The instability of framing

But the situation is more complex than this. Measurement and calculation do not only have anti-political effects. They do notjust have the effect of restricting political controversy in the economic field.

The fragility of metrological regimes

While the enormous efforts of the authorities to measure pollution look impressive from a distance, the whole apparatus of measurement and calculation is much more fragile than it first appears. In part this fragility is a function of the weakness of standardised metrological regimes when judged against the more complex analyses that often derive from on-going research. In the case of routine air pollution measurement, for example, from the point of view of research scientists such information is massively overproduced.¹⁷ Measurements of air pollution in the city only occur at particular points and it is very difficult to say what air pollution is at other places in the city, even a hundred yards away from the point at which a

permanent air quality monitoring device is placed.¹⁸ Certainly, it is almost impossible to establish an exact correlation between air pollution measurements and health statistics, except in extreme conditions. Greater levels of pollution certainly do contribute to bad health but it is very difficult to say how much (Barry 2001: 170). Moreover, despite the efforts of garage mechanics to measure the levels of pollution from car exhausts, such measurements may actually have surprisingly little relation to the performance of cars on the road. In the mechanic’s garage, the composition of the exhaust fumes from the car is measured when the engine is idling.¹⁹ On the road, however, the exhaust is likely to contain the highest concentrations of pollutants when the car is starting, or in a traffic jam, or accelerating rapidly, and there is no straightforward relation between such concentrations and the concentrations of common pollutants as measured in the garage.

Certainly, routine monitoring and measurement activities often have anti­political effects. Governments, environmental organisations, laboratories and firms may all apply themselves to the task of increasing the quantity and enhancing the accuracy of environmental data. Attention is devoted to improving the practice of monitoring and testing. Garage mechanics are required to do further training to make sure that their measurements are comparable with each other. But this vast metrological regime is much more fragile than one might imagine. The situation has been framed - in the sense that a whole series of other questions about urban pollution and urban politics have been displaced (Rydin 1998) but the organisation of any regime is always open to the possibility of contestation. There is an agreement, whether consensual or forced, to accept the truth of measurements and the legitimacy of the regulatory practices with which they are associated, but this agreement is not grounded in science, but in the much more uncertain procedures of metrological practice (Latour 1999a: 258-9).

In these circumstances the work of scientists can have political rather than anti­political effects. For scientific work can identify the weaknesses in this vast exercise in routine monitoring and measurement. Potentially at least, far from restricting the space of contestation, further scientific calculations may serve to open it up. They may reveal the limitations of all the investments of governments and

The anti-political economy 91 environmental organisations in monitoring pollution levels. Indeed, in the case of air quality measurement it is relatively easy for research scientists to demonstrate the weaknesses of existing metrological regimes.²³

To be sure, there is no general political crisis about how car exhaust fumes are measured or how air quality is monitored on the street. Consumers simply accept the validity of the test, and environmentalists press for further increases in the level of monitoring. But in other cases the weaknesses of metrological regimes are revealed. Consider, in particular, the case of BSE. Despite the huge investments in the inspection of farms and abattoirs, for example, the limitations of such inspection regimes have become apparent. After many years, there is still no commonly accepted method for ensuring that beef is safe to eat across Europe.

Given that routine metrology is so weak to further interrogation, how do metrological regimes survive? There are two possible options. One option is to protect the fragility of metrological regimes. Measurements and assessments are routinely conducted, by abattoir inspectors, garage mechanics, structural engineers, financial analysts, auditors and so on, but the form and content of their assessments are rarely subject to wider scrutiny and debate. To be sure, the ways in which the fragility of metrological regimes are protected are various. They depend on the use of a range of anti-political devices. An explanation of why it took so long for BSE to be recognised might include, for example, discussion of the particular culture of scientific advice in Britain, the existence of a culture of secrecy in government, the role of the farming lobby and a failure to adopt a precautionary policy in the context of scientific uncertainty.²⁴ In the other instances, the fragility of metrological regimes relies on the fact that metrology itself is given a market value.

Serco has a particularly strong culture that embodies people, ethics and an ethos that enables us to work easily with public sector bodies. We adopt a stance towards customer relationships that can be reconciled with the provision of public services through the private sector. This is not an artificial stance that we have taken with a view to a particular market - it is a culture that has become embedded because it describes the organisation we wish to be.²⁵

This contractual arrangement - amongst other factors - has made it difficult for scientists to raise questions about the reliability of metrological regimes even if they so wanted. Once their knowledge is sold to government, they may simply be unaware of how their knowledge is represented in public by government officials and ministers.²⁶ At the same time, because they depend on government for business,

laboratories that sell their services may be less likely to produce conclusions that criticise government policy

The second option is to open up metrological regimes to greater scrutiny and to acknowledge the weaknesses of measurement and the uncertainties of economic and scientific calculations. In the wake of the BSE crisis, amongst other problems, there are some indications of movements towards this option.²⁷ In Britain and France there have been calls for demands for more constructive debates between experts and publics. Some writers have framed the need for a greater level of dialogue between experts and publics in terms of the notion of risk.²⁸ “When society has problems with science, it is often over questions of uncertainty and risk” (House of Lords 2000: 7).

The inventiveness of measurement

Metrology puts new objects into circulation. It multiplies realities by creating objects that can be regarded neither as reflections of reality, nor the expressions of the social subjects who created them. Reality is not a blank screen onto which social categories can be projected. Metrology creates new objects that make a difference in the world. When it is presented as information, measurements do not merely inform - they make demands on those who should be informed (Strathern 1999, 2000). In so far as it is treated as the source of information, metrology has performative and regulative consequences.

