George Yarrow is an Emeritus Fellow of Hertford College, Oxford, and was a co-founder of the Regulatory Policy Institute.
In the context of the decision to close the Tavistock Clinic, Kemi Badenoch wrote recently about her personal experiences as a Minister in dealing with the Civil Service. It was the latest in a longish string of expressions of dissatisfaction with this relationship. Tony Blair complained about “the scars on my back” from his attempts at public sector reform; John Reid, when Home Secretary, told MPs that the Home Office’s Immigration Directorate was “not fit for purpose”; and in his Speaker’s Lecture of 2017, Francis Maude identified a number of system-wide failings in Civil Service conduct and performance, the general message being that ‘things ain’t what they used to be’.
The issues raised are of utmost importance because of the centrality of the relationships between Ministers and civil servants in the development and implementation of public policies. Major failures in this process, which appear to have been increasing in frequency in recent decades, can place very heavy burdens on the public, particularly in economic policy areas (the focus of the remarks that follow).
To address the challenges posed by institutional failures, it’s advisable to start by examining some of the root causes of dysfunctions. I will summarise four.
First, and considered as a whole, the Civil Service is an entrenched, hegemonic institution. It enjoys monopolistic powers over its domain and, as Adam Smith put it, monopoly “is a great enemy of good management”. Competitive pressures motivate sustained search to discover better ways of doing things, whatever the activity, be it science, sport, or business. When it comes to the Civil Service, this source of pressure to raise performance standards is very weak: organisational inefficiency and limited of drive for improvement can therefore be expected to be the norm.
Second, the existence of strong power of influence attracts those who would want that power exercised in ways favourable to their own partisan, generally narrow, interest. The interests can be of a financial or ideological nature, or the two combined. The aim is the same whatever the cause: to tilt Ministerial agendas away from a broader balancing of public interests by prioritising a particular, specific aspect of them. And, in its dealings with Ministers. the Civil Service is itself partisan. It brings its own interests to the table, for example in defending or seeking to increase its budgets and powers,
Third, there is a selection bias in recruitment: it attracts candidates who have a more than average keenness for exerting influence themselves. The message, sometimes made explicit in recruitment advertisements, is ‘you may be able to get a higher salary elsewhere, but here you will be able to influence government policy’. The mischiefs that flow from this are rooted in the way work on policy development is structured and sequenced. There is devil in the detail and the detail is largely left to middle ranking officials, done prior to the final judgment of a Minister. Settling the various issues of detail requires judgments to be made and these can be heavily influenced by the beliefs and sense of priorities of the individual officials involved. It’s remarkable how a few changes in words can significantly change the effects of a regulation. ‘Here you will be able to influence government policy’ is not a fraudulent promise.
Fourth and finally, there is the obstinate fact that in most circumstances, at most times, most Ministers do not express a strong demand for truthful, impartial information and advice. They are themselves partisans, though distinguished from, and elevated above, other partisans by having been elected. Their special interests in today’s politics tend to be ‘the narrative’ and ‘the retail offer’ (to the public) and, if those things become ministerial preoccupations, the Civil Service will tend to have greater scope for deviating from the pursuit of impartiality.
So, given these sorts of factors at work, what might be done about them?
The key insight is, I think, recognition that the supply of truthful information and objective advice is not an inherently monopolistic activity. If major dysfunctions are rooted in monopoly/hegemony, then de-monopolisation is the obvious go-to policy principle, best done as a process, taken step by step. There should be no great concern about the size of early steps: it’s the direction of travel that matters more, via its effects on expectations today, which affect behaviour today.
My suggestion is to establish a new structure to institutionalise/normalise Ministerial choice of sources of information and advice beyond the regular channels, to put at least some competitive pressure on the established structures (stronger if usage had negative implications for departmental budgets) and to promote cognitive diversity. It could also be expected to have indirect effects: a Civil Service with weakened influence on Ministers would be a less attractive target for partisan lobbying, could help mitigate selection bias, and be a less attractive option for individuals keen on influencing government policy.
We can know that the introduction of competing alternatives to the status quo is feasible, because it has been done. UK vaccine procurement is the outstanding example of recent times, done with great success.
