International Trade Negotiations and the Trans-Border Movement of People: A Review of the Literature
Recent decades have seen a large reduction in barriers for many traded goods and services, as illustrated by the trade agreements reviewed in the previous section. There has also been considerable liberalisation of markets for financial capital. However, the removal of barriers to the movement of people has been relatively limited. Of course, growth in immigration to developed economies has been significant, with the share of migrants in the population of high-income countries almost doubling between 1970 and 2000. However, of the current world population of 6.6 billion no more than 3 percent are resident in a country other than their country of birth. This contrasts markedly with trade, where the ratio of worldwide imports to GDP is around 10 percent (Hatton, 2007). While for OECD countries the ratio of immigrants to population is around 6 percent on average, the ratio of imports to GDP is four and a half times this at 27.5 percent (Hatton, 2007). This suggests enormous potential for further increases in migration.
Freeman (2006a) also compares the relative importance of people, trade and capital flows in globalisation by looking at several different measures. There is no simple way of comparing the flows. Nevertheless, international merchandise trade as a fraction of the global sales of goods, and international capital flows as a fraction of the global capital market are much more significant than immigrants are as a proportion of the global labour market. Another measure is that of price or wage dispersion. Wages of similar occupations around the world are much more dispersed than prices of goods and the cost of capital. Freeman therefore concludes that the least liberalised part in the global economy is the labour market.
In this section, we examine the interactions between trade and investment flows and migration. Some of the movement of skilled labour between countries is closely aligned with the movement of capital and trade. Given the relatively mild regulation of the temporary movement of people, most of the cross border people flows tend to be of a temporary nature (Iredale, 2000). Migration that is intended to be long term or permanent, faces in most cases relatively strong restrictions. Particularly strong restrictions tend to be placed on unskilled workers, family-linked migration and the admission of people on humanitarian grounds. Both Iredale (2000) and Freeman (2006a) argue that there will be a loosening of such restrictions when developed economies with ageing populations start to realise the benefits of opening the border to labour flows. On the other hand, concerns about diminishing social cohesion may lead to tighter restrictions.
With increasing trade in goods internationally, the relocation of people may be necessary to facilitate transactions. Without more careful coordination of trade and migration policies, firms will be 'harder and harder pressed to meet their increasing need for more permeable borders that allow the flexible and fleet movement of high-skilled employees' (Keely, 2003). Both Harris (1998) and Kikuchi (2003) argue that communication networks, such as the internet, are important in facilitating business services trade, including the services of skilled labour.
Goods trade and people movement: Substitutes or complements?
In models of international trade it is argued that
'if country A has more labour relative to capital than country B, it can send labour to country B directly though immigration or indirectly through the export of labour-intensive goods. Restrict immigration, and trade [of goods] should increase. Restrict trade, and immigration should increase' (Freeman, 2006a, p.160).
This then argues for the substitution case. In other words: movements of goods, or factors of production other than labour, may be able to substitute for the movement of people (Luterbacher and Theler, 1994; Feenstra, 1998; Freeman, 2006a).
Collins et al. (1997) argue that if a quota is set on the number of immigrants this will create an incentive for trade in goods to increase, causing more competition with local suppliers in the quota-setting country. On the other hand, if tariffs are significant, then it is likely that countries will receive more pressure from potential immigrants to allow access to their labour markets. This is because the immigrants' home countries' exports will be lower than otherwise, causing lower demand for labour used to produce such export commodities. In addition, prices of goods and wages tend to be relatively high in developed countries with protected domestic industries, and the gains to immigrants may be significant if they can gain access to these labour markets.
Apart from substituting finished goods for labour movement, it is also common for countries to demand low-skill labour intensive intermediate goods from abroad. This will result in a relative fall in demand for low-skilled workers in the importing country. On the other hand, unskilled labour demand will then rise in the exporting economies. Activities which need large amounts of low-skilled labour may be outsourced to low wage economies, given the constraints of moving people between countries. For example, in the case of the well-known Barbie doll, materials such as plastics and hair are retrieved from Taiwan and Japan, with the assembly undertaken mainly in the low-wage labour markets of Indonesia, Malaysia and China (Tempest, 1996; Hatton and Williamson, 2005).
