Although one of the central questions in the global strategy field is how multinational firms successfully navigate multiple and often conflicting institutional environments, we know relatively little about the effect of conflicting labor market institutions on multinational firms’ strategic choice and operating performance. With its decision to invest in manufacturing operations in nearly every one of the world’s largest welding markets, Lincoln Electric offers us a quasiexperiment. We leverage a unique data set covering 1996-2005 that combines data on each host country’s labor market institutions with data on each subsidiary’s strategic choices and historical operating performance. We find that Lincoln Electric performed significantly better in countries with labor laws and regulations supporting manufacturers’ interests and that Lincoln Electric performed significantly better in countries that allowed unconstrained use of incentive-payfor-performance. Furthermore, we find that in countries with labor market institutions unfriendly to manufacturers, the company was still able to enhance its performance significantly by what we term flexible intermediate adaptation.
Key words: global strategy; institutions, labor market; adaptation; complementarity History: This preliminary draft is dated January 25, 2007. Please do note cite, quote, or circulate without the authors’ permission.
The Essay on Secondary labor market
Many Liberal feminists believe that gender role of socialization is the major reason for sexual division of labor. The main act which is responsible for this change is because of the activities and nurturing that is been followed right from the home of the girl or the boy. At homes there is mainly manipulation and canalization where in girls or boys are forced to use things or do acts which are ...
1. Introduction
One of the most significant questions in the global strategy field is how multinational firms should navigate their way through multiple and often conflicting host-country institutional environments (Ghoshal and Westney 1992).
In spite of the fact that most FDI is still conducted by manufacturing companies whose profitability derives in large part from achieving labor productivity, we still know surprisingly little about whether labor market institutions, defined as formal and informal rules governing the labor market, matter for the operating performance of foreign direct investment, and if so which ones matter and how much they matter. Despite earlier calls for research in this area (Rosenzweig and Singh 1991), there has been little work done on the effect of adaptation to labor market institutions on multinational firms’ strategic choice and performance (see one notable exception by Zaheer, 1995).
The following questions are still largely understudied. Under what conditions should multinational firms simply avoid institutionally incompatible environments when the institutions affect basic rules of work, and what in fact makes an environment institutionally incompatible? How far should multinational firms go in adapting to different institutional environments? When should a multinational firm hold the line and introduce what at first appears to be an incongruent organizational practice into a host country? Whereas some past studies have examined the likelihood of transfer of organizational practices abroad, remarkably few have examined the effect of strategic adaptation on firm performance. Adaptation is one of the important strategies in international business (Prahalad and Doz, 1987; Bartlett and Ghoshal, 1989; Ghemawat 2003), and yet a recent literature review shows that little is known about the optimal level of adaptation by multinational firms to any local market institutions (Dow 2006).
To answer these questions, we leverage a unique quasi-natural experiment. Lincoln Electric is one of the most studied companies of all time in MBA programs, and it operates in a representative manufacturing industry where profits have historically derived in no small part from labor productivity. Furthermore, Lincoln Electric pioneered a diverse set of incentive practices in the first part of the 20th century, many of which—particularly the discretionary bonus—have been increasingly adopted. One of the reasons that there have been so few studies of multinational firms interacting with conflicting labor market institutions is that it is costly to go to each country and measure the labor market institutions, and it can be even more costly and often infeasible to interview every operations manager of every host country for even a single multinational firm.
The Essay on Oligopolists Market Price Firms
Oligopolists There are four market structures in our economy today: Perfect competition, monopolistic competition, oligopolies and monopolies. This essay shall describe the oligopoly market. The definition of an oligopoly states that in an industry, a small number of firms dominate the market. There are a low number of firms in the industry, becase and adding to the barriers to entry. The barriers ...
We have been able to spend the past year interviewing Lincoln Electric managers around the globe, and we have at the same time interviewed local labor market experts and studied local labor laws and regulations around the globe. Together, these efforts have allowed us to implement what is to our knowledge the first quasi-natural experiment on the effect of diverse labor market institutions on a global firm’s strategic choices and performance. We emphasize that this is a quasi-natural experiment because Lincoln Electric long ago decided that it needed to be present in nearly every one of the world’s largest welding markets across four out of five continents; through this decision rule, Lincoln Electric was not more or less likely to enter a specific large market because of its labor market institutions. That is what makes this study a quasi-natural experiment. Lincoln Electric is notable within the welding industry for its investment in a diverse range of the largest 20 markets (in North America, South America, Europe, Asia, and Australia).
Yet while Lincoln Electric is unusual even in its industry for its broad global footprint, it is important to point out that it is likely representative of a wider universe of horizontal U.S. multinational enterprises (those producing and selling directly in foreign markets to foreign consumers).
Bognanno et al. (2005) show how U.S. MNEs during 1982-1991 did not like host countries with strong unions and work councils but would still enter when the market was large. Our analysis is supported by an effort to control for alternative explanations (including alternative institutional explanations) as well as an effort to correct for serial autocorrelation in performance within countries over time. We find that Lincoln Electric performed significantly better in countries with labor laws and regulations that support manufacturers’ interests and that Lincoln Electric performed significantly better in countries that allow unconstrained use of incentive-pay-for-performance. Furthermore, we find that in countries with labor market institutions that were unfriendly to manufacturers, the company was still able to enhance its performance significantly by what we term flexible intermediate adaptation.
