2. Sources and methods

Researching organizational change is hard. Some of the key questions, particularly those around blame-avoidance behaviour, concern motives, which are impossible to observe and which may not be fully known even to the relevant actors themselves. Participants have different and inconsistent perspectives, and their recollections are clouded by the speed of events and the flow of information that accompanies a major governmental emergency. Assigning responsibility for events is, thus, an intrinsically difficult task. Moreover, this paper draws upon a set of interviews conducted over a short period of time, the 8 months between October 2021 and May 2022, when the memory of the change was still recent but the record of access to official documents closed. Against this background, we have primarily relied upon three different sources of information:

1. (1) Published literature, most notably the COVID-19 inquiry by the joint Commons Science and Technology Committee and the Health and Social Care Committee published in 2021 together with the associated record of evidence, and which we refer to as the Joint Committees (2021b) for the report and Joint Committees (2021a) for the transcripts of evidence. We also rely upon the judgement in Harris & Gardner, High Court (2022), particularly its narrative section and reports from the National Audit Office (2014, 2020, 2021). In addition, a rapid narrative literature review was undertaken rather than a systematic review because of the nature of the information the project was seeking to collect and explore and the time limit for the study. Apart from the reports mentioned above, a range of sources were utilized based on a cascade approach linked to key articles concerning the closure of PHE and those identified by the collaborators. These sources comprised of: Government websites of policy papers and press releases (42); Parliamentary Select Committee inquiry reports (35); King's Fund publications (12) and publications from The Health Foundation (4); NHS England (8); British Medical Journal (6); The Lancet (1); PHE blog (8); published medical and health care articles, including Royal College reports (65) (Littlejohns et al., (Reference Littlejohns, Khatun, Knight, Hunter, Markham, Coultas, Kelly and Ahuja2022)).

One author (TK), a postdoctoral researcher within the NIHR ARC South London read all the papers and undertook an inductive thematic analysis (Hayfield, 2021) – coding and theme development was directed by the content of the data. These themes were discussed with two other authors (PL, DJH). The interviews (13 in total) were conducted virtually by two researchers. Potential interviewees were identified from the literature review analysis. The list included prominent commentators on public health policy, members of relevant government departments and PHE as well as local government. The questions were based on the literature review findings.

1. (2) A stakeholder workshop where preliminary findings from the first two sources were presented for verification and criticism (Littlejohns et al., (Reference Littlejohns, Khatun, Knight, Hunter, Markham, Coultas, Kelly and Ahuja2022)).

3. The competing interpretations

What are the principal propositions advanced in each of our two interpretations of the demise of PHE?

On the official interpretation, held primarily but not exclusively by government figures, the key element in the decision to close PHE was its inability to scale up its test and trace operation to the required level, an inability that revealed structural flaws in its organization. Given the failure of PHE to administer mass testing, it was inevitable that any such programme would have to be moved to another organization, leading to the creation of Test and Trace. Joining the excellent science of PHE with the operational practice of Test and Trace eventually required a new integrated structure.

In support of this view, critics of PHE urge a number of points. Despite its success in developing a reliable diagnostic test for COVID-19, PHE fell short on mass testing. On 12 March 2020, testing for COVID-19 other than in hospitals was halted and in mid-March, responsibility for the testing strategy was taken over by the Department of Health and Social Care, although it was not until 18 May 2020 that widespread community testing was resumed in the UK (Joint Committees, 2021b: 62). In late January, it had already been reported to SAGE that PHE only had operational capacity to administer 400–500 tests per day. The Joint Committees (2021b: 61) wrote that ‘it rapidly became apparent that the scientific expertise in identifying the virus and the ability to deploy that operationally were very different’. The Committees went on to contrast the numbers being tested in the UK – 27,476 between 25 January and 11 March 2020 – with the number of tests in South Korea, Hong Kong and Germany. For example, in Germany, some 50,000 people were being tested daily.

In his evidence to the committees, Matt Hancock explained the situation from his point of view as Secretary of State for Health and Social Care:

Hancock returns to the question of testing capacity along the same lines more than once in his evidence.

