Robots may not want your job – they like to watch

Paul Krugman says the decline in truck drivers’ wages is “not a technology story … robot truck drivers are a possible future, but not here yet … the obvious thing: unions.”

Certainly, the collapse of union power in trucking had a lot to do with the collapse of wages. But that does not mean that technology was not a factor. In trucking, technology has done little to change the hours of work, or the level of skills, required to deliver a load. But technology has improved management surveillance of truck drivers. By standard efficiency wage models of the labour market, this improved surveillance technology (“monitoring”) would cut wages even without any change in collective bargaining; there is also good reason to believe it would undermine the voice function of unions. Improved surveillance technology is thus an example of what Peter Skott and I call power-biased technological change (PBTC) – technological change which affects the balance of bargaining power between employers and workers.

Starting in the late 1980s, it became common to equip trucks with data recorders, which keep track of the truck’s speed at all times (Figure 1). By the 2000s it was common for these recorders to have satellite feeds, either batch or real time, and to combine with GPS location of the truck. Most deliveries are now tracked electronically as well – giving managers the time each delivery is signed for.

Figure 1

Such improved surveillance is not restricted to truck driving. Consider retail clerks, whose wages, like those of truck drivers, collapsed starting in the late 1970s, hitting bottom in the 1990s and never recovering (Figure 2). In the mid 1970s, supermarkets started installing cash registers equipped with bar code scanners. Adoption was rapid, becoming near total in retailing by the 1990s. The cash register is a worker surveillance device from the late 19th century – its use ensures that cash collected from customers equals cash turned over to the employer. Entering prices by scanning takes surveillance one step further: by taking control of price entry away from the clerk, it prevents “under-rings” which could (accidentally?) give substantial discounts to the clerk’s friends.


Figure 2

The association of falling wages with improved surveillance in such cases could be coincidental, and I would certainly never argue that it is a complete explanation for changes in the distribution of earnings. But that improved surveillance would result in a fall in wages is a straightforward prediction of efficiency wage theory, or indeed any agency-based theory of earnings determination.

And it does not apply just to these two occupations: we are surrounded by such applications of information and communications technology. Much of it is used to spy on (“monitor”) us. Some applications further undermine employee power by allowing managers to bypass any worker or group of workers who try to hold up production for bargaining purposes. For instance, customer service jobs which are no longer delivered face-to-face at scattered locations (e.g. bank branches) but from phone centers; not only are they more closely monitored (“this call may be monitored for quality assurance…”), they are also part of a system where the queue of customers can be redirected to any employee at any location – no single employee, not even any local group, is in a position to hold up production. The replacement of paper with electronic records makes it possible for information flows to bypass any troublesome employees – individual or group – who have record-keeping responsibilities, while flexible manufacturing systems and improved transportation facilitate alternative flow paths for physical products. Such changes as these seriously undermine the bargaining power of workers.

Which was it – technology, or institutional changes such as the decimation of unions? Our answer is “both”, and also that changes in technology have more than a little to do with the weakening of unions.


Peter Skott and I have developed this line of argument in a series of papers.

Skott, P. & F. Guy (2007) A Model of Power-Biased Technological Change. Economics Letters, 95 124-131 (paywall, or access via a university library). This is the shortest and most technical version of the argument. Our theory (power-biased technological change, PBTC) and the hegemonic theory (skill-biased technological change, or SBTC) are both stories about why the introduction of new information and communication technologies (ICTs) should lead to greater inequality of earnings (income from work). PBTC, which sees ICT as a set of monitoring technologies, predicts that work intensity – effort – should increase even as low-end wages fall; SBTC makes no such prediction; Francis Green and Steven McIntosh’s research finds that work intensity has risen with the adoption of new ICT, consistent with the PBTC prediction.

Guy, F. & P. Skott (2008) Information and Communications Technologies, Coordination and Control, and the Distribution of Income. Journal of Income Distribution, 17, 71-92 (PDF). This paper does have the equations, but also has a (relatively) plain English version of the argument. Simulations in this paper show that monitoring technologies will not only increase earnings inequality, and increase work effort even when wages and unemployment are low, but will also raise the profit share – again, consistent with real-world outcomes since the ICT revolution.

Guy, F. & P. Skott. 2008. Power, productivity, and profits. In Power, Freedom, and Voting, eds. M. Braham & F. Steffen, 374-392. Heidelberg: Springer. In this paper we explore the implications of our model for technology choice and efficiency. There is a long tradition – Marglin, Stone, Bowles, Gintis, among others – of arguments, theoretical and historical, that the most profitable technologies – and thus the ones adopted by employers – may not be the most efficient (that is, may produce less output for the same real (labour, capital & material) inputs. This can happen when the less efficient technology reduces worker bargaining power. We show the conditions under which the availability of improved monitoring technologies will lead to investment in technologies that are more profitable but less efficient. This paper has often be mentioned on the excellent Stumbling and Mumbling blog.

Guy, F. 2014. Technological Change, Bargaining Power, and Wages. In Our Work Here is Done: Visions of a Robot Economy, ed. S. Westlake. London: NESTA. A brief summary of the argument – all words, no equations – in a book where the rest of the papers are all about whether or not robots will take our jobs. My point: even if they don’t take your job, they will help your boss boss you. Entire book free to download.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s