The Term Paper on Remuneration Labor Market
Abstract The real wage of the average worker in the United States has fallen 13 percent in the last 20 years, while the average chief executive officer (CEO) has received a pay raise of over 300 percent (Crystal, 1991). This glaring contrast has sparked an explosion in academic research on executive compensation. The argument that Crystal (1991) proposed, ! ^0 just about all of the rational ...
Together these findings suggest that the theory of strategic complementarity (Milgrom and Roberts 1992) needs to be made more globally and institutionally contingent. A bundle of managerial practices may be complements with each other (with the payoffs for each increasing in the presence of the others), but that complementarity may be directly predicated on the precise characteristics of fundamental capitalist institutions in the country, such as whether the institutions carry more of a pro-labor or a procapitalist orientation. Practices that are complements with one another in the United States may often not be complements with one another in a different institutional environment. While the laws of efficiency do not vary radically across the globe, the optimal bundle of practices may vary to a moderate degree according to the institutional variation across countries.
The multinational firm should be cognizant of the effect of labor market institutions on performance, and if the effect is large enough the firm may need to consider moving its resources to friendlier institutional environments. If a firm needs to be in a given market for other reasons (such as market size), the optimal mix of labor market practices is not radically different but needs to be customized in what we term a flexible, intermediate form of adaptation. The optimal mix of labor practices for Lincoln Electric was often different in each country, and the successful adaptations were based on managers optimizing between the bundle of practices that worked most efficiently in Lincoln’s flagship Cleveland, Ohio operation and the practices that could work best in each host institutional environment.
We proceed as follows. First, we discuss past research on labor market institutions and their effect on strategic choice and performance. Most such studies look at a single country rather than studying the global firm encountering multiple institutional environments. Second, we briefly introduce Lincoln Electric and its historical business strategy. Third, we discuss our method for implementing the quasi-natural experiment using data around the world from Lincoln Electric. Fourth, we discuss the data we have collected. Fifth, we discuss our results, and finally we conclude by relating how our study can inform future research.
The Term Paper on Small Firm Financing Company Companies Market
Financing a small firm can be achieved in three ways. The most preferable but at the same time the least likely is self financing from retained earnings, otherwise, the firm will have to resort to either one of the two following financial markets. Debt capital and equity capital (which strictly speaking is the same as retained earnings, both having their advantages and disadvantages. Only after ...
2. Labor Market Institutions and Strategy
Most prior research in this area has looked at firms within a single country and shows that prolabor oriented institutions mean lower company productivity and/or profitability (i.e., Clark 1984; Bemmels 1987; Doucouliagos 1995; Frank 2007).
Frank (2007) most recently implements an ingenious natural experiment in Germany showing that just after the fall of the Berlin Wall, companies with work councils and labor unions suffered a deteriorating relative cost position and had to outsource more and more activities to outside firms that did not have both work councils and labor unions. Of course, a small minority of studies argue that pro-labor-oriented institutions need not automatically lead to poor firmlevel performance, but these studies suggest that firms bear significant adjustment costs for overcoming these institutions (i.e., Mefford 1986; Boal 1990).
Surprisingly, there has been no work to our knowledge that has sought to explain multinational firms’ performance as a function of labor market rules and regulations in each host country. One study looks at the presence of unions across a multinational firm’s 31 plants and finds that the negative effect of unions on performance can potentially be counteracted but at cost in retraining the management and shifting away from labor-oriented production (Mefford 1986).
Moreover, while there has been a great deal of work within the international human resource management literature focused on Japanese labor practices and what explains why some Japanese firms transfer their practices to the U.S. while others do not, there has been very little work overall on how labor market institutions affect strategic choice in multiple host nations. The literature seems sharply divided between those authors who hypothesize that multinational firms should adapt as little as possible to local labor market institutions, and those authors who hypothesize that multinational firms should maximally adapt to the full range of local institutions. Actual empirical tests of the effect of adaptation on performance are few and far between (Dow 2006).
The Research paper on Unfair Labor Practice and Apparel Companies
This paper draws on interviews with Michael Kobori, Vice President Corporate Code of Conduct, LeviStrauss at his office in San Francisco in 2006 and Caitlin Morris, Director of Engagement and Integration, Labor Compliance Department, Nike, at Nike’s Headquarters’ in Beaverton, Oregon in 2006, as well as discussions and/or e-mail communications with: Neil Kearney, General Secretary, International ...
Among those who advocate maximum adaptation to local labor market institutions, the main arguments were inspired by neo-institutionalization theory. Kostova (1999) defines institutional distance based on three dimensions from institutionalization theory–regulatory, cognitive, and normative—and goes on to argue that if the organizational practice was in conflict with local laws and regulations, then the foreign employees would not be willing to take the risk of instituting the practice.