In complementary evidence, Dominic Cummings, the Prime Minister's Chief of Staff, also noted the failure of PHE to scale up testing capacity and gave this as his reason for drafting in an official to take over the task:

Significantly, according to Cummings, his adverse judgement of PHE was shared by leading figures in government, including the Chief Scientific Advisor (CSA) and the Chief Medical Officer (CMO):

In short, inside government, there was a sense that not only had PHE been unable to scale up testing capacity to the required level, but also that it would be incapable of doing so.

The bias of PHE to its own scientific processes was also of concern to others outside government. For example, PHE was reluctant to take up offers from universities and others to use laboratory capacity. In April 2020 the Director of the Francis Crick Institute, Sir Paul Nurse, was critical of the over-centralized testing regime run by PHE. The non-specialist laboratories in the public and private sectors could play a crucial role (see Parker et al., Reference Parker, Cookson and Shrimsley2020: 7). The Joint Committees (2021b: 7) concluded: ‘The test and trace operation followed a centralised model initially, meaning assistance from laboratories outside PHE – particularly university laboratories – was rebuffed’.

So, the official interpretation can be summarized as follows. It would have been unwise simply to have thrown money at PHE in the hope that it could have scaled up its testing capacity in the required time. The problem needed the shock of the new and it was better to establish a separate organization to conduct the testing. PHE could then later be merged with that organization and the newly formed body headed by someone who lacked the caution of the traditional civil service.

By contrast with this official interpretation, the sceptical interpretation sees the decision primarily as a piece of blame-deflection behaviour on the part of policymakers, particularly Matt Hancock and Dominic Cummings.

A Lancet (2020) editorial put the case succinctly:

The editorial then went on to criticize the appointment of Dido Harding to head Test and Trace, as showing political bias. Harding, it went on to say, ‘has limited experience of public health but is married to a Member of Parliament for the governing Conservative Party. Harding's suitability for the role, even temporarily, is further called into question by the record of NHS Test and Trace, which since mid-June has failed in its target to reach 80% of people who test positive for COVID-19’. In a similar vein in the BMJ Iacobucci (Reference Iacobucci2020) wrote that PHE was ‘being scapegoated by ministers who were ultimately responsible for the country's response to COVID-19’.

McKee (Reference McKee2020) reported the views of various leaders of the medical profession:

McKee went on to cite others who were similarly critical of the decision including Chris Hopson, chief executive of the hospital body NHS Providers, and Dr Chaand Nagpaul, chair of the British Medical Association's ruling council.

Unsurprisingly, similar opinions were held by all of our interviewees from the public health community. They repeatedly ascribed the decision to abolish PHE to blame avoidance and scapegoating of a team of dedicated public health professionals. One saw PHE as ‘a fall guy’, a necessary role when things go wrong in government. Interviewees also stressed the extent to which the problem that PHE was dealing with were a major challenge, both in terms of the severity of transmission and the disjointedness of the global surveillance system, the latter not being PHE's responsibility. If the core of the failure of PHE was its inability to scale up its testing, then its defenders say that it was never designed to undertake mass testing and it would have been hopeless expectation for government to think that it could have done so. This was both a matter of the initial design of PHE and of funding over 7 years of its existence.

Among our interviewees, those who accepted that some blame can be laid at the door of PHE still questioned the need for its closure in the middle of a pandemic. Some saw not only an attempt at blame-avoidance but also opportunistic behaviour. According to this view, Dominic Cummings, who was well-known to have highly critical views of the UK civil service, took the chance to destroy an organization that did not live up to his idiosyncratic expectations. There were those in the Conservative Party who were sceptical of the public health agenda. The appointment of Dido Harding gave rise to the suspicion that the pandemic provided the occasion to constrain public health policies that many found ideologically unpalatable.

The official and the sceptical interpretations are obviously contrasting. That does not mean that there cannot be elements of truth in each of them. For example, even if there were substantive problems with the performance of PHE, the decision to deal with those problems by abolition, rather than reform, might have been motivated by blame-avoidance behaviour. Since, in the UK system of government, accountability moves up the system through the principle of ministerial responsibility, a government minister may well hope to take pre-emptive measures to protect against blame. Conversely, although it may be true that PHE was not set up to deal with large-scale test and trace operations, its activities around pandemic planning, given the experience of South Korea, might have alerted it to the need to put some surge capacity in place. The mandates of governmental organizations evolve and develop and it is the mark of a strong organization that it can take on new tasks.