A much earlier example, from which I think significant learning can be had, was the Peacock Committee on Broadcasting (1985/6). In those days broadcasting was the responsibility of the Home Office. Doubting the capacity of the Office to handle questions concerning the effects on the broadcasting ecology of allowing the BBC to draw finance from advertising, Sir Alan Peacock called in two small teams of economists to each answer the questions. The ‘outside’ teams were instructed not to talk to each other, or to the Home Office team. In the event, the two ‘outside’ teams took different lines of attack on the problem, but came up with closely similar quantifications of the likely effects. The Committee accepted the relevant conclusions and the Government accepted the Peacock recommendations. With high likelihood, the right decision on BBC financing in the circumstances of the time was made.
Such ‘bypasses’ of the regular Civil Service have been infrequent, depending on serendipitous conjunctions of individuals and circumstances. Hence my stress on the institutionalisation of the bypass option, so that it is always at hand for a Minister, just as the regular Civil Service is.
To fly some kites, its characteristics might be as follows. Provisional name: the Government Economic Intelligence and Strategy Service (GEISS). Separated from the regular Civil Service in all respects. Staffers subject to a Code based on the Civil Service Code (old principles redux), but tightened and broadened to encompass conduct in teams. Staffed by specialised, highly skilled, highly paid, full-time professionals, relatively few in number, working in small platoons (the Bezos two-pizza rule is a rough guide: if the team can’t be fed on two pizzas. the team’s too big). No administrative clutter. Sustained deliberation, but fast responses to calls for urgency. Main base in a sizeable city without a parliament, but with a leisure-friendly hinterland (important for recruitment). No talking with the Civil Service (see Peacock). No meetings with lobbyists (documentary communication only). Ministers the only clients.
The initial focus might be on risk assessment, economic strategy development, and ‘change management’, these being activities where the biggest economic policy errors have been occurring (in the strategy case by tolerating a void). Massively costly assessment mistakes have been made in relation to the 2008 financial crash, post-referendum Brexit options, Covid regulations, and a whole sequence of energy policy decisions.
As for sporadic bypass we can be confident this could potentially work, because institutionalisation of risk assessment and of strategy is observable in the intelligence and security services and in the armed forces (see GCHQ in Cheltenham, strategic analysis in the higher military ranks) and there we find an area of government where performance has been in the better part of the spectrum.
George Yarrow is an Emeritus Fellow of Hertford College, Oxford, and was a co-founder of the Regulatory Policy Institute.
In the context of the decision to close the Tavistock Clinic, Kemi Badenoch wrote recently about her personal experiences as a Minister in dealing with the Civil Service. It was the latest in a longish string of expressions of dissatisfaction with this relationship. Tony Blair complained about “the scars on my back” from his attempts at public sector reform; John Reid, when Home Secretary, told MPs that the Home Office’s Immigration Directorate was “not fit for purpose”; and in his Speaker’s Lecture of 2017, Francis Maude identified a number of system-wide failings in Civil Service conduct and performance, the general message being that ‘things ain’t what they used to be’.
The issues raised are of utmost importance because of the centrality of the relationships between Ministers and civil servants in the development and implementation of public policies. Major failures in this process, which appear to have been increasing in frequency in recent decades, can place very heavy burdens on the public, particularly in economic policy areas (the focus of the remarks that follow).
To address the challenges posed by institutional failures, it’s advisable to start by examining some of the root causes of dysfunctions. I will summarise four.
First, and considered as a whole, the Civil Service is an entrenched, hegemonic institution. It enjoys monopolistic powers over its domain and, as Adam Smith put it, monopoly “is a great enemy of good management”. Competitive pressures motivate sustained search to discover better ways of doing things, whatever the activity, be it science, sport, or business. When it comes to the Civil Service, this source of pressure to raise performance standards is very weak: organisational inefficiency and limited of drive for improvement can therefore be expected to be the norm.
Second, the existence of strong power of influence attracts those who would want that power exercised in ways favourable to their own partisan, generally narrow, interest. The interests can be of a financial or ideological nature, or the two combined. The aim is the same whatever the cause: to tilt Ministerial agendas away from a broader balancing of public interests by prioritising a particular, specific aspect of them. And, in its dealings with Ministers. the Civil Service is itself partisan. It brings its own interests to the table, for example in defending or seeking to increase its budgets and powers,
Third, there is a selection bias in recruitment: it attracts candidates who have a more than average keenness for exerting influence themselves. The message, sometimes made explicit in recruitment advertisements, is ‘you may be able to get a higher salary elsewhere, but here you will be able to influence government policy’. The mischiefs that flow from this are rooted in the way work on policy development is structured and sequenced. There is devil in the detail and the detail is largely left to middle ranking officials, done prior to the final judgment of a Minister. Settling the various issues of detail requires judgments to be made and these can be heavily influenced by the beliefs and sense of priorities of the individual officials involved. It’s remarkable how a few changes in words can significantly change the effects of a regulation. ‘Here you will be able to influence government policy’ is not a fraudulent promise.
