What would a next-generation organizational structure look like for governments?
Photo by Alain Pham on Unsplash
The success or failure of organizations is critically dependent on how they are organized. Governments today are large bureaucracies with organizational structures that have changed little over the years. They have also expanded vastly over time. Under its first president, George Washington, the US government started with three cabinet departments, State, Treasury, and War. The federal government employed 3,905 employees in 1802. It now has 15 executive departments, 2.1 million civilian federal workers, and 15 levels of pay. According to the World Bank, public sector employment accounts for 16% of total employment globally and 37% of formal employment.
The advent of digital technologies calls for a very different type of organizational structure. To illustrate the point, the Bank of America, founded in 1924, employs 213,000 people to serve approximately 66 million clients. In contrast, Ant Financial, founded in 2014 as a digitally native organization, has 16,600 employees but counts 731 million monthly active users. Additionally, Ant Financial offers a broader range of services to its customers than the Bank of America. MYBank, a subsidiary of Ant Financial, follows a 3–1–0 system for processing loans. It takes customers three minutes to apply for a loan, one second for approval and involves zero human interaction.
The way governments leverage digital technologies can make a big difference to citizens. For example, tax authorities in Norway send taxpayers a pre-computed tax return every year, which they can either confirm or contest. Norway makes every individual’s income, net worth, and tax payments publicly available on a government web portal. In contrast, the tax compliance burden for US taxpayers in 2020 is estimated at 6.081 billion hours with a financial burden of $304 billion, or nearly 9% of total taxes collected by the Internal Revenue Service (IRS). The release of President Trump’s tax returns continues to be the subject of lengthy litigation.
With the rapid growth of the digital economy, we need to confront increasing levels of complexity and compressed decision cycles. The siloed structures of the government and the sheer cognitive load of complex decision-making require us to radically rethink how governments are organized. Many of the challenges faced by societies today require a whole of government response. Dealing with pandemics, climate change, disruptions in labor markets, and rising inequality calls for cross-sectoral and cross-functional responses. Existing functional hierarchies are ill-suited to deal with such challenges.
Features of a futuristic government
While thinking of next-generation organizational structures for governments, we should consider the features of a modern twenty-first-century government. Ideally, such a government would have:
1. Extreme client-focus in the design and delivery of hyper-personalized, proactive services with a minimal burden of compliance.
2. Seamless capture and integration of data across the government and the private sector.
3. A liquid workforce that can be rapidly reconfigured combining both internal and external expertise.
4. Continuous learning and horizon scanning to progressively improve capabilities within government.
5. A culture of inspiring, meaningful, and satisfying work.
Realizing these objectives requires a digitally native organizational structure for the government. This structure must be user-centered and data-driven. Processes will need to be reconfigured and embedded in software. Artificial intelligence should be used to design and deliver hyper-personalized and anticipatory/predictive services. The composition of government expertise will also have to change. It will be necessary to develop capabilities to algorithmically configure cross-functional teams comprised of internal and external staff. Continuous learning will have to become an integral part of work, not an afterthought.
There are several organizational models available, each with its advantages and disadvantages. The classical Weberian bureaucracy can be efficient within functional boundaries but has difficulties dealing with cross-boundary coordination. We have the matrix organization which tries to overcome this problem. Matrix teams were first used in 1947 by General Chemicals and in the public sector by NASA in the early 1960s. The structure involves adding dotted lines of authority to vertical reporting lines for cross-functional teams. This structure became necessary as NASA had to tap expertise from different departments and functions to solve complex engineering problems. The matrix form of organization has its merits in cutting across silos but suffers from complexity, and the vertical lines of authority often prevail over horizontal dotted lines.
More recently, companies such as GitHub, Valve, and WL. Gore have implemented a “flat” organizational structure. For example, Valve’s Handbook for New Employees states, “we don’t have any management, and nobody “reports to” anybody else. We do have a founder/president, but even he isn’t your manager.” Flat organizations may not be genuinely flat and may conceal power structures and absolve individuals of accountability. In 1972, feminist author Jo Freeman wrote an essay, The Tyranny of Structurelessness, where she asserted that “there is no such thing as a structureless group.” Regardless of their nature, any group of people that gathers for any purpose, for any time, inevitably structures itself in some way.
The late Tony Hsieh introduced the concept of holocracy in Zappos with no leaders or subordinates. The terms “position” and “employee” were replaced by the term “role.” Each role had a distinct purpose, and employees were empowered to make independent decisions that ensured and improved the performance of that role. While holocracy empowers employees, the transition from a traditional model of governance can be lengthy and complex. 14% of employees left Zappos after the adoption of holocracy.