Outsourcing can lead to concerns about the hollowing out of domestic industries, with large shifts in the location of production to abroad. This has been a particularly prominent issue in countries such as Japan (e.g., Cowling and Tomlinson, 2001). In recent years, there has been rapid growth in outsourcing of services (e.g. call centres) where the main benefit is not the abundance of unskilled labour in the developing economy, but the wage differential between developing and developed countries in the cost of skilled labour. A major concern that countries seem to have with outsourcing is the loss of jobs for local citizens, although some argue that the quality of services can also be affected. From a purely economic perspective, the benefits will likely outweigh the costs, with production shifting according to comparative advantage. Gregory Mankiw, a very well-known and respected economist, has stated that 'outsourcing is just a new way of doing international trade' (Drezner, 2004).
Outsourcing of services has become increasingly possible due to new technologies, the lowering of communication costs and the compatibility of software packages. Some of the main business functions or activities that are outsourced include: customer service, telemarketing, and document management. In addition, professional services in medical transcription, tax preparation, and finance are also outsourced (Drezner, 2004). The literature in this area generally concludes that outsourcing leads to a redistribution of job opportunities. While lower-wage jobs are lost due to outsourcing, they are replaced by higher-wage jobs in the local economy (Bardhan and Kroll, 2003; Drezner, 2004; Bhagwati et al. 2004; Yomogida and Zhao, 2005).
There are some concerns that large numbers of the services jobs may be lost to countries such as India (e.g., Ganguly, 2005). However, Bhagwati et al. (2004) argue that this fear is unfounded. For example, around 70 percent of jobs in services in the US require both the consumer and supplier to physically be in the same place. This means that these jobs cannot be outsourced. Examples include retailing, catering, restaurants, hotels, tourism, and personal care. Hufbauer and Stephenson (2007) agree that 'available evidence suggests that widespread fears of massive job losses in industrial countries due to outsourcing are overblown'.
Yomogida and Zhao (2005) note that outsourcing of services and immigration are 'two sides of the same story' in that when a firm cannot hire cheap labour through immigration, outsourcing will be undertaken to substitute for this. In most countries, outsourcing at the firm level faces fewer restrictions than immigration. This raises an interesting issue for research as the net benefits of outsourcing as compared with immigration have yet to be assessed. Meta-analysis of the available evidence suggests that immigration has very little impact on wages and employment opportunities of native workers (Longhi et al. 2005a; 2005b). This is partially due to the consumption of locally produced goods and services by immigrants and through faster capital accumulation. With outsourcing, short-run domestic demand may decrease and wage and employment opportunities of local workers may be more negatively affected until long-run adjustments have been made.
Jain et al. (2006) consider the dilemma that governments of developed countries face: restricting immigration may encourage firms to outsource services abroad, but permitting immigration may also lead to distributional and social costs. Jain et al. (2006) argue that there is more political resistance to the outsourcing of service sector jobs than to the loss of manufacturing jobs (e.g. due to trade liberalisation) because of the general purpose nature of information technology that is used in services. This implies that outsourcing of services could potentially affect a greater share of the labour force than an increased vulnerability of manufacturing jobs.
International production factor movements, such as migration and capital flows are not always substitutes for trade. Markusen (1983) defines the flows as complements when trade increases as international factor mobility increases, and substitutes when trade falls due to an increase in factor mobility. The migrant labour input into production may be complementary to the locally born workers in a number of different ways. For example, as Freeman (2006a) points out, Mexican low-skilled workers have on average three years less schooling than low-skilled American workers, which may be a reason for the disproportionate representation of Mexicans in certain job areas. Furthermore, Jones (2005) notes that immigrant labour may differ in some unmeasured ways from local workers. One possible way may be a willingness to work in jobs that locals perceive as undesirable. For example, the seafood industry in Thailand makes use of workers from Myanmar willing to undertake unskilled work such as opening shellfish. This unskilled job is increasingly unattractive to Thai workers who have other opportunities. Therefore, immigrants and local workers may be complements in production. If the inflow of migrants is large enough, migrant inflows may lead to an increase in the marginal product of locals. Jones (2005) concluded that under certain circumstances both immigration and outsourcing of labour intensive production may lead to higher wages.