However, although it is possible to reconcile in some measure these two interpretations, they clearly are competing. The official interpretation makes the need for a new organization inevitable; the critical interpretation makes the option of reform of the existing organization possible. The official view sees a government dealing with a problem of unprecedented novelty and scale that finds that one of the principal organizations on which the implementation of its policy relies is not capable of rising to the challenge, so that it has to improvise a new body. The sceptical view sees an experienced and dedicated set of professionals who find themselves the victims of poor institutional design and cumulative budget austerity.

How can we adjudicate between these interpretations? To do so, we look at four sets of issues:

1. (1) How far were the alleged failures of PHE uniquely attributable to it or how far should responsibility be shared?

2. (2) How far were any failures of PHE a product of its origins and history? How far, in other words, was PHE cursed with original sin in its design? And how far was that original sin compounded by failures of funding?

3. (3) Was it necessary to create a new organization to conduct testing that later was merged with PHE, and how far did the process of creating a new overarching organization for public health show a bias towards the private sector and against the place of public service?

4. (4) How far can the abolition of PHE as a free-standing organization be regarded as an example of blame-shifting behaviour by the government?

We take each of these topics in turn.

4. Shared responsibility?

One criticism of PHE is that it should have been better prepared for a pandemic like COVID-19. The huge task of scaling up testing may have been made more difficult than otherwise by the assumptions about the character of future pandemics that PHE had made. In October 2016 it had led the Cygnus contingency planning exercise, where the working assumption was that the next major pandemic would be like flu. No consideration seems to have been given to the possibility that transmission could be from asymptomatic individuals. Moreover, Cygnus only simulated the treatment and escalation phases of a pandemic taking as its starting point the announcement of a pandemic by the WHO. So, it was claimed by the Joint Committees (2021b: 17–18) that PHE never really learnt from the East Asian experience of SARS and MERS, particularly the test and trace programme in South Korea.

However, these planning assumptions were broadly shared in the UK public health community. Commenting on this fact, Dame Sally Davies said to the Joint Committees (2021b: 19, para. 27):

So, although PHE has to carry some responsibility for its insufficiently cautious planning, so do other agencies and individuals.

A similar point can be made in respect of asymptomatic transmission, one of the most challenging aspects of COVID-19 from a control point of view. However, this was not fully appreciated at the start of the pandemic. For example, at its meeting on 13 January 2020, NERVTAG took the view that the virus did not appear to be very transmissible, let alone asymptomatically transmissible (High Court, 2022: para. 30). Others involved in giving scientific advice were sceptical about testing asymptomatic adults. On 28 January 2020, for example, SAGE advised against testing asymptomatic individuals because of the possibility of false negatives. So, as with assumptions about the character of a future pandemic, a correct understanding policy problem was not self-evident. Reliable policy relevant understanding necessarily emerges from a community of agents operating in the domain of policy decision making. Responsibility for the failure to plan for a pandemic of the type of COVID-19 goes more widely than PHE. This does not mean that there were no grounds for abolishing PHE, but it suggests that anyone looking at the issue impartially would have thought there was a wider structural issue in relation to pandemic planning.

5. Original sin?

Defenders of PHE are prepared to admit that it failed in respect of its inability to scale up its testing operations, but ascribe the source of that failure to its design, holding that PHE was never set up to deal with something on the scale of the COVID-19 pandemic. In its early days PHE had to bring together the large number of separate laboratory services for which the organization had become responsible. However, in other aspects of its work, it was clear that PHE lacked capacity. For example, in its report on the implementation of the Lansley health reforms, the National Audit Office (2013: 23) noted that PHE was under-staffed to the tune of some 12% of posts, with many vacancies being in the immunization and screening services.

