Modern Control is Soviet Control: the NHS as a case study
but not in the way you might think...
All modern bureaucracies operate much like their Soviet counterparts once did. And paradoxically, it is the drive to make them less bureaucratic and more metrics/market-driven over the last four decades that has made them more Soviet in nature.
This is not a novel argument. Aviezer Tucker has shown how European universities operate on a “managerial model that resembles the late Soviet model of industry during the Brezhnev era”. However, the the best example of this evolution of all modern bureaucracies towards a more Soviet control regime is the evolution of the British National Health Service (NHS) since its creation after WW2. Not only does its current state resemble a late Soviet-era bureaucracy, but its entire lifecycle has also mimicked the evolution of the Soviet economic system.
The lifecycle of the NHS
Decentralised fiefdoms
At the beginning, the NHS had a structure that was as far removed from 'command and control' as possible1. The structure was largely decentralised and managed through exhortations rather than commands. Rather predictably, this structure led to local fiefdoms. The long period of stability inevitably led to these local fiefdoms being more and more resistant to the government's exhortations. Richard Crossman, the Secretary of State for Social Services in the 60s, described his position as follows:
You don’t have in the regional hospital boards a number of obedient civil servants carrying out central orders... You have a number of powerful, semi-autonomous boards whose relation to me was much more like the relations of a Persian satrap to a weak Persian emperor. If the emperor tried to enforce his authority too far he lost his throne, or at least lost his resources, or something broke down.
The presence of fiefdoms, whether they were regional or ministerial, was a constant problem for the Soviet Union. Stalin was aware of the problem of local ‘nests’ and the violence of the Stalinist era was partly aimed at countering this problem. Khruschev too lamented the controllability of local officials but was ousted partly because he tried too hard to break up the fiefdoms in the system. Brezhnev essentially gave up the fight with his ‘stability of the cadres’ policy but at the cost of permanent decline.
Integration
The first hint of ‘command and control’ came in 1974 when the NHS was integrated and the services that had been left in local hands (e.g. midwifery, ambulance services) were brought into the NHS. Even by this stage, the consensus-driven approach of the NHS had become bureaucratic and sclerotic.
Integration and centralisation, by definition, demand more control and this came in the 1980s with targets, performance indicators and performance reviews. There is a deep irony that, in the 1980s, “just as the targets system was collapsing in the USSR, the same basic approach came to be much advocated for public services in the West by those who believed in ‘results-driven government’”2. However, much more than neo-liberalism, it was increased computerisation that also enabled a move towards metrics-driven control.
Every reform adds more layers of management
In 1985, the NHS got a management board but this was just another layer of management that immediately ran into conflicts not only with the chairs of the regional health authorities (whom Richard Crossman had called “feudal barons”) but also with the ministers themselves.
This tendency for every series of ‘reforms’ to add layers of control is a universal feature of the evolution of all codified systems. For example, every “reform” that sought to take power away from the all-powerful Soviet Ministries only served to create more committees and regional ‘mini-ministries’ with no dilution in either the size or the power exerted by the Ministries3.
Mimicking Markets
The 1990s saw the introduction of “internal markets” to the NHS where:
Instead of health authorities directly managing hospitals, there was to be a purchaser/provider split. The NHS, arguably for the first time, would decide what care it wanted and then purchase it on contracts from “self-governing” hospitals to be known as NHS trusts, which were given a limited range of new operational freedoms. Hospitals, in effect, would have to compete for patients. The belief was that the injection of at least an element of competition would increase efficiency, and – it was hoped – quality.
Predictably, this too added more and more layers to the NHS structure and “mounting political sensitivity over ‘the men in grey suits’ because the purchaser/provider split was driving up management costs as hospitals, health authorities and GP fundholders negotiated with each other”.
As I have described in a previous essay, the Soviets too went from simple output targets to “profit” targets and instituted an increasingly elaborate bonus system for managers in an attempt to mimic the behaviour of a capitalist market economy.
The Tyranny of Targets, Metrics and Incentives
However, the NHS only truly became Soviet during Frank Dobson’s tenure as health secretary under Tony Blair’s government from 1997 to 1999 which saw “the most determined effort in NHS history to run the service by “command and control” from Whitehall”. As Gwyn Bevan and Christopher Hood have argued, this approach was a “watered down version” of the Soviet system - targets, positive and negative incentives (bonuses and the risk of being sacked) and an elaborate performance monitoring system4.
The intention was to replace the internal market with centralised control but what followed over the next decade was a mix of targets, market-mimicking structures and a plethora of new bureaucratic agencies. As Gwyn Bevan and Christopher Hood have noted, “what lay behind the system of governance by targets in health care in the early 2000s amounted to an institutionally complex and frequently changing set of overseers, inspectors and assessors”.