Consider the importance of measurements of, and experiments on, metal fatigue in rail tracks following the rail crash at Hatfield, on the main line north of London, in October 2000.³⁰ At Hatfield, a train travelling from London King’s Cross to Leeds derailed while travelling at 115 miles per hour. It soon became clear that the accident had been caused by a broken rail. According to the owners of the track, Railtrack plc, within hours Railtrack engineers had “identified some significant deterioration in the condition of the rail”.³¹ Later it was stated in public that this was likely to have been caused by “gauge corner cracking”, itself the result of metal fatigue. Despite reports inJanuary 2000 by engineers working for the Railtrack sub-contractor, Balfour Beatty, that the Hatfield track should have been re-railed, the track had not been re-railed, nor had speed restrictions been imposed in order to reduce the stress on the weakened track.

The anti-political economy 93 conditions of track that had been made were not systematic, coordinated, nor acted upon. The Hatfield crash was not an accident.³²

To be sure, metal fatigue in railway track is difficult to detect, to investigate and to analyse. First, laboratory simulations are extraordinarily inexact. Tracks in laboratories can never be subjected to the same range of temperature variations, surface contamination, and stress resulting from complex variations in train speed and load, which they will be subjected to in commercial use. The level of track lubrication, which varies considerably with weather conditions and the form of track maintenance, significantly affects crack growth, for poorly understood reasons.³³ Second, it may be difficult to locate or detect cracks caused by metal fatigue on rails in use. Indeed, according to the UK government Health and Safety Executive, the ultrasound equipment commonly used to detect cracks on railways may not have been able to detect the kind of cracks found in the rails at Hatfield. The Hatfield cracks may have been too slanted from the vertical:

The [ultrasound] techniques [used on British railways] allow for rail to be classified as ‘untestable’ under certain circumstances, in particular, when metal has been lost from the rail surface by a process known as ‘spalling’ which appears to have been the case at Hatfield. Laboratory examination has revealed that the transverse fatigue cracks present in the Hatfield rail were located in the angular range 20°-35° from the vertical. Hence, had the rail been testable it is possible that all the fatigue cracks would not have been detected.

(HSE 2001: 13)

But, following a crash, however difficult they are to interpret, and however inexact and contested earlier observations and measurements of rail track fatigue had been, they have to be taken into account. They demand a response on the part of those who are in receipt of such information. They have an immediate regulative effect. “Railtrack acted swiftly to accept responsibility for the terrible accident at Hatfield. Railtrack’s actions since the accident have been taken in the light of the lessons learned from the initial investigation and the engineering understanding of gauge corner cracking.”³⁴ Given the uncertainties in the science of metal fatigue and the evident weaknesses of its existing metrological regime, Railtrack responded with extreme caution. Speed restrictions were immediately imposed throughout the UK railway network, causing chaos as it was no longer possible to keep to existing timetables.

Sociologists have sometimes wanted to find political or sectional interests embedded in the calculations of experts (Callon and Latour 1992). But the case of the rail track suggests a different conclusion. In this case, the observations of the Health and Safety Executive and the calculations of experts about fatigue in rail tracks had political effects precisely because such calculations are not reducible to politics. Metallurgy and mechanical engineering - the most material of disciplines, which contain no obvious traces of political impurity - proved to be the most profoundly political in their effects.³⁵ Because they were not associated with any particular political doctrine, the mechanics of metal fatigue raised questions about

the viability of a model of economic organisation based on the regulation of private monopolies. Far from having anti-political effects, the calculations of engineers had political resonances.³⁶ They flooded across the political field. Demonstrations of the state of rail tracks were political demonstrations of a kind, but they were not ones that could be associated with any identifiable political actors. As a result, they were much more difficult to police than demonstrations conducted in the street. Fatigue cannot be explained away as the expression of particular political interests. It exists, as Whitehead would say, as a stubborn fact.³⁷

In Isabelle Stengers' account, an event can be understood as a creator of difference. It is in the middle of a field of effects, which does not mean that it is the cause of such effects. An event has a factual existence, yet this existence does not pre-determine the response to it, but creates the necessity of a response (Stengers 1997: 215-16). The force of an event is itself in question. It is resisted only when its existence is ignored:

The scope of the event is part of its effects, of the problem posed in the future it creates. Its measure is the object of multiple interpretations, but it can also be measured by the very multiplicity of these interpretations: all those who, in one way or another, refer to it or invent a way of using it to construct their own position, become part of the event's effect. In other words, every reading - even a reading that denounces the event as a fake - still situates the one who proposes the reading as an heir, as belonging to the future whose creation the event contributed to.

(Stengers 2000: 67-8)

In this sense, the failure of the track at Hatfield turned out to be a political event. It became the centre of actions, questions and responses on the part of a vast range of agencies that were compelled to respond to the fact of the crash. The stubborn fact of metal fatigue demanded a response, yet the course of this response could not be predicted. This response was not merely technical. However briefly, it raised questions concerning the relations between the organisation of markets, the management of companies, the role of engineers and the performance of technology. It served to reveal the inadequacies of the particular form of market organisation, which had been established through the privatisation of the railways. For a few months, ‘Hatfield' came to occupy the centre of field of political activity that went far beyond the local problems of repairing a particular stretch of track.