While complementarity between goods and factor movements is possible, there are economic theories that equally well suggest that migration and trade can be substitutes (Schiff, 2006; Carbaugh, 2007). The literature on this topic, reviewed in for example Nana and Poot (1996), concludes that in theory, trade and factor movements can be either complements or substitutes. The more realistically the theoretical models are formulated, the harder it becomes to draw firm conclusions. For example, Felbermayr and Kohler (2006) design a trade and migration general equilibrium model with three skill groups of labour and three types of outputs. However, analytical solutions in these types of models are quite messy. It is clear, therefore, that the substitution/complementarity question requires empirical testing (Bowen and Wu, 2005).
Bowen and Wu (2005) suggest that the literature in this area has neglected to look at how the type of immigrant employment may affect the issue of whether trade and immigration are complements or substitutes. Many immigrants work in sectors which produce non-traded goods, and some immigrants may face barriers in moving across sectors in the host country. Furthermore, a growing portion of industrial countries' economies consist of the activities of non-traded services sectors. In this case immigration leads to a larger domestic economy, more income and therefore a higher demand for imports. In response to these issues, Bowen and Wu (2005) develop a simple model for an open economy in which two internationally traded goods and one non-traded good are produced. The authors' model predictions, using a panel of OECD countries, show that trade and immigration are complements. The model shows that the larger the fraction of new immigrants employed in the non-traded sector, the more likely it is that trade and immigration are complements. However, the complementary relationship can be reduced, and may even be reversed, by some immigration policies. For example, guest-worker programs which recruit immigrants who substitute for domestic workers in traded goods sectors may lead to fewer imports from the countries that send the migrants. The Bowen and Wu (2005) model also suggests that it matters from which country an immigrant originates. When immigrants come from countries with language or culture common to the host nation, they are able to integrate more easily into the local labour force. If, in addition, they have the skills necessary to work in the traded goods sectors, this will lower the positive effect of their presence on the non-traded goods sector and cause trade and immigration to become substitutes. In a recent paper, which also addresses the issue of substitutes and complements, Schiff (2006) analyses twenty different scenarios that show possibilities for both.
In summary, we find an interesting paradox. As has also been elucidated formally through economic theory by Borjas (1999a), the economic benefits to a country from immigration are the greatest when migrants and the locally-born are the least similar. However, this is precisely the case when the social costs of immigration are the largest (e.g., Schiff, 2000). The argument is that people establish attachments more easily with those who have similar customs, values, language, history and culture. For this reason migration can be thought of as creating negative externalities, or unintended adverse impacts, in terms of social cohesion. On the other hand, diversity may have positive externalities if it creates a more vibrant and entrepreneurial society (Poot, 2008). It is thought that migration can not be optimal from both the host and sending countries' point of view unless the negative externalities are somehow internalised (Schiff, 2000).
In the New Zealand and Australia context, Nana and Poot (1996) evaluated the impact on the two countries of lowering trade barriers with other (that is, third) countries. The study used a two-country multi-sectoral computable general equilibrium (CGE) model. The model takes account of the fact that production in Australia is on average more capital intensive than in New Zealand and this leads to higher real wages in Australia and a tendency for net migration from New Zealand to Australia. It is found that when trade barriers are removed with respect to third countries, microeconomic adjustments will be made that expand production in those sectors where the respective countries have a comparative advantage. Australia then demands relatively more capital and New Zealand relatively more labour. While this could lead to some return migration of New Zealanders from Australia, professional workers and capital are complements in production. Trade liberalisation could then lead to a 'brain drain' from New Zealand to Australia. Nevertheless, the impact of these adjustments on the movement of people between Australia and New Zealand is possibly quite small and may be offset through the impact of macroeconomic policies or changes in the business cycle.
Trade facilitation is currently a key area of interest in international trade. Originally, trade facilitation tended to be narrowly focused on the logistics of moving goods through ports or customs. However, the definition has broadened over time and can now cover the entire environment in which trade takes place (Wilson, 2003). This includes transparency in regulatory authorities, harmonisation of standards and conformance to international norms, as well as the available technology. It is also referred to as business facilitation or, in the popular jargon, as 'cutting red tape' (Woo and Wilson, 2000; Kleitz, 2002; McMaster and Nowak, 2006). Trading procedures that are inefficient are a discriminatory tax on economic activity between different countries, resulting in a deadweight loss. Making such procedures more efficient, or removing such procedures entirely, can result in a positive-sum game (Kleitz, 2002).