Earlier work by Rosenzweig and Singh (1991) makes a related argument, namely that legal requirements in the host countries are the strong form of adaptation constraint because they represent a form of “coercive isomorphism” (DiMaggio and Powell, 1983).
Rosenzweig and Nohria (1994) argue that of all organizational practices, labor practices are the most likely to be adapted to the local environment, given that they are often mandated by local laws and regulations or else enforced by informal norms and conventions. Moreover, if the practice were in conflict with the cognitive and normative institutions in the local environment, then local employees would have difficulty understanding, interpreting, and judging the practice correctly (1999, 315).
Multinational firms would likely face resistance if the practice violated the different assumptions and value systems of the local culture (Beechler and Yang 1994).
Yet there are few empirical tests of these general institutional hypotheses. Beechler and survey of foreign multinational firms operating in the U.S. Still, perhaps U.S. unions were simply weaker in pressing their demands for local adaptation then unions in some European countries might be (Rosenzweig and Nohria, 1994: 249).
In their survey of multinational firms operating in Greece, Myloni, Harzing, and Mirza (2004) find mixed evidence in support of this hypothesis. Greek unions may have been able to push successfully for local adaptation on some issues but not on others. Rosenzweig and Singh (1991) posit that market-seeking horizontal MNEs were dependent on local resources and thus had a greater need to gain local legitimacy both to access these local resources and to secure local sales. Thus, local legitimacy depended on adaptation to local organizational practices. Zaheer (1995) hypothesizes that a market-seeking horizontal MNE would see the greatest need to adapt to local practice, because these MNEs compete on a local-for-local basis (i.e., with products produced and sold locally) and therefore have greater need to engage in local isomorphism. Zaheer (1995) in her analysis finds, however, that American MNEs in the foreign exchange trading industry, in spite of being market-seeking, horizontal MNEs, actually benefited by not imitating local Japanese practices related to using microcontrols and not using market controls.
The Review on The Relationships Between Reward Systems Being Implement in the Company and Firm Performance
1.0 Introduction Understanding what motivates people is necessary at all levels of management. Deci, E., Koestner, R. & Ryan, R.M. (1999) had said that motivation is generally linked to reward, and it is widely recognized that reward management is central to the regulation of the employment relationship. The reward system varies from organization to organization, and comes in various and ...
Few authors argue, in contrast, that firms should not adapt to local labor institutions. When a firm has a universally superior organizational practice, it can transfer the practice successfully to even institutionally distance countries and still see a performance benefit from doing so. Zaheer (1995) shows that in the case of foreign exchange trading rooms, market controls were clearly a superior organizational practice whereas microcontrols on traders were clearly a practice detracting from performance. Market controls are defined as rewarding each employee “in proportion to their contribution” (Ouchi, 1979: 835), and included basing a high proportion of traders’ total incomes on performance-linked bonuses, hiring experienced traders from the external market, and employing high turnover among traders. Microcontrols involve extensive use of rules and the legitimacy of authority, and in Zaheer’s industry context meant firms’ setting detailed limits on intraday and overnight open positions by currency and by trader in an attempt to micromanage speculation by individual traders.
Zaheer finds that American firms with market controls benefited from transferring those practices to Tokyo, and furthermore that American firms without microcontrols benefit from not adapting to the Japanese practice in Tokyo of using microcontrols. Even if an organizational practice is not universally superior, it may be successful in a culturally distant market because it allows the firm to operate in a manner that some sector of the population would prefer. Gamble (2003) describes the example of a UK-based retailer that transferred its flat hierarchy and use of first names to its store in Shanghai, China. These practices actually may have attracted the retailer’s desired set of workers and may have facilitated superior performance in Shanghai. Using a related logic, Szulanski and Jensen (2006) argue that if the firm does not fully understand the complexity of interactions among its management practices and activities (Rivkin 2000), then the firm should not “presumptively” adapt to different local institutions and should not presumptively change its complex interaction of practices and activities.
In summary, past work would suggest that pro-labor-oriented institutions negatively impact firm performance, but there have been virtually no studies of multinational firms assessing which among the possible set of labor market institutions matter most and how much they matter. We do not know whether multinational firms from countries with institutions favoring firms should systematically avoid countries with pro-labor-oriented institutions or whether any institutional difference can be bridged through adaptation to local practice. Moreover, we know very little about what is the optimal amount of adaptation to local practice.
3. Method
One explanation for the paucity of work on labor market institutions, adaptation choices, and multinational firm performance is found in a combination of data constraints. To test for causality, ideally one would hope to have a natural experimental design, in which the same firm interacts with widely varying institutions. Even when one finds a firm that has decided as it decision rule that it needs to be present in all large markets, data availability has been the primary constraint. To measure the interrelationship between institutions, strategic choice, and performance, one would desire to go inside a multinational firm and hand-collect data from across its subsidiaries. We spent the past 12 months collecting such data (which will be described below), and we use the data to estimate the following OLS regression for country-years during 1996-2005:
(1) ROA it = β 0 + β 1 ∗ Labor market institution it
+ β 2 ∗ GDP Growth Rate it + β 3 ∗ Only Consumables it + β 4 ∗ Only Machines it + ε it ,
where subsidiary i’s ROA in year t is determined by the labor market institution of interest, the GDP growth rate, whether the subsidiary only produces consumable welding products in the host market, and whether the subsidiary only produces machine products in the host market. Because the labor market institutions are time-invariant, we cannot use a fixed-effects model. However, we are able to cluster the standard errors by host country and, to correct for serial autocorrelation, we later run a robustness check using the Cochrane-Orcutt correction.