Fourth and finally, there is the obstinate fact that in most circumstances, at most times, most Ministers do not express a strong demand for truthful, impartial information and advice. They are themselves partisans, though distinguished from, and elevated above, other partisans by having been elected. Their special interests in today’s politics tend to be ‘the narrative’ and ‘the retail offer’ (to the public) and, if those things become ministerial preoccupations, the Civil Service will tend to have greater scope for deviating from the pursuit of impartiality.
So, given these sorts of factors at work, what might be done about them?
The key insight is, I think, recognition that the supply of truthful information and objective advice is not an inherently monopolistic activity. If major dysfunctions are rooted in monopoly/hegemony, then de-monopolisation is the obvious go-to policy principle, best done as a process, taken step by step. There should be no great concern about the size of early steps: it’s the direction of travel that matters more, via its effects on expectations today, which affect behaviour today.
My suggestion is to establish a new structure to institutionalise/normalise Ministerial choice of sources of information and advice beyond the regular channels, to put at least some competitive pressure on the established structures (stronger if usage had negative implications for departmental budgets) and to promote cognitive diversity. It could also be expected to have indirect effects: a Civil Service with weakened influence on Ministers would be a less attractive target for partisan lobbying, could help mitigate selection bias, and be a less attractive option for individuals keen on influencing government policy.
We can know that the introduction of competing alternatives to the status quo is feasible, because it has been done. UK vaccine procurement is the outstanding example of recent times, done with great success.
A much earlier example, from which I think significant learning can be had, was the Peacock Committee on Broadcasting (1985/6). In those days broadcasting was the responsibility of the Home Office. Doubting the capacity of the Office to handle questions concerning the effects on the broadcasting ecology of allowing the BBC to draw finance from advertising, Sir Alan Peacock called in two small teams of economists to each answer the questions. The ‘outside’ teams were instructed not to talk to each other, or to the Home Office team. In the event, the two ‘outside’ teams took different lines of attack on the problem, but came up with closely similar quantifications of the likely effects. The Committee accepted the relevant conclusions and the Government accepted the Peacock recommendations. With high likelihood, the right decision on BBC financing in the circumstances of the time was made.
Such ‘bypasses’ of the regular Civil Service have been infrequent, depending on serendipitous conjunctions of individuals and circumstances. Hence my stress on the institutionalisation of the bypass option, so that it is always at hand for a Minister, just as the regular Civil Service is.
To fly some kites, its characteristics might be as follows. Provisional name: the Government Economic Intelligence and Strategy Service (GEISS). Separated from the regular Civil Service in all respects. Staffers subject to a Code based on the Civil Service Code (old principles redux), but tightened and broadened to encompass conduct in teams. Staffed by specialised, highly skilled, highly paid, full-time professionals, relatively few in number, working in small platoons (the Bezos two-pizza rule is a rough guide: if the team can’t be fed on two pizzas. the team’s too big). No administrative clutter. Sustained deliberation, but fast responses to calls for urgency. Main base in a sizeable city without a parliament, but with a leisure-friendly hinterland (important for recruitment). No talking with the Civil Service (see Peacock). No meetings with lobbyists (documentary communication only). Ministers the only clients.
The initial focus might be on risk assessment, economic strategy development, and ‘change management’, these being activities where the biggest economic policy errors have been occurring (in the strategy case by tolerating a void). Massively costly assessment mistakes have been made in relation to the 2008 financial crash, post-referendum Brexit options, Covid regulations, and a whole sequence of energy policy decisions.
As for sporadic bypass we can be confident this could potentially work, because institutionalisation of risk assessment and of strategy is observable in the intelligence and security services and in the armed forces (see GCHQ in Cheltenham, strategic analysis in the higher military ranks) and there we find an area of government where performance has been in the better part of the spectrum.