What difference does distance make?

'Offshoring' - the process of hiving off jobs to lower-cost economies - began in manufacturing, as multinationals shipped the less sophisticated parts of their production-process abroad. But as the IT revolution has slashed the costs of long-distance communication, it has also become commonplace for services to be performed thousands of miles away from a company's HQ. A firm in Basingstoke can have a back-office in Bangalore.

Thomas Friedman, in the bestseller The World is Flat, enthusiastically called this process 'globalisation 2.0'; but the idea of wholesale offshoring of services jobs causes widespread alarm in the world's richest countries.

In a new paper, CEPR Researcher Thierry Mayer and co-authors Keith Head and John Ries attempt to assess the risk that thousands of services jobs are set to disappear overseas. They use a concept of economic 'distance', previously applied to manufacturing, to quantify the costs to firms of having services performed thousands of miles away, instead of on-site.

They begin by constructing an economic model of an international market for services, where each job in the sector is filled by the person who costs the least - but crucially, this includes not just wages, but the costs of 'delivering' a unit of the service.

In manufacturing, these costs might include the tariffs applied to exports and imports by governments, as well as the price of transporting a product to its destination market.

The authors construct an equation that they can use to test what factors contribute to 'distance' in an international market for services. They suggest that physical distance itself will create 'delivery costs'; but so will differences in time zones, languages, and the legal system.

Offshoring services functions to a country with a different language is likely to imply translation and training costs, for example; while lack of a shared legal system may make it harder to create and enforce the complex, binding contracts essential to making offshoring work successfully.

The authors also want to test the impact of shared colonial origins on 'distance'. The legacy of colonialism can be lasting commercial and cultural links, which might help to overcome some of the problems of geographical distance.

They suggest that the impact of offshoring services to a different time-zone could in theory be either positive or negative. It might cause problems for colleagues struggling to co-ordinate with each other, because their working days barely overlap - what the authors call the 'synchronisation effect.' On the other hand, stretching across time-zones allows a business to operate twenty-four hours a day, seven days a week: the 'continuity effect.'

To establish the strength of these various factors that could make up 'distance' in the global services market, the authors carry out a regression analysis, testing the impact each factor has on the volume of trade in services.

To do this, they use data collected by eurostat, the European statistics agency, covering bilateral services trade between 64 countries - including EU members, the US, and Japan - over the period 1992 to 2004. The eurostat data allows the authors to disaggregate different services industries, so that they can focus on the particular areas where offshoring has become more common, including finance and computing.

Since the mid-1980s, both goods and services have been traded more widely, so that the value of exports and imports as a proportion of economic output has increased. But services trade has been expanding even more rapidly than trade in goods.

The composition of services trade has also changed. Over the past two decades, the category called 'computer, communication and other services' has rapidly become more important, accounting for more than 40% of total services trade by 2004; while the more traditional transport and government services sectors have declined in importance.

By performing the regression analysis, the authors find that - as they expected - colonial relationships, a common language, and a shared legal system all have significant, positive effects on services trade, helping to lessen the impact of geographical distance alone. As far as stretching across time-zones is concerned, their results suggest that the 'continuity effect' of spanning the globe outweighs the problems created by the 'synchronisation effect' of getting people to work together at different times of day and night.

Because they have more than a decade's worth of data, the authors are able to calculate the size of the 'distance effect' in 1992, and measure how it has changed. They show that the distance effect has declined markedly over time, supporting the idea that falling telecoms costs have helped to reduce the need for costly face-to-face meetings - and made geographical distance matter less.

Distance effects for the services sector are now similar to those for the trade in goods, according to the authors' analysis; but while distance effects for goods have if anything been increasing slightly over time, in services they are on a declining trend.

In order to give a better idea of how important proximity is for services - and thus how powerful the threat of offshoring may be - the authors then use their findings to calculate what premium businesses would be prepared to pay to hire someone locally, rather than thousands of miles away.

Assuming the productivity of home and offshored workers is similar, they calculate the 'wedge' in wages that an employer would be prepared to pay, to get a service performed locally. They find that a London-based firm would be willing to pay between 30% and 80% more in wages for a services supplier based 83 miles away in Oxford, rather than in Dublin; and a 53% to 163% wage-premium for a service provided in Dublin, rather than Bangalore.

And workers in Oxford could be paid something between 99% and 373% more than workers in Bangalore, and still be attractive to a London firm. In other words, employers are prepared to pay close to five times as much, to get a service provided locally (within 100km), rather than remotely (more than 10,000k away).

These findings do provide some comfort for those fearing that in a hi-tech, globalised world economy, distance no longer matters and in the services trade is effectively costless. The authors are able to show that there is a clear 'distance effect', even for sectors such as IT, which seem most prone to large-scale offshoring. However, there is also evidence that in services, distance effects are waning - and the world is getting smaller.

DP6542 How Remote is the Offshoring Threat?

Keith Head, Thierry Mayer and John Ries