4. A Wider Range of Labor Institutions
While prior studies have focused in large measure on a dummy variable for union strength, we propose to examine a richer variety of labor market institutions that might substantially affect multinational firms’ performance abroad. In this study we look at five distinct types of labor institutions. The first set of institutions focuses on government’s attempt to mandate working conditions and to protect workers from disability and other injury incurred in the workplace. Governments enact these mandatory constraints on working conditions in order to protect workers from perceived abuses of employer market power. We focus in this study on legal limits on the number of hours in the workweek and on the law defining who pays for disability insurance (the worker, the employer, or the government itself).
We examine whether all mandatory constraints hinder multinational firm performance and whether some government interventions, such as the local government’s assuming the costs of disability insurance, actually raise multinational firm performance.
A second set of institutions focuses on the ability of workers to join together for collective action. We are interested in institutions such as closed shop rules make unions more powerful in demanding wage gains and other manufacturer concessions. Prior work has also shown that national collective agreements are on average less flexible than local or firm-specific union agreements in allowing firms to adapt to technological change and changes in market demand (ie, Locke 1992).
We are therefore also interested in the percentage of workers in a country that are covered by national collective agreements. The third set of institutions focuses on empowering workers to participate in management. Some countries legally empower workers to participate in management through workers’ councils. The effect of these institutions may be to limit management flexibility in investment and in flexible work rules on the plant floor. Alternatively, participation may also result in greater long-term worker commitment and
flexibility. The net effect of this set of institutions is unclear and is fundamentally an empirical question. The fourth set of institutions focuses on legal rights for workers during a labor dispute, and specifically whether workers receive any legal protection during a strike and whether a third-party arbitrator is authorized by law to resolve the dispute. These institutions clearly can serve as an incentive for workers to strike as a means of realizing their demands. Also in this category are institutions that make dismissing workers relatively difficult or easy.
The fifth set of institutions focuses on the ability of companies to use pay-for-performance incentive practices. Whereas the four sets of labor institutions described above have data for the late 1990s compiled by Botero et al. (2004), this latter set of institutions has surprisingly not been the subject of past study or data collection efforts. Hence, we develop our own database and collect our own data for this set of labor market institutions. We focus first on companies’ ability to use piecework, not because piecework specifically is used by many companies but because piecework is a proxy for overall pay-forperformance incentives. Next, we focus on companies’ freedom to pay discretionary bonuses to workers. Some countries actually restrict companies from paying a bonus one year and none the next. In other countries, companies are required to pay out a set percentage of profits to all eligible employees, regardless of individual contributions. The net result is that companies may be limited in their ability to use pay-for-performance incentives in host countries.
5. The Case of Lincoln Electric
Lincoln Electric is a welding manufacturer based in Cleveland, Ohio, the company over the past 60 years has been the subject of repeated academic and teaching interest for its use of incentives. Founded in 1895, the company produces both welding machines and consumable products for those machines. In 2005, the company had manufacturing operations in 19 countries and $1.86 billion in revenues. Historically, the company has been able to outlast and defeat significant multinational competitors in the welding industry such as General Electric and Westinghouse.
The company’s historical competitive advantage has stemmed from its industry-leading productivity, which has been attributed largely to the company’s management system. The system consists of four main components: the use of piecework, a discretionary annual bonus based on individual and company performance, an individual merit rating used to determine the annual bonus, and a voluntary employee advisory board which works to generate productivity-enhancing innovations. But in addition to those four main components, there are also a number of complementary management practices. Lincoln uses few supervisors on the plant floor and attempts to assign a great deal of trust and autonomy to its factory employees. In Cleveland, there were tens of thousands of piece rates set, and workers have been trusted for decades to record their output accurately. Lincoln began expanding abroad in the 1940s, but its first major foreign investments occurred in the late 1980s and resulted in initial failure.
There were numerous potential explanations for the difficulties faced after this first wave of expansion. The company wanted its new foreign subsidiaries to operate in Lincoln USA’s image, and the company’s international managers were expected to introduce piecework, a bonus system, and an advisory board.1 Many workers in Western Europe in particular did not want to adopt the company’s recommended practices (Hastings 1999).
The company was also unlucky in having bought companies in Europe just prior to a global economic downturn. Past case histories of Lincoln Electric’s experience during this period of time hypothesized that the company’s strategic success was incumbent on a host country’s having pro-capitalist/pro-manufacturer institutions (Chilton 1993a, 1993b; Dawson 1999; Hastings 1999; Maciariello 2000).