A study by Kim and Park (2005) looked at four trade facilitation indices: customs procedures; standards and conformity; business mobility; and information and communication technology. It was found that each index has a positive and significant effect on trade between three Northeast Asian countries, namely China, Korea and Japan. For this reason, the authors recommend that free trade agreements should stress trade facilitation, rather than just following the standard tariff reduction agenda. Woo and Wilson (2000) note that trade facilitation among APEC countries may yield an extra 0.25 percent of real GDP by 2010 (Woo and Wilson, 2000). Although this may seem a small percentage, trade facilitation can also result in an overall improvement in the world trade environment (Kim and Park, 2005).
Trade facilitation involves the recognition and reduction of trade barriers, including barriers due to a range of different standards and requirements. It can be argued that trade facilitation is a substitute for migration: when administrative trade barriers are high, skilled migrants residing in the importing country can act as agents to deal with 'red tape'. High transaction costs due to technical trade barriers are likely to particularly impact on developing country exporters who may not have the skills and resources to deal with these barriers (Saner and Fasel, 2003). When the barriers are removed, both exports and imports between host and migrant source countries will increase but the need for these types of services provided by expatriates is diminished. The presence of an expatriate population is likely to lead to greater imports 'from home' to that host country in any case. This issue will be explored further in Section 4.1.
There is increasing recognition of the importance of liberalising the services sector, which often comprises a very significant part of a country's economy and continues to face high barriers to international trade. The introduction of GATS in 1995, as discussed in Section 2.1 of this report, is clear evidence of this. Some estimates suggest that trade liberalisation of the service sector would offer much more significant gains than liberalisation of international goods trade (e.g., Dee and Hanslow, 2000). Services trade has become an important engine of world growth, and despite the large barriers faced, it has grown faster than world merchandise trade over the past two decades (Hufbauer and Stephenson, 2007).
Negotiation issues in the services sector have particularly significant implications for the movement of labour. Karsenty (2000) attempts to decompose the services trade volume into the four GATS modes of supply. Using 1997 data, Karsenty finds that 41 percent can be attributed to mode 1, 19.8 percent to mode 2, 37.8 percent to mode 2 and only 1.4 percent to mode 4. While these are 'very rough estimates' the temporary movement people under mode 4 is clearly a very small contributor to the overall value of services trade. This relatively low level of mode 4 services trade is due to the many barriers in place and there are potentially large gains to be made from liberalising people movement (Winters et al., 2003). As Iredale (2000) explains '[g]lobal labour markets now exist in a range of occupations, and a person's skills are their greatest assets to be bought and sold'.
GATS mode 4 is concerned with the movement of 'natural' persons for the purpose of providing services, therefore viewing the restricted movement of people as a barrier in exporting services rather than an issue of immigration (Walmsley and Winters, 2005). There exists quite an extensive literature on GATS mode 4, as introduced in Section 2.1. Much of the literature discusses the lack of progress under GATS, with Ng and Whalley (2005) going so far as to suggest that a new body should be specifically created for global negotiations on visas and work permits (see also Pasquetti, 2006). The overall conclusion of the literature on services and the movement of people is that there has been very little liberalisation in this area and that there are potentially many gains to be made. Trade in services cannot prosper if the movement of people is not promoted (Walmsley and Winters, 2005; Chaudhuri, et al., 2004; Ghosh, 2005).
While trade in services using mode 4 is not very wide-spread at this stage, and most countries stand to gain from freeing up the movement of natural persons, the challenge is to negotiate how the movement of natural persons should occur. Winters (2003b) notes that practical implementation of mode 4 liberalisation is very sensitive and it will be difficult 'to convince immigration officials that mode 4 does not undermine border integrity, or labour officials that mode 4 does not undermine labour law or local job markets.' Mattoo (2003) classifies the main barriers to trade in services under mode 4 into three broad categories: immigration issues, particularly visa-related barriers; discriminatory treatment of foreign providers of services; and inadequate recognition of qualifications. Winters (2003b) suggests four specific procedural issues that need to be resolved: social security contributions; the classification of occupations; the recognition of qualifications and certification and the codification of economic needs tests. He argues that countries need to carefully consider how to balance the economic gain with political sensitivities.