More generally, two features of the Lansley reforms are particularly important in understanding the focus and limitations of PHE. PHE was established as an executive agency to provide national leadership for public health while locally, the lead responsibility for public health returned to local government where it had been located prior to the 1974 NHS reorganization. This was in contrast to the National Institute for Health and Care Excellence (NICE) that was re-established as Non Departmental Public Body in the same reforms. In a report of 2015, the House of Commons Public Accounts Committee concluded that although PHE ‘has made a good start in its efforts to protect and improve public health’, it felt that the organization lacked ‘strong enough ways of influencing local authorities to ensure progress against all of its top public health priorities’ as measured against the outcomes framework that had been adopted for public health (House of Commons Public Accounts Committee, 2015: 3). This followed a critical report from the National Audit Office which was unimpressed with PHE's efforts to provide local authorities with evidence and decision support tools (National Audit Office, 2014). Endorsing the criticism, the PAC recommended improved responsiveness on the part of PHE to local authority requests for support, including help with understanding the evidence base and cost implications of different public health interventions (Public Accounts Committee, 2015). In policy terms, PHE could only seek to influence, and not direct, local authorities to make good progress in improving the public's health (Hunter, Reference Hunter, Exworthy, Mannion and Powell2016). The tension is its remit was built into its set-up. On the one hand, it was accountable for securing improved health outcomes but, on the other, the levers available to it to ensure that such outcomes were realized were few and limited in their reach.

Lansley emphasized that his reforms had a focus on maintaining and enhancing the scientific quality of health protection. So the bias towards science rather than operational testing was present at conception. However, it can be argued that governmental agencies in the UK are not so tightly defined by their statutory mandate that their functions cannot evolve over time. The paradigm case in this regard is NICE. Originally established to undertake cost-effectiveness analysis on pharmaceuticals and medical devices, within 10 years the scope of its work expanded to deal with public health and clinical standards among others. Could not PHE have similarly expanded the scope of its work to include operational effectiveness in the face of a major pandemic? Counterfactual speculation is always hard. However, there are some significant differences between the development of NICE and the development of PHE. NICE received high-level political backing from its beginning, with Frank Dobson as Secretary of State being willing to give it strong support in the face of opposition from some pharmaceutical manufacturers. By contrast, on at least two issues – minimum unit pricing of alcohol and obesity – that PHE took up, elements of the governing Conservative Party, including Andrew Lansley, were sceptical of what they regarded as ‘nanny state’ public health interventions.

NICE also expanded at a time when the Labour government was significantly increasing public expenditure on health care. By contrast, PHE's initial years occurred in a period of severe public expenditure real-term reductions. In such a period, the NHS was always going to be the beneficiary of any spare money. Figure 1 shows the scale of the reductions. As can be seen from the chart, PHE lost nearly £100mn from its budget between the first and second year of its full operation, and even after that major cut continued to suffer reductions in nominal as well as real terms. At a time when there was continuous pressure to increase funding to the NHS, commitments on the public health side were always vulnerable to losing money.

Figure 1. PHE annual budget spending allocation (2014–2021).

To see the scale of the difficulty PHE faced, consider the following thought-experiment. Suppose that the Cygnus exercise had been conducted on the assumption that any future pandemic would require a response on the scale of the Korean programme developed in the light of the 2015 MERS outbreak (compare Joint Committees, 2021b: 61, para. 169). Suppose too PHE that had argued vigorously in negotiating its public expenditure settlement for an increased allocation of expenditure sufficient to provide a contingency for testing in such a major pandemic. Suppose also that a realistic estimate of what such a contingency would be would have amounted to something like the £37bn over 2 years that was eventually allocated to Test and Trace, a figure that the Joint Committees (2021b: 70–71, para. 201) pointed out was more than the Home Office and Justice Department's annual budgets and more than twice the spending on the whole of scientific research. It is hard to believe that such an expenditure bid, even on a contingency basis, would have got anywhere in Whitehall negotiations given the other claims on resources. Only experience would counteract the sentiment that in making such a bid the agency was crying wolf. It would take the focusing event of the pandemic to shift government opinion, but by then it was too late for PHE.