Still, none of the studies ever tested this hypothesis more than anecdotally.
Starting in 1996, companywide profitability returned, and the company renewed its global expansion. Over the next several years, Lincoln Electric chose to expand to nearly every large welding market in the world. It is because of that decision, and because there is a remarkable diversity of labor market institutions across those markets, that we have a unique quasi-natural experiment to implement. 6. Data
The dependent variable is ROA in the host country, measured as operating income divided by total assets. The sample is the entire set of country operations including the U.S. belonging to Lincoln Electric in its welding business in years 1996-2005. A small operation that Lincoln Electric owned in Ireland, but which was an entirely different business line and was managed separately, was purposely excluded from the analysis. Also, Lincoln Electric owned a minority stake in its Taiwanese joint venture partner, but had no say in the Taiwanese company’s operation, and thus Taiwan was excluded from the analysis. Lastly, at the very end of the period, Lincoln Electric purchased a plant in Colombia, but since Lincoln Electric had barely closed the acquisition at the end of the period, Colombia is not part of the sample. The results are substantively similar with or without the small Irish operation. The results are also substantively similar with or without the U.S. operation.
The data were given to the authors by the company, and the sample covers years 1996-2005. As a robustness check, we will test in Table 9 whether the same results obtain when we use a productivity proxy, operating profits per employee, as the alternative dependent variable. We use operating profits per employee because gross margin per employee in this industry appears to be primarily a function of country-level wealth. Also, we will show that operating profitability is not correlated with wage levels, but is closely connected to labor productivity. Therefore, the proportion of the gross margin pie that remains as company operating profit is based in large part on labor productivity. We thus use as our alternative variable operating profits per employee, which in turn reflects the fact that final operating profit in this industry is not dependent on country-level wealth but is directly dependent in large measure on labor productivity.
The first set of independent variables focuses on labor market institutions in the host country. b97hours_week measures the maximum duration of the regular workweek (excluding overtime), normalized from 0 to 1, where higher values mean less hours of work (higher protection).
If there is no legally mandated limit, the variable equals zero. b97work_man_c measures the protection of the workers’ right to participation in management in the country’s constitution. The variable equals one if there is a ‘right’ to participation in management expressly granted by the constitution. It equals 0.67 if participation in management is mentioned as a state public policy or public interest (or simply mentioned within the chapter on rights).
It equals 0.33 if participation in management is mentioned in the constitution. It equals zero otherwise. b97index_work_man1 measures the legal rights of workers to participate in the management of the companies. This index is computed as the normalized sum of: (i) worker participation by law; and (ii) right to worker participation in management in the constitution. Public disability benefits measures the number of months of contributions or employment required by law to qualify for sickness benefits, normalized from 0 to 1, where higher values mean less contribution (higher protection).
Data for these four variables come from Botero et al. (2004).
The next set of variables focuses on what legal protections workers enjoy during a labor dispute. b97index_arbitr equals one if compulsory third party arbitration during a labor dispute is mandated by law or if the government is always entitled to impose compulsory arbitration on the parties to a labor dispute. It equals zero otherwise. b97index_col_disp1 measures the level of protection of workers during a collective dispute. The index is computed as the normalized sum of: (i) legal strikes; (ii) procedural restrictions to strikes; (iii) employer defenses; (iv) compulsory third party arbitration during a labor dispute; and (v) right to industrial action in the constitution. b97index_industrial1 measures the level of protection of industrial (collective) relations laws. The index is computed as the normalized sum of: (i) subindex of collective bargaining; (ii) subindex of worker participation in management; and (iii) subindex of collective disputes. Data for these three variables come from Botero et al. (2004).
Next, we control for the real GDP growth rate. Data come from the Economist Intelligence Unit Country Data, for years 1996-2005. We also control for the product market coverage of Lincoln’s local subsidiary. Only Consumables is set equal to 1 when the local subsidiary only produces consumable welding products, and is set equal to zero otherwise. Only Machines is set equal to 1 when the local country operation only produces welding machines, and is set equal to zero otherwise. In a set of robustness checks, we also test for the effect of a series of related and collinear business-level variables. We test for Lincoln’s prior year market share, its prior year share of the “accessible market,” its selfevaluation of its product range relative to its best local competitor in machines, and its self-evaluation of its product range relative to its best local competitor in welding consumable products. Data come from Lincoln Electric.
Next, we examine the effect of specific labor laws and regulations related to piecework and discretionary bonuses. To collect data on these variables, we implemented an extensive survey of local country labor laws in addition to conducting interviews with local lawyers and labor experts in most of the 16 individual countries. We measure (1) whether the law allows the unrestricted use of piecework and discretionary bonuses as compensation; (2) whether the law allow wages to be paid on a piece-rate (ie, by the piece, rather than by the hour) basis; (3) whether the law requires companies to meet a minimum wage when paying piecework; (4) whether the law requires that workers being paid piece rate also received paid vacation; (5) whether the law requires that workers being paid on a piece-rate basis be given the same amount of paid vacation as equivalent workers, based on rank or seniority; (6) whether the use of piecework is restricted to certain industries; (7) whether the law permits employers to pay discretionary bonuses as the only form of bonus or variable pay; (8) whether the law permits employers to pay discretionary bonuses, but only as a supplement to some other required profit-related bonus or variable pay; (9) whether the country has regulations that prescribe a calculation formula or a lower/upper bound for discretionary bonuses; and (10) whether the law of the country requires a certain distribution of discretionary bonuses (eg., that bonuses be paid to all if paid to any, that bonuses be paid to all workers in a given type of position).