Gaming the Targets
Sooner or later, all systems that rely on metrics and incentive payments succumb to Goodhart’s Law and the reforms of the 2000s were no different. The most famous example was the four-hour target for treatment in the A&E. To meet the target, ambulances were often parked outside the hospital (so that the clock does not start). In 2016, “20,000 patients in south England were found to have had their ambulance to hospital deliberately delayed in order to help the hospital reach targets for ambulances” and “an ambulance Trust was found to have ‘downgraded’ 111 calls for ambulances in order to meet performance targets”5
The ‘trolley-wait’ target that a patient must be admitted to a hospital bed within 12 hours of admission led to trolleys being turned into beds6. And "the waiting time target for new ophthalmology outpatient appointments at a major acute hospital had been achieved by cancellation and delay of follow-up appointments, which did not figure in the target regime. Recording of clinical incident forms for all patients showed that, as a consequence, 25 patients lost their vision over two years, and this figure is likely to be an underestimate"7.
The gaming persisted not just because of the ingenuity of those subject to the targets but because there is no incentive for even the supervisors to detect gaming. As Bevan and Hood have argued, the same incentives applied in the NHS as applied in the Soviet system:
In the Soviet system, as all bodies responsible for supervising enterprises were interested in the same success indicators, the supervisors, rather than acting to check, connived at, or even encouraged, gaming. In the English NHS, ‘hard looks’ to detect gaming in reported performance data were at best limited.
From Stability to Deteriorating Control and Permanent Crisis
Although the NHS today seems to be in a state of permanent crisis, it was remarkably stable and unchanged from its inception till 1974. Since then, it has been subjected to
something like 20 reorganisations, depending on precisely how you count them. In other words, on average, around one every two years – to the point where “organisation, re-organisation and re-disorganisation” might well be dubbed the NHS disease.
The same was true in the Soviet system with the period since the 60s seeing an almost continuous series of economic reforms, each of which only served to make the system less efficient and more sclerotic.
Lurching between Centralisation and Decentralisation
If one set of changes aimed to centralise and integrate the NHS, the next set of reforms aimed to decentralise the system. The NHS started as local fiefdoms, followed by “integration” (centralisation), followed by “internal markets” (decentralisation), “command and control” (centralisation) and finally Andrew Lansley’s reforms in the 2010s which were supposed to be “bottom-up”. As is universally the case, every such set of changes is purely additive and “reorganisation has often been cyclical, with new governments or ministers reinventing structural arrangements that their predecessors abolished, seemingly unaware of or uninterested in past reorganisations”8.
The Soviet experience was no different. The reforms oscillated between increasing the power of the enterprises and local territories and increasing the breadth and intensity of the control exerted by the centre.
Death by Complexity
By the 2010s, the NHS had reached the stage that all such control projects eventually reach, a state of total paralysis thanks to an incomprehensible level of complexity. After the Lansley reforms, “it ceased to be possible to draw an organogram that explained, with any truth, how the NHS worked”9 and the NHS now has a "bewildering range of regulatory bodies"10.
Complexity also increases because, as with every system of targets and incentives, the system “becomes a race between the ingenuity of the regulatee and the loophole closing of the regulator, with a continuing expansion in the volume of regulations as the outcome”11. More complex control cannot overcome Goodhart's Law.
The Soviet economic system went through the same transition from relative simplicity to paralysing complexity. Post-Stalin, the Soviet leadership recognised the shortcomings of the simplistic planning system. Thus followed a constant treadmill of ‘reforms’ many of which were changed, reversed or not fully implemented but nevertheless resulted in the inexorable buildup of cruft and ‘legacy code’. The sponsors of each change believed that they had found the magic bullet to cure all of the system’s ills yet every change only made the system even more illegible.
Compared to the 40s, the Soviet manager of the 80s faced a more complex system with many more bosses, more committees, more indicators used for control (from several tens indicators in the 60s to several hundred indicators in the 80s), and an elaborate bonus system that the manager could barely understand. As a result, the performance of the system and the Soviet enterprises deteriorated to a level that led to systemic collapse.
The Unavoidable Lifecycle of Control
The NHS is not so much over-managed as it is over-controlled. However, this is not a condition unique to the NHS or even to the Soviet Union. It is the unavoidable nature of codified, "legible" control. Over time, all control becomes more complex, more expensive and less effective. What starts out as a simple, legible process eventually transforms into a complex, illegible mess.
Most of the information regarding the timeline comes from these two excellent reports - https://www.kingsfund.org.uk/sites/default/files/2018-05/worlds_biggest_quango_ifg_may2017.pdf and https://www.kingsfund.org.uk/sites/default/files/field/field_publication_file/never-again-story-health-social-care-nicholas-timmins-jul12.pdf
Joseph Berliner in ‘The Innovation Decision in Soviet Industry’
A brief description of the Soviet system can be found in my previous essay ‘More complex control cannot overcome Goodhart's Law: The Soviet experience’.
Margaret McCartney in ‘The State of Medicine’
Timmins, Nicholas. The Five Giants [New Edition]: A Biography of the Welfare State . HarperCollins Publishers. Kindle Edition.
'The Public Use of Private Interest' by Charles Schultze