While GATS is largely concerned with the movement of skilled labour, Walmsley and Winters (2005) argue that freeing up the movement of unskilled labour offers even larger potential gains. Their very innovative study employs a version of the well-known Global Trade Analysis Project (GTAP) general equilibrium trade model. Results from the Walmsley and Winters study suggest that even a small quota increase of 3 percent for both skilled and unskilled workers in developed countries could increase world welfare by US$156 billion. Only $46 billion of this is due to increasing the international mobility of skilled workers, whereas increasing the international mobility of unskilled workers leads to a welfare gain of $110 billion. They compare the assumed small increase in immigration quotas of 3 percent to removing all the remaining restrictions on goods trade. The latter is estimated to increase global welfare by US$104 billion. This leads to the conclusion that benefits from increasing efforts in liberalising both skilled and unskilled labour movements may be significantly larger than the benefits that can be realised from goods trade liberalisation. This conclusion is consistent with other studies summarised by Hatton (2007).
World Bank modelling suggests significant economic gains for migrants themselves, as well as for origin and destination countries when there is increased migration from developing to high-income countries (World Bank, 2006). This research calculates gains that are more than double those suggested by Walmsley and Winters (2005). The World Bank estimate is supported by more recent estimates by Walmsley et al. (2005b) using a dynamic version of the GTAP model and an updated database. Walmsley et al. (2005b) conclude that their earlier work may well have significantly underestimated the gains from migration liberalisation.
Although allowing unskilled workers into host countries will potentially lead to the most significant economic gains, most legally admitted immigrants are relatively highly skilled. For example, immigrants in Australia are twice as likely to have a university degree as locals are. In the United Kingdom, 60 percent of migrants entering the country are professionals (Freeman, 2006a). This brings about the 'brain drain' concern which will be addressed in Section 5.1 of this report.
Manning and Sidorenko (2007) examine for the ten ASEAN countries the barriers to migration of professionals, with a particular focus on the healthcare and IT sectors. The authors evaluate these regulations by putting the countries into three groups in terms of their stage of development, as defined in terms of per capita income and industrial structure. The overall finding is that within ASEAN the intensity of regulation is broadly inversely related with development, while the intensity of migration tends to be positively related to the level of economic development. The more developed countries (mainly Singapore, Brunei and Malaysia) have more liberal policies for immigration of health professionals and other high skilled migrants. Some countries in the ASEAN group have more open regimes for the movement of temporary migrants in the IT sector in comparison to healthcare, where professional bodies regulate the industry. In several countries, the restrictions on foreign professionals severely limit the access of such professionals to such potential host countries. The authors recommend 'streamlining visa and work permit regulations for professionals across the region (including short-term entry of independent service providers); improved education and professional standards (for example, using the recent APEC and ASEAN initiatives to examine nursing standards and develop a regional Mutual Recognition Agreements in nursing); overcoming the language barrier to mobility by allowing foreign-trained doctors and nurses employed in export-oriented hospitals to be exempt from the language tests for temporary registration purposes; promotion of industry self-regulation and certified training programmes in IT; and a more systematic approach to data collection on international stocks and flows of professional manpower' (Manning and Sidorenko, 2007). Multilateral trade in healthcare services is also extensively analysed in a recent book edited by Blouin et al. (2006).
Technology both affects and is affected by the movement of people. Modern communication technologies, such as email and the internet, lower the costs that migrants face to keep in touch with their relatives and friends, and help them to remain informed about their home country (e.g., McCann and Poot, 2008). This increases the proportion of the population that would contemplate migration. On the other hand, migrants are also a channel for the international transfer of technology themselves. For example, professional migrants may bring with them new processes and innovative ideas that can contribute to productivity growth (Hoekman et al., 2005). To date, however, very little systematic research as been undertaken on how immigrants affect innovation and productivity growth in specific sectors (Poot, 2008).
Because R&D expenditure in service sectors is growing faster than in goods production (Gera et al., 2005), there is growing demand for professionals at the high-technology end of service provision that may be partially met by immigration. One example is the information technology (IT) sector, discussed in the previous section, which undertakes a great deal of R&D and continues to grow very rapidly. This has resulted in the increased demand for skilled and knowledgeable immigrants in this sector (e.g., Coppel et al., 2001). However, not only highly developed countries have expanding IT sectors but some developing countries are also rapidly catching up. For example, India has put significant resources into developing software and related services. Although it took some time to reap the rewards, these have now become very large (Hoekman et al., 2005). The repatriation of systems engineers, programmers, etc. leads to knowledge transfers back to the original migrant source country.