Next, we control for the role of labor costs. We collect data on manufacturing average hourly wage costs by country in U.S. dollars from the KILM data set of the International Labour Organization
(ILO).
We have hourly wage cost data covering individual years 1996-2004, and as a robustness check we also use the average for the 1996-2004 period. Next, we control for the host country’s level of development by taking the natural logarithm of GDP in 1996 constant dollars as well as the natural logarithm of GDP per capita. Data come from the Economist Intelligence Unit Country Data and are available for years 1996-2005. To test for the role of policy instability in influencing FDI flows, we use Henisz’s (2000) political constraints index (POLCONIII).2 This measure estimates the feasibility of policy change (the extent to which a change in the preferences of any one political actor may lead to a change in government policy) by considering each political actor’s ideology and veto power. Specifically, we take the squared distance of origin and host country-year observations.
To test for the role of corporate tax differences across countries, we use data on corporate tax rates from the World Tax Database of the University Michigan Office of Tax Policy Research (OTPR).
We take the origin country’s top corporate statutory tax rate and subtract from it the host country’s top corporate statutory tax rate. Desai, Dyck, and Zingales (2005) use the OTPR data and, in their examination of the impact of taxation on corporate governance, they emphasize the role of so-called high statutory corporate tax rates. Because the Michigan corporate tax data is available only through year 2002, we went to the original source of the data, the Center for International Trade and Economics (CITE) at the Heritage Foundation, and were able to augment the Michigan data set with corporate tax rates for 2003-2005.
Next, we control for the rule of law. We take the squared difference between the United States and each host country’s scores on the Rule of Law (legality) index for 1998, the latter being an index of perceived compliance with protection of legal entitlements (property and contractual rights), law and order, etc. Data come from Kaufmann, Kraay, and Mastruzzi (2003).
Next, to control for the role of geographic distance from the United States, we take the natural logarithm of the great circle distance between capital cities. Because we do not wish to exclude the 2 United States from our data set, and following standard methodological practice in economics, we take the natural log of (great circle distance in kilometers + 1).
This allows the U.S. to take a zero value on the natural log measure. We check and confirm that we get similar results with or without the United States and with or without adding 1 to the great circle distance. Data come from the Gleditsch-Ward Data Set on Great Circle Distance Between Capital Cities, accessed in January 2006 from http://dss.ucsd.edu/~kgledits/capdist.html.
Distance data for any missing country pairs was accessed in January 2006 from http://www.airport-accommodation.co.uk/worlddistances.php. Next, we examine data on actual adaptation of the Lincoln management system to each host country. We interviewed Lincoln managers on the ground to determine the local use of four major elements of Lincoln systems. We created dummy variables for whether the company uses piecework in its local operation; for whether the company pay discretionary bonuses in its local operation; for whether the company uses individual merit ratings in its local operation; and for whether the company uses an employee advisory board or an equivalent organizational practice. We set dummy variables equal to 1 if the local operation used a particular practice, and zero otherwise. We also used these variables to set up interaction variables to count the combined use of multiple Lincoln system elements and also separately to measure the combination of specific system elements.
For a final robustness check, we include controls for the host country’s real interest rate, the host country’s relative price of capital, the real price of foreign exchange, the annual change in real wages, and the volatility in the host country’s foreign exchange rate. rlint_long measures the real interest rate, calculated by adjusting the nominal interest rate using a ratio of GDP deflators for the current and succeeding years. relprcap_n measures the relative price of capital, defined as the fixed investment deflator divided by the GDP deflator. rlprfx_cpi_av measures the real price of foreign exchange, calculated as the average exchange rate (in terms of local currency/US dollars) adjusted by a ratio of U.S. CPI to local-country CPI. The resulting values for each country were then normalized about their (country) means and multiplied by 100. realwg_pct measures the percentage change in hourly wages in local currency adjusted for inflation, over the previous year.
Finally, we use multiple measures of foreign exchange rate volatility. To do so, first we take the real price of foreign exchange and then calculate the standard deviation of quarter-to-quarter changes in the real price of foreign exchange within each year. We examine the volatility using alternatively either the nominal quarter-to-quarter change in the real price of foreign exchange or else the percentage quarter-to-quarter change. As a further robustness check, we confirm that the results are similar no matter whether we use the U.S. CPI or the U.S. PPI to calculate the real price of foreign exchange. Data for the previous five variables come from the Economist Intelligence Unit Country Data, the OECD, and Global Financial Data.