Smaller, multidisciplinary teams, squads, pods, or cells operating with some degree of autonomy can help boost agility and responsiveness. Their focus on specific objectives and bringing together disparate skillsets can help break down functional silos and hierarchies and accelerate innovation and citizen engagement.
Companies have long considered Conway’s law and Dunbar’s number while setting up teams and designing organizations. According to Conway’s law, solutions or software produced by a team or organization, and its underlying code, will tend to reflect the group’s or company’s communication and organizational structure. Dunbar’s number sets a maximum of 150 people with whom effective collaboration is possible and five people who can form a close circle of intimate relationships.
The flip side to working with teams is that members tend to develop strong team loyalties at the expense of organizational goals. It can also be challenging to scale up teams in large organizations.
General Stanley McChrystal introduced a new structure for the US Special Forces Command, elaborated in his book, Team of Teams. To overcome the constraint of scaling up compact teams, General McChrystal started an exchange program where members of different teams would spend at least six months in other divisions. A Navy SEAL, for example, would be assigned to an Army Special Forces team, and vice versa. This exchange helped connect teams and bring greater cohesion in information sharing. It significantly improved the success of operations against terrorist organizations in Iraq.
In addition to these models, we also have the infocracy and algocracy models of organization. An infocracy is a type of organization that operates based on information flows. The introduction and use of digital technologies and information systems in governments can be termed as an infocracy. In an algocracy, authority is exercised through algorithmic systems. Algorithm-based decision-making can potentially improve human sense-making as it can help make decisions more rational, fact-oriented, and reliable. Predictive policing is an example of an algocratic system.
Next-generation organizational models for governments
While thinking of next-generation organizational structures for governments, we may need to consider not one but multiple options that could co-exist with each other. No single organizational structure can fit all requirements within the government. A futuristic organizational structure would be one that can quickly shapeshift depending on changing needs. Think of an organization that can be rapidly reconfigured along the dimensions of space, time, people, technology, and context. A unified data fabric would be critical to allow for such a shapeshifting organizational structure.
Organizing data by governments
Successful digital companies try maximizing data capture in real-time from each customer interaction. Governments need to do the same. Data should be systematically captured and connected across all parts of the government. The data function should ideally be centralized and anchored at the highest level in the government with direct oversight over data capabilities across the government. Singapore’s GovTech is anchored in the Prime Minister’s Office and has staff embedded in each government agency.
Besides creating a unified data function across the government, it will also be essential to ensure data access from all parts of the government through standardized APIs (Application Programming Interfaces). A flexible API-based approach can help data flows across the government. Amazon has demonstrated the value of modern data and technology architectures in enabling rapid growth, responsiveness, and scale. In 2002, Jeff Bezos issued the now famous API mandate. The mandate required all teams to expose their data and functionality through service interfaces designed from the ground up to be externalizable. This mandate enabled a common backbone connecting AWS, Zappos, Audible, Kindle, and Prime, enabling Amazon to scale up its operations efficiently and quickly.
Similarly, empowering each part of the government to share real-time data through API calls can improve collaboration and result in more informed decision-making at the edges closer to citizens and businesses.
Using digital twins for organizational design
Digital twins combine various technologies to create digital versions of real-world places, objects, people, and processes. Creating digital twins of government processes and employees, for example, can prove an effective means for an organizational redesign. A digital twin of an employee can capture, for example, the skills, qualifications, work experience, aptitude, and interests of the employee. This data can be used while constituting cross-agency, cross-functional, and cross-sectoral teams. Digital twins can be used to go beyond organizational boundaries to tap external expertise as well. According to the Institute for the Future, going forward, we will see pop-up enterprises where algorithms help connect expertise within and across organizational boundaries. The use of algorithms could create a liquid workforce that can be rapidly reconfigured depending on emerging needs and contingencies.
Government as a learning organization
Training of staff is often seen as a low priority for many government organizations. Governments will have to focus on constant skilling, reskilling, upskilling, and multi-skilling their workforce to prepare for future challenges. It will be critical to also equip the workforce with the skills necessary to work with data, machine intelligence, and cutting-edge technologies.
Governments, therefore, need to create digitally native organizational structures with the ability to shapeshift based on situation and context. An appropriate data architecture combined with digital twins and a learning organization can enable such organizational agility. Constituting teams to focus on mission projects like averting the next pandemic or achieving a significant reduction in energy consumption or greenhouse gas emissions can inspire teams and make them more capable of dealing with the grand challenges of development.