While migrants can be a transmission channel for the international transfer of new knowledge and practices, new information and communication (ICT) technologies can also substitute for migration flows. The increased sophistication of IT hardware and software makes the e-labour concept more feasible then ever before. Internet-based services in particular represent a new frontier of services trade, with significant opportunities and challenges (Hufbauer and Stephenson, 2007). Professions such as software engineering, data entry, translation services, and distance teaching are all examples of sectors where knowledge-intensive services can be exported without 'exporting' the workers. India is the most commonly cited case, with its many call centres providing services for other countries (Gera et al., 2005). This concept of outsourcing was introduced in Section 3.2.
So far in this report, we have considered the issue of trade and immigration as one of relative production factor abundance. Developing countries have abundant labour and tend to 'specialise' in exporting either that labour or labour-intensive goods. Developed countries can similarly specialise in exporting capital or capital-intensive goods. Such patterns only hold for production sectors where the available technology is similar across countries. When countries' have superior technologies they can have a monopoly in the world market for goods and services produced with such technologies. Such a monopoly improves the terms of trade of such countries in that export prices increase relative to import prices. Davis and Weinstein (2002) argue that an influx of immigrants into a technologically superior country may remove the significance of this 'leadership' and lead to a fall in the terms of trade and a welfare loss for the migrant-receiving country (also see Findlay, 1982). Davis and Weinstein use the United States as an example and observe that technological superiority in this country coincides with inflows of all production factors, not just labour. In this case a further inflow of labour leads to welfare losses to the receiving country and gains to the sending country (also see Commander et al., 2003).
A general conclusion from this and previous sections is that there are many theoretical perspectives on the links between international migration and other forms of international interaction. Conclusions drawn from theoretical models in this area are rather sensitive to the assumptions used and the overall direction of any causal links can only be established by careful empirical analysis.
Just as in the case of international trade, investment agreements and foreign direct investment (FDI) flows will also impact on, and be impacted by, the movement of labour internationally. Gross foreign direct investment as a percentage of GDP averages 8.8 percent in the world, but the post-1990 growth has been phenomenal (Poot, 2004, Table 1.1). Globalisation and trade agreements played important roles. For example, in the NAFTA area, investment in Canada by the US more than doubled; US investment increased in Mexico fourfold; and Canada and Mexico more than doubled their investments into the US. In addition to this, integration between Canada-US and US-Mexico financial markets was enhanced, mainly due to mergers and new corporate subsidiaries across each others borders (Duran, 2003; Hufbauer and Schott, 2004).
Flows of goods, investment and labour were traditionally seen as substitutes. However, it is becoming increasingly clear that countries wishing to attract FDI need to ensure an adequate supply of professional workers (Chia, 2006). To invest across borders there is a need to have knowledge of that market, in addition to monitoring the investment activities. This creates the need to have people in the foreign markets to take up management positions that will oversee activities and make decisions. In addition, FDI may trigger temporary movement in the form of business trips and (semi)-permanent movement in the form of intra-corporate transferees. International investment flows are often followed and facilitated by skilled migrant flows (Gera et al., 2005; Freeman, 2006a). Therefore, FDI can be a complement to, as well as substitute for, the movement of people.
Guellec and Cervantes (2001) suggest that some FDI may flow to countries where there is skilled labour available, but at a globally competitive wage. Examples are Israel and India. At the same time, the presence of high-skilled immigrants may cause an inward flow of FDI as well (Ivlevs, 2006). Thus, people movement and FDI may be complementary flows in both directions. However, Gera et al. (2005) argue that if multi-national enterprises relocate facilities to another location for the purpose of gaining access to cheaper labour (the main driver of the 'hollowing out' of, e.g., the Japanese economy referred to earlier), then people movement and FDI may be substitutes.