7. Results
Summary statistics are presented in Table 1, and a correlation matrix is presented in Table 2. As shown, a number of the labor institutions are highly correlated with one another. They should therefore be seen as members of an institutional set, and one should be careful in attributing causality to any one specific labor institution. Still, as members of a coherent institutional set, it is clear that some countries create protections for workers that limit the workweek, that empower them to participate in management, that protect workers’ interests during a labor dispute, and that provide public assistance in case of individual employee injury. These institutions tend to go together; in other words, a country that limits the workweek also tends to empower workers to participate in company management and also tends to protect workers’ interests during a labor dispute.
What this means is that these pro-labor institutions should be treated as a set of complementary institutions, and causality should be assigned to the set and not to an individual component. Moreover, because of collinearity among these pro-labor institutions, it is best not to enter them together into a regression and instead to enter them one at a time to test their statistic significance, or even better, to conduct a principal component analysis and use the first component. The exception to this institutional set is that countries that provide public assistance in case of individual employee injury appear to do so as a means of assisting employers’ expenses. These countries are actually less likely to limit the workweek or to empower workers to participate in company management.
The results of our initial model specification are presented in Table 3. Labor institutions that limit the workweek and that give workers the right to participate in management are negatively associated with Lincoln Electric’s country profitability in a statistically significant manner (p < .05).
More proemployer rules of the game, including one that uses public assistance to cover employee injury, are positively associated with Lincoln’s country profitability (p < .01).
Still, it is important not to assign causality to these individual regulations. They are instead part of a set of labor institutions that either seek to empower and protect labor, or else seek to protect companies and labor jointly (as in the case of public disability benefits).
We further find that company profitability is aided by the individual country’s real GDP growth rate as well as from the company producing both machines and consumable products locally.
Because we were curious to see whether machines or consumable production was more important to country profitability, we enter “Only Machines” and “Only Consumables” and used “Both Machines and Consumables” as the reference case. It is clear that company profitability was aided by the joint production of machines and consumables. As shown in Table 3, it is not clear whether machines or consumables were more profitable in countries where the company produced only one general product type. The overwhelming result is that joint local production of machines and consumables was positively associated with country profits.
In Table 4, we consider the effect of other variables from Lincoln’s internal data on market share and product range relative to the best local competitor in each region. We find that both market share and product range are positively and significantly associated with profitability. Still, the various measures of labor institutions continued to be significant even after controlling for GDP growth and either of the market share or product range measures.
Next, we examine whether our results are simply being driven by differences in labor costs. As shown in Table 5, we find that hourly labor costs averaged for 1996-2001 are simply unassociated with company profits. Indeed, based on labor costs, we would expect countries such as China and Indonesia to be more profitable; yet in fact, countries with relatively high labor costs such as the U.S. and Germany are among the most profitable. This result may be evidence that productivity in the individual plant and not unit labor costs were primarily responsible for performance, a finding that is consistent with Cushman’s (1987) earlier evidence that productivity and not unit labor costs were most important for attracting overall U.S. foreign direct investment during the period 1963-1981. Next we examine whether Lincoln’s operating performance depends upon the development level of the host country. We find in the bottom of Table 5 that the log of host country GDP has no significant association with profitability. Furthermore, the log of host country GDP per capita is positive but not statistically significant either.
In Table 6 we examine the effect of specific institutions that either enable or constrain the use of piecework and discretionary bonuses. We find in Model 1 that Lincoln Electric’s profitability is significantly greater in countries that allow unrestricted use of piecework and discretionary bonuses as the main form of worker compensation. Interestingly, as shown in Models 2 and 5 of Table 6, it is not enough to have unconstrained use of just piecework or just discretionary bonuses. The one type of specific piecework regulation that is directly tied to country-level profitability is shown in Model 4. Lincoln Electric’s profitability is significantly lower in countries that require that paid vacation be paid on the same basis for piecework employees as for hourly/salaried workers. Other individual piecework and discretionary regulations are statistically insignificant, and in Model 7 we confirm that it is the unconstrained use of both piecework and discretionary bonuses that that remains highly significant even after controlling for public disability benefits (our earlier proxy for the pro-manufacturer nature of overall labor regulations).
We then examine whether the transfer of specific Lincoln management system components is essential to company profitability. At first glance in Panel 1 of Table 7, the transfer of specific Lincoln components is positively associated with profitability. But when labor institutions are included as a further control variable in Panel 2, the statistical significance goes away. We conduct a principal component analysis of the various labor institutions analyzed individually in this study (those shown in Tables 3 and 6) and use the first component in Panel 2 of Table 7. The first component is the same as taking a weighted average of the various labor institutions. A higher value on the first component signifies stronger labor protections, and a lower value on the first component reflects a pro-employer set of labor institutions. We find that labor market institutions are fundamental factors driving both the transfer of specific management system components as well as overall company profitability. As shown in Panel 3 of Table 7, if a host country has pro-labor-oriented institutions, Lincoln Electric more often only transplanted either one element of its management system or no elements at all.
