Hunt (994): Wage Bargaining Structure, Employment and Economic Integration, Economic Journal 04, 58-54. Estrin, S., V. Perotin, A. Robinson and N. Wilson (997): Profit-Sharing in OECD Countries: A Revie and Some Evidence, Business Strategy Revie 8, 7-3. Feenstra, R. and G. Hanson (999): The Impact of Outsourcing and High-technology Capital on Wages: Estimates for the United States 979-990, Quarterly Journal of Economics 4, 907-940. Freeman, R. B. (995): Are Your Wages Set in Beijing?, Journal of Economic Perspectives 9, 5-3. Glass A. J. and K. Saggi (00): Innovation and Wage Effects of International Outsourcing, European Economic Revie 45, 67-85. Geishecker, I. and H. Görg (008): Winners and osers: A icro-evel Analysis of International Outsourcing and Wages, Canadian Journal of Economics 4, 43-70. Görg, H. and A. Hanley (005): abor Demand Effects of International Outsourcing: Evidence from Plant-evel Data, International Revie of Economics and Finance 4, 365-376. Hasan, R., D. itra and R. V. Ramasamy (007): Trade Reforms, abor Regulations, and abor-demand Elasticities: Evidence from India, Revie of Economics and Statistics 89, 466-48. Holmlund, B. (990): Profit Sharing, Wage Bargaining and Unemployment, Economic Inquiry 8, 57-68. Hummels, D., D. Rapoport and K.-. Yi (998): Vertical Specialization and the Changing Nature of World Trade, Economic Policy Revie 4, 79-99. Hummels, D., J. Ishii and K.-. Yi (00): The Nature and Groth of Vertical Specialization in World Trade, Journal of International Economics 54, 75-96. Jerger, J. and J. ichaelis (999): Profit Sharing, Capital Formation and the NAIRU, Scandinavian Journal of Economics 0, 57-75. Jones, D. C. and J. Pliskin (99): The Productivity Effects of Profit Sharing and Worker Representation on the Board, Working Paper Hamilton College, No. 9. Koskela, E. and R. Schöb (00): Outsourcing of Unionized Firms and the Impact of abor arket Policy Reforms, Revie of International Economics 8, 68-695. Koskela, E. and R. Stenbacka (009): Equilibrium Unemployment ith Outsourcing under abour arket Imperfections, abour Economics 6, 84-90. Koskela, E. and R. Stenbacka (00): Equilibrium Unemployment ith Outsourcing and Wage Solidarity, European Economic Revie 54, 376-39.
Kruse, D.. (99): Profit Sharing and Productivity: icroeconomic Evidence from the United States, Economic Journal 0, 4-36. ommerud, K. E., F. eland and O. R. Straume (009): Can Deunionization ead to International Outsourcing?, Journal of International Economics 77, 09-9. Naylor, R. (998): International Trade and Economic Integration hen abour arkets are Generally Unionized, European Economic Revie 4, 5-67. Naylor, R. (999): Union Wage Strategies and International Trade, Economic Journal 09, 0-5. Pendleton, A., E. Poutsma, J. van Ommeren and C. Brester (00): Employee Share Onership and Profit Sharing in the European Union, European Foundation for the Improvement of iving and Working Conditions, Dublin. Perry, C. R. (997): Outsourcing and Union Poer, Journal of abor Research 8, 5-534. Pohjola,. (987): Profit Sharing, Collective Bargaining and Employment, Journal of Institutional and Theoretical Economics 43, 334-34. Senses,. Z. (00): The Effects of Offshoring on the Elasticity of abor Demand, Journal of International Economics 8, 89-98. Skaksen, J. R. (004): International Outsourcing hen abour arkets are Unionized, Canadian Journal of Economics 37, 78-94. Skaksen,. Y. and J. R. Sorensen (00): Should Trade Unions Appreciate Foreign Direct Investments, Journal of International Economics 55, 379-390. Spencer, B. J. (005): International Outsourcing and Incomplete Contracts, Canadian Journal of Economics 38, 07-35. Weitzman,.. (987): Steady State Unemployment under Profit Sharing, Economic Journal 97, 86-05. Yeats, A. (00): Just ho big is Global Production Sharing?, in S. W. Arndt and H. Kierzkoski (eds.): Fragmentation: Ne Production Patterns in the World Economy, Oxford University Press, Ne York, 08-43. 3