More generally, however, trade liberalisation and foreign direct investment are strongly related. If FDI and migration are on balance complements rather than substitutes, trade liberalisation will increase the volume of migration between countries that intensify their economic integration (Gera et al., 2005). Again, the issue of substitution versus complementarity can only be resolved by careful empirical analysis. Kugler and Rapoport (2007) argue that FDI and migration substitute one another in the international matching process between workers and firms. On the other hand, migrants provide information about future investment opportunities in their country of origin. Kugler and Rapoport (2007) provide evidence from U.S. data that is consistent with contemporaneous substitutability and dynamic complementarity between migration and FDI.
Negotiation or bargaining is a form of strategic interaction that can often be better understood by so-called game-theoretic models that have become very popular in economic theory since the publication of von Neumann and Morgenstern's (1944) Theory of Games and Economic Behavior. However, this approach has had as yet surprisingly little application to the issue of how countries decide unilaterally or multilaterally about admitting foreign citizens to become residents of their own territory. An important issue in this context is the nature of the payoff function: the translation of policies and actions into the estimated net benefits to the citizens of a country. Trade and migration are in this context not seen as achieving similar goals. Greenaway and Nelson (2006) argue that trade policies are essentially about material wellbeing while immigration policies are much more complex. We have already noted earlier in this report that trade and migration policies are administered in any case by different government departments.
Luterbacher and Theler (1994) apply a game-theoretical model to analyse the decisions made by governments on restricting migration, taking into account possible tradeoffs and effects on other countries. Luterbacher and Theler consider the standard case of production by means of labour and capital inputs. They consider two stylised countries, one from the North and the other from the South. Both countries aim to maximise their own welfare. The North's welfare is positively affected by its GDP and by revenue from capital exported to the South; and negatively affected by remittances of Southern immigrants, and by these immigrants becoming too numerous (due to declining wages and diminishing social cohesion). The South is likewise positively affected by its own GDP, and also by the inflow of remittances, but negatively by the payments to capital owned by the North, by the presence of too many North-owned firms and by too much unemployment. It is intuitively clear that in this situation autarky (i.e. closed borders for capital and labour) is not optimal. There are some gains to both countries from opening up the border to some extent.
The contribution of Luterbacher and Theler (1994) is that they show in theory that it is not certain that migration and foreign direct investment will be permitted. A so-called Prisoner's Dilemma may occur, in which both countries know that they would benefit from co-operation but each country has an incentive to cheat on any joint agreement, leading to a stable outcome (referred to as the Nash equilibrium) that is not optimal for either country. This stable outcome is a closure of the border to migration and capital flows. The fact that in practice some migration and foreign direct investment is permitted can be explained in game theory by the fact that the bilateral negotiation between the North and the South does not occur just once, but is repeated over many years. In that case, an agreement to co-operate to have open borders is more likely to be stable.
Besides migration negotiations between governments of countries, there can also be game-like negotiations between the government and various lobby groups within a country. Facchini and Willmann (2005) present a 'factor protection' game, in which domestic interest groups support a democratically elected government dependent on the extent to which the government is willing to protect the interests (usually income) of that group vis-a-vis other groups. Nash equilibria can be determined for such games. Governments have in such strategic games the option to either restrict inflows (capital or immigrants) or impose tariffs (essentially imposing a tax on immigrants). Facchini and Willmann demonstrate that these policies have theoretically identical impacts.
 Another implication is of course that the rates of return to capital (and real interest rates) are higher in New Zealand than in Australia. Capital would therefore tend to flow from Australia to New Zealand.
 Welfare in this context is defined by an equivalent variation (EV) in income, with EV being the change in income that is equivalent in its effect on people to the policy change being simulated.
 FDI of a country is here measured as the sum of the absolute values of inflows and outflows, not just the inflows.
 For an accessible introduction to game theory, see for example Osborne (2004). Correa (2001) provides a non-technical introduction to the use of game theory as an instrument for the analysis of international relations.
 There is, however, some literature on how game theory can be generally used to analyse trade negotiations. See, for example, Harrison and Rutstrom (1991).
 The Prisoner’s dilemma received its name from a simple two-person game in which two prisoners who have committed a crime would both benefit from pleading innocence (in which case the case against them cannot be proven), but the attraction of plea bargaining is too large and both will succumb to the sub-optimal strategy of pleading guilty.
 This result is also shown theoretically by Luterbacher and Theler (1994).