Healthcare.gov, strategic leadership, and lessons for UK local government

(This post was written as an assignment for MLD670M, a course at the Harvard Kennedy School of Government)

“It ought to be remembered that there is nothing more difficult to take in hand, more perilous to conduct, or more uncertain in its success, than to take the lead in the introduction of a new order of things. Because the innovator has for enemies all those who have done well under the old conditions, and lukewarm defenders in those who may do well under the new. This coolness arises partly from fear of the opponents, who have the laws on their side, and partly from the incredulity of men, who do not readily believe in new things until they have had a long experience of them.”

― Niccolò Machiavelli, The Prince

While clearly Machiavelli didn’t have the implementation of government digital projects in mind when he penned the above quote, I think his insights accurately apply. Having spent some time studying the failures of healthcare.gov, and the broader landscape of government digital projects, I’ve become convinced that strategic leadership is key to enabling success.

In this piece, I want to look at the lessons of the course through the prism of Mark Moore’s strategic triangle, which I’ve learnt about through my work on Jorrit de Jong’s Innovation Field Lab course. I’ll first sketch out key questions under the headings of the ‘strategic triangle’ framework, and then apply the framework to potential digital innovations in UK combined authorities.

Public Value — What is the scale and nature of the problem you are trying to address through digital innovation? What is your theory of change? What are the knowns, unknowns, and unknown unknowns?

The first area of the strategic triangle is public value, which requires innovators to think about the issue they are attempting to solve through their intervention. One of the key insights from healthcare.gov is that if governments are going to engage in large-scale digital innovation, they need to adjust the way they approach the definition of public value. Dominant historical approaches (as reflected in federal procurement documents) have emphasised tight definitions of technical problems alongside lengthy specifications of how government would like a vendor to solve them. If you need a bridge built, then this approach can work fairly well, as the public value proposition is clear — if you’re looking for a vendor for a social service then less so, and if you need a vendor to produce a digital platform then this is even less the case. This is partially to do with three factors:

· Precedent: Has a similar project been undertaken before?

· Government expertise: Are there individuals within Government departments that fully understand this area?

· Static problem: Is the nature of the problem fixed and unlikely to change during development or after deployment?

In all three of these areas, the answer for large-scale digital innovations is usually ‘no’. Defining public value for these digital projects is therefore an exercise in establishing desired end results, but not tightly defining the mechanism by which that value can be realised. This has significant implications, particularly for the style of procurement adopted by leaders. Steven Kelman has indicated that over the past two decades there has been a move away from hundreds of pages of specifications, and Rohan Bhobe from Nava supported this in indicating that from a vendor side he was seeing movement from ‘statements of work’ to ‘statement of outcomes.’ However, this appears to be a very nascent shift in the way that leaders define projects and translate them into request for proposal (RFP) documents.

Authorising Forces — What decisions need to be taken as this project develops? Do individuals leading and acting in this team have the ability to take those decisions? What avenues exist for ultimate decision-makers to be reached, and is there an environment in which they can make informed decisions?

The second area of the strategic triangle looks at authorising forces, focusing attention on stakeholders who have an influence on whether your innovation can be successful. An interesting feature of speakers on both sides of the healthcare.gov project — pre-launch and rescue — is that they identify access to decision-makers as a key determinant of success. Henry Chao, project manager of healthcare.gov within CMS, shared his frustration about the inability for senior leaders within federal government to sign-off decisions about the architecture of the site. In particular, he emphasised that the political decision not to proceed with aspects of the policy framework during the 2012 presidential campaign paralysed his team. Ryan Panchadsaram, former White House Deputy CTO and member of the rescue team, reinforced this message and emphasised that ‘air cover’ was a key reason he and his colleagues were able to move quickly with fixes. Ryan was able to get decisions from the White House and HHS within three hours, while Henry Chao often went months without a clear steer.

Given the discussion above regarding the need for public value definition to be an iterative process, this question about authority becomes particularly pertinent. It is likely that fundamental decisions about the nature of the project will need to be taken as it develops, not specified comprehensively at the outset. There are therefore two options — either fully empower teams to make decisions, or open up channels to senior decision-makers so that they can be taken quickly. Within a government context, the first option is understandably difficult given the nature of democratic accountability and the ultimate responsibility of elected officials. Projects therefore need clear channels by which team members can share challenges open and frankly with either politicians or their advisors, and then receive decisions. Here, healthcare.gov offers sobering lessons — the stark contrast between challenges faced by the delivery team and comments on the Whitehouse lawn prompts the question: ‘Why did nobody tell the President?’.

Operational Capacity — What environment is most conducive to solving the challenges that have been identified? What methodology should you adopt with your team in order to manage risk and maximise the potential of digital? What skills do you need to enable delivery?

The final area of the strategic triangle looks at operational capacity — in the context of digital implementation, this could clearly mean a wide range of factors. Here, I want to focus specifically on capacity as it relates to leadership. Adopting Ronald Heifitz’s framework on leadership, David Eaves has indicated that a key element in overcoming the capacity challenge in many large-scale digital implementations is identifying them as adaptive challenges, not technical problems. This means that they require changes in beliefs, working relationships, attitudes, and fundamental approaches — not just customising a pre-built piece of technology or putting a paper form online.