6. Why a new body?

If we accept that PHE not only had been unable to conduct testing on a mass scale, but was unable to do so given the organizational features that it had inherited from its initial design, we can still ask whether it was necessary to create a new organization? Could PHE not have been funded adequately to carry out the required work?

In answering this question, we note that the creation of UKHSA went through three stages. The first occurred in mid-March 2020 when the Department of Health and Social Care took over responsibility for testing from PHE. The second occurred at the end of May 2020 when Test and Trace was established. And the third occurred in August 2020 when the plan to merge PHE with Test and Trace into a new agency was announced. Of these, clearly the most crucial was the first step. It rested on the assumption that PHE lacked the organizational capacity to conduct testing at the required scale, and that a new organization was needed. The organization that was to become UKHSA was then the logical outcome of the creation of Test and Trace, which also contained the Joint Biosecurity Centre. Once Test and Trace was created, it would seem anomalous to have two different organizations separate from the core department involved in securing public health protection. In this context, it is important to note that, though the initial government announcement referred to the ‘abolition’ of PHE, PHE was effectively merged into a larger organization in which Test and Trace was also located. So the question then becomes why Test and Trace was set up as a separate organization rather than providing the funding to PHE.

The answer to this question is that both Matt Hancock and Dominic Cummings judged that PHE would not be able to rise to the challenge. Moreover, if the testimony of Cummings before the House Joint Committees is to be believed, both Sir Patrick Vallance and Chris Whitty shared their scepticism. If testing was going to be an integral part of the control strategy, then action on a massive scale would be urgently required and the consensus at the core of government was not the organization to undertake action on the required scale.

However, the suspicion exists in some minds that the creation of Test and Trace revealed a set of anti-public services prejudices on the part of leading actors in the decision. Dominic Cummings had been a long-standing critic of government organizations, accusing them of a bias against decisive action and a failure to use modern data analytic techniques. In June 2020 he was reported as saying that a ‘hard rain’ would fall on the civil service, which he regarded as being too large for effective management (Johnstone, Reference Johnstone2020). (It should be noted, however, that Michael Gove appearing before a parliamentary select committee denied that Cummings, who had previously worked for Gove, ever used the phrase. For the denial, see Dunton, Reference Dunton2020.) Moreover, given the appointment of Baroness Harding to head Test and Trace, the climate of opinion in government seems to have been that private sector expertise was required, despite the failures during Harding's time as CEO of TalkTalk, when the organization was subject to a data breach of major proportions.

Of course a private sector background is not of itself inimical to good public administration. The Vaccine Task Force, chaired by Kate Bingham, who had a background in venture capital, is thought to have done a good job. However, the life of Test and Trace in its first year from May 2020 onwards was not a happy one. It failed to achieve the number of tests that its capacity allowed for. It failed to prepare adequately for the expected surge of cases that would be needed in autumn 2020 and so, despite the money allocated to it, it did not help prevent a second lock-down that year. There were delays in establishing new laboratories, delivering test equipment and appointing staff. It made unsubstantiated claims about its success in reducing R. And its turnaround time for test results, so vital an element of response, was too slow over too long a period. A particular and significant failing of Test and Trace, however, was its initial reliance on a system of centralized private tendering in which Serco and Sitel were awarded some £720mn to run the programme. The privatization bias was thus shown in the decision to bypass local government, even though it is the directors of public health in the local authorities who possess the requisite knowledge make a testing system work effectively. (For these criticisms, see National Audit Office, 2020, 2021; Joint Committees, 2021b: Ch. 4.)

A number of our interviewees questioned whether it was necessary to announce the new organization in the middle of the pandemic and to do so at such a stressful time. A defender of the government's decision could argue, of course, that no doubt things could have been done better in the way the change was made and that it was unfortunate that the staff were given no warning. But a defender of the decision could also claim that all reorganization is painful, and there is no obviously right time to inflict the pain. However, the issue of timing is quite distinct from the manner in which the reorganization was announced, which itself is related to the question of how far the whole story can be understood in terms of blame avoidance behaviour by the government, and by Matt Hancock as Secretary of State in particular. Moreover, the rapid changes in titles of the two bodies to replace PHE were indicative of a government acting in haste with little clarity concerning its objectives.