On the other hand, if a host country has pro-employer-oriented institutions, Lincoln Electric more often transferred all four of the main elements of its management system. What is most interesting is that we discovered through the course of our interviews that multiple Lincoln Electric subsidiaries have implemented interesting intermediate combinations of Lincoln practices best suited for the local host environment. We find in Table 7 that those truly intermediate and customized adaptations involving two of the four Lincoln components were not collinear with pro-labor or pro-employer orientation of the local labor institutions. When we interviewed Lincoln Electric subsidiary managers, we found that they had gone back and forth between Cleveland and their host country before deciding on an intermediate set of practices that would provide strong incentives but also operate smoothly in the local institutional environment. Furthermore, the results in Table 8 show that these customized/intermediate adaptations were positively and significantly associated with company profitability (p always < .01).
Again using the principal component of labor market institutions, we find pro-labor institutions are negatively associated with Lincoln’s country-level profitability (p always < .05 and often < .01).
As shown in Table 8, while customized intermediate adaptation helps profitability at the country level, pro-labor institutions at the same time hurts profits. Customized intermediate adaptation is further shown to be robust in Model 10 to the inclusion of variables for all levels of adaptation (with zero adaptation as the reference case).
We test and find that the results are not just statistically significant but economically significant as well. Using the final Model 10, a shift from 0 to 1 on the flexible intermediate adaptation variable is associated with an 8.2 percent nominal increase in ROA. The effect of moving from the country with the most manufacturer-friendly labor institutions to the country with the most labor-friendly labor institutions is associated with a 10.7 percent nominal decrease in ROA. A one-standard deviation increase in the prolabor orientation of local institutions is associated with a 2.9 percent nominal decrease in ROA.
We also run a series of other robustness checks. We find in Table 8 that the results hold even after the inclusion of real GDP growth rate, the earlier business mix variables, Henisz’s POLCONIII index used as the distance between political stability in the host country and the U.S., the difference in corporate taxes between home and host countries, the distance in the rule of law between the U.S. and host country, geographic distance, and host-country log of GDP per capita. Finally, we show that our results are not the result of serial autocorrelation. We perform the Cochrane-Orcutt correction for AR(1) and show at the end of Table 8 that our main results still stand. Among the control variables, the PolconIII distance measure was significant but then turns insignificant after implementing the AR(1) correction. Geographic distance is negatively associated with company profitability and highly significant (p < .01) both with and without the AR(1) correction.
In a final robustness check not reported here, we find that Model 8 is further robust to the addition of the host country’s real interest rate, the host country’s relative price of capital, the real price of foreign exchange, the annual change in real monthly wages, and the volatility in the foreign exchange rate. Because of high collinearity among those last four control variables, we only include two of them at a time to avoid all of them turning insignificant. Finally, we run a robustness check to further confirm that the profitability results are in fact consistent with our labor productivity story. In Table 9, we find that in fact our proxy for labor productivity, operating profits per employee, is determined in large part by the same labor market institutions and adaptation choices as is ROA. This is consistent with the fact that in this industry gross margin is highly correlated with country wealth, but final operating profit is determined instead in large part by labor productivity.
8. Discussion and Conclusion
In summary, we show in this paper that the global profitability of a major and often-studied multinational manufacturer known for its use of pay-for-performance incentives depends on a combination of factors. Even after controlling for a host of alternative explanations, we find that local labor institutions exercise a direct effect on company profits. Moreover, the company’s ability to find an intermediate form of adaptation to a host country institutional environment is positively associated with profitability. This study suggests that prior studies of FDI performance may have missed an important set of institutional variables, namely those related to the labor market. While other institutions examined in prior studies, such as policy stability and the rule of law, are no doubt important, it should be evident from this study that labor market institutions may have a comparably large effect on FDI performance. The obvious next step is to look at the effect of labor market institutions on a large sample of multinational firms, both those in manufacturing and those in service industries. We hope to undertake that project next.
This study suggests that intermediate and flexible adaptation is better for multinational firm performance than either wholesale transplantation of home-country practices or a total adaptation to local institutions. In future work we hope to examine how companies find these intermediate and flexible forms of adaptation and whether there are generalizable lessons from the Lincoln Electric case for the wider population of manufacturing and service sector multinational firms. Finally, this paper focuses on the firm, Lincoln Electric, that has been the poster child for teaching the virtues of complementarity (Milgrom and Roberts 1995).
Yet we find that there is no universal bundle of practices that is optimal for every institutional environment. Instead, there often is a unique bundle that needs to be recalibrated to fit with the relevant local institutions. Our study strongly suggests that the theory of complementarity needs to be made more institutionally contingent. Optimality in a firm’s bundle of management practice is likely dependent on the nature of local institutions, and in no small part dependent on the pro-labor or pro-manufacturer orientation of these institutions that vary significantly across countries.
Acknowledgments
We acknowledge the Harvard Business School Division of Research for funding support. We are grateful to Lincoln Electric for cooperation, but the analysis, findings, and views expressed in this study are entirely our own. We thank Botero and colleagues (2004) for making their data publicly available online.