Much of the discussion during our study of healthcare.gov has focussed on the contrast between waterfall and agile, which maps on to the technical/adaptive distinction. While waterfall focusses on sequential delivery, with clear specifications at the beginning and full launch at the end, agile focusses on iterative and simultaneous progress, with regular user-testing and deployment of portions of the product once they are completed. Eaves argued that waterfall is sometimes appropriate for technical problems — if you need to replace a database, then you can do so in an ordered, sequential way given clear specifications. However, adaptive challenges require an agile approach. They are inherently relational and ill-defined — as a project develops, you will encounter new problems and opposition from new parties, and will need to adapt in response. David Moskovitz, head of Accenture’s Federal Service, reinforced the view of digital implementations as adaptive challenges when he indicated that AFS look for government clients who understand the need to manage ‘inherent tensions’ in the design and delivery of digital solutions. These clients, he argues, are more willing to be creative and innovative in co-developing systems because they recognise that traditional solutions using enterprise software won’t work.

These two types of problem, and two approaches, require different types of leadership. Technical problems need leaders who can design solutions themselves, address failure as it occurs, and manage a team through implementation. Adaptive challenges in the digital space need leaders who design solutions iteratively based on user needs, fail forward, and encourage teams to experiment. The challenge for government is therefore building capacity among its cadre of leaders to be comfortable with the complexity and ambiguity of adaptive challenges, and competent with the appropriate methodologies to address them.

Applying the lessons of healthcare.gov to UK combined authorities

In May 2017, six new ‘metro-mayors’ were elected in England to lead combined authorities — regional government units that have been given unprecedented powers and budgets in order to drive growth and address social challenges. These bodies have enormous potential to deliver results for the communities they serve, and I’m hoping to work with them in the future to support their mission. One of the tools available to mayors will be digital innovation, and in this section I’ll suggest what healthcare.gov could teach them about how to be successful.

Firstly, mayors will need to consider carefully how they define public value. A less ambitious and risky approach could be, for example, to integrate the databases used by training providers within individual councils, to provide greater functionality and share information. This would potentially lead to improvements, and create some public value. However, the injection of funds, devolving of powers, and resetting of relationships with the private sector could warrant a complete reconceptualisation of public value as it relates to digital innovation. For instance — an entirely new online workforce skills system could integrate training opportunities, job advertisements, casework management for unemployment support workers, apprenticeship development pathways, and other elements of the education, skills, and employment pipeline. With public value defined this broadly, there are no clear specifications for what a final system would look like, nor defined milestones for how it would be developed. Yet adopting the kind of modular, agile approach that Todd Park and Michael Wilkening effectively advocated for with California’s Child Welfare System could yield sustained results, by focussing public value statements on outcomes. Mayors could announce a vision for an integrated, user-friendly online platform that ignores the historical silos and departmental distinctions between education, training, work, and welfare — and then design RFPs in such a way that a system could be developed piece-by-piece to meet that outcome.

Secondly, mayors will need to develop pro-active strategies for managing authorising authorities. As leaders of city regions, their initial position should be to empower the team developing the programme to make decisions within a pre-agreed framework, and then be available to make ad-hoc policy decisions on a timely basis where necessary. However, whilst metro-mayors have been given a lot of statutory authority, in practice they sit in a precarious position between a national government that is fundamentally uncomfortable handing down power and local councils that don’t want to pass it up. This means investing a significant amount of time upfront working with stakeholders across both local and national government. To continue the example of an integrated online workforce skills system, this would mean proactive discussions with:

· Prime Minister’s team in Number 10 (to get overall policy clearance)

· Department for Work and Pensions (access to Job-Centre Plus information)

· Department for Digital, Culture, Media, and Sport (access to digital infrastructure and funds)

· Department for Business, Energy, and Industrial Strategy (relationship to skills and industrial strategy)

· Information Commissioners Office (regarding potential information rights issues)

· Trade Unions (representing training providers, public sector workers, and teachers)

· Business groups (e.g. Institute of Directors and Confederation of British Industry)

· Third Sector Groups (e.g. Springboard, Nesta, Joseph Rowntree Trust)

This initial period of engagement could build positive relationships, and would further mean that mayors (on behalf of the team developing the platform) could quickly escalate and address issues that emerged, whether they were administrative, ethical, or policy-related.

Thirdly, mayors need to have the personal leadership capacity to embark on these projects as adaptive challenges. The significant changes in cross-government culture, citizens’ relationship with government, and the relationship between business and society would result in a sustained period of disequilibrium. It is likely, for the reasons outlined in the quote at the beginning of this piece, that mayors would personally be under attack for the digital innovation they were introducing. To a degree, the agile, modular approach and proactive stakeholder engagement outlined above would mitigate this challenge. However, mayors would still need the resilience to remain open to failing forward in the face of public criticism. It has been regularly highlighted in class discussion that agile doesn’t prevent failure, it just means you can fail quickly and in a way that is less impactful, and then learn from mistakes to drive improvement. Ultimately, this is a difficult political pill to swallow — a modular roll-out with numerous hiccups at the early stages of your period in office can be more challenging than a behind-the-scenes, over-budget and behind-schedule IT programme that is yet to be completed even after your second term finishes. Whilst a long-term legacy of non-delivery can be embarrassing, a regular trickle of negative courage can lead to death by a thousand cuts. This means that mayors should be aware of the leadership capacity required before embarking on these programmes, and further emphasises the importance of communicating clear public value at the outset. As well as their own personal leadership, mayors should ensure that they have talent within their teams that are both familiar with agile methodology and have the leadership capacity to exercise similar level of resilience. This may mean audits of in-house capability, training programmes around managing software development and implementation, and even hiring externally if required to initiate the project.

Healthcare.gov is by no stretch of the imagination the only ‘digital disaster’ from the public sector — it’s just one of the only ones we’ve ever heard of. Yet this publicity is to be welcomed — the lessons it provides, and the perspectives of its protagonists, allow the extraction of key insights that, when applied to the strategic triangle, facilitate a more general application to UK local government. However, learning these lessons is a necessary but not sufficient condition. To be successful, leaders embarking on digital innovations will still need to be alert, attuned, determined — and unfalteringly curious.