7. Was it really scapegoating?

The view that the demise of PHE was the result of scapegoating by the Secretary of State rests in part upon a judgement that the government's handling of the pandemic had been poor. In the weeks before and in the early stages of the pandemic, the government had been fixated on completing the UK's exit from the European Union, the promised to ‘get Brexit done’ that Johnson had campaigned on in the December 2019 election. Johnson was absent from crucial Cabinet sub-committee meetings at the start of the pandemic; the government was slow to lock down; at times there seemed to be a policy of herd immunity; death rates were measurably higher than in comparable countries; in line with his political manner, Johnson constantly talked up the quality of the government's response, repeatedly referring for example to the creation of a ‘world-beating’ test and trace system; and the government wasted the summer, when there was a decline in infections, as an opportunity to prepare for further waves of infection in the autumn. Because public expenditure settlements had been very tight since 2010, the NHS entered the pandemic short of beds, facilities and medical staff. In these circumstances, so it is alleged, the government focused on the failure of PHE as a way of deflecting attention from its own poor performance.

The theory of blame-avoidance has been extensively studied. As Hood (Reference Hood2010: 5) has explained, the theory begins with the assumption that people in general wish to avoid blame and its consequences, and politicians in particular wish to avoid blame because it damages the chances of their re-election. As a result the logic and politics of blame avoidance can triumph over good governance. Of course, politicians also wish to claim credit when things go well, but the theory of blame-avoidance posits a ‘negativity bias’, such that failure is more prominent in public attention than success. There are a variety of ways in which blame may be avoided or deflected, including presentational strategies involving ‘spinning’ the policy, policy design strategies involving the introduction of automatic mechanisms, and, most pertinent to our study, agency strategies by which responsibility is delegated to bodies so that they act as lightning conductors to deflect blame from high-level office-holders according to the principle of ‘find a scapegoat’ (see Hood, Reference Hood2010: Table 1.1 and Ch. 4). The particular agency strategy that Hood identifies and which is most relevant to the case of PHE is what he calls ‘defensive reorganization and staff rotation’, where the motto is ‘that was then, this is now’ and the frequent examples are to be found in services where there is a high potential for serious blame, for example child welfare services (Hood, Reference Hood2010: Table 4.1).

Although scapegoating in the form of defensive reorganization seems plausible, it would be premature to jump to the conclusion that scapegoating explains a large part of the decision. In the first place, there is the danger of committing the logical fallacy of affirming the consequent. The argument ‘if there was scapegoating then PHE would have been abolished, PHE was abolished, therefore its abolition was a result of scapegoating’ is about as pure a case of the fallacy as one can get. The scapegoating explanation will only stand scrutiny if there is no other more plausible explanation. It requires justification by the exclusion of other possibilities.

The obvious candidate for an alternative to the scapegoating explanation is the official one: once PHE failed to produce an operational testing regime at the requisite scale, the Department had to step in and then Test and Trace had to be created. The logic of the situation then involved brigading the two organizations together. The demise of PHE as a separate organization was simply a consequence of this chain of events. No untoward motives need to be posited.

There is one further line of argument that casts some doubt upon the scapegoating hypothesis. Responsibility, and therefore blame-attraction, sticks to ministers in the UK system of government. It is unlikely that the inquiry into the government's conduct of its COVID-19 policy will exonerate the members of the government purely on the basis of the failures of PHE. Already the Joint Committees (2021b) have been very critical. In the case of Gardner & Harris the High Court (2022: para. 298) found a breach of a common law duty in respect of two policy documents issued by PHE and the Secretary of State on the discharge of patients from hospital and the guidance on admission to care homes, neither of which took into account the possibility of asymptomatic transmission. Ironically, the criticism that PHE, as an executive agency, was not sufficiently independent of the department meant that the DHSC was also liable for misleading advice that was issued. Had PHE had the independence that some of its critics said that it ought to have, it would have been easier for the Secretary of State to escape blame. In short, if blame-avoidance was the basic explanation of the demise of PHE, the institutional structure was poorly designed to facilitate the process and it is unlikely to be successful in any case.