3. The application of data in the public sector to generate public value

Collecting and generating

This is the starting point for the application of data within government. The data accessed by public servants can take many forms and come from multiple sources. They could involve the consumption of a third party’s published dataset, whether as open government data (OGD) or via an application programme interface (API). They could be using the data generated by another piece of technology, perhaps an Internet of Things (IoT) device. They could be data requested as part of the design of a service, like forms collecting information from the public or logged in customer relationship management software following subsequent follow-up enquiries. They could be data produced as the output of government activity, such as one that involved the creation of government contracts. They could also be data held by private sector actors working in conjunction with the public sector to deliver goods and services.

While much of those data are generated by government activity, it is also possible that this first stage in the cycle involves non-governmental sources. This highlights the importance of universal standards in data collection and handling in both the private and public sectors (and forms part of the data infrastructure and architecture discussions in Chapter 2). Although the implications of data involved in this stage are internal to the public sector and influence internal decisions, having good-quality data helps in its reuse in later phases. As these decisions shape the interactions through which data are collected and lay the foundations for future use, this phase of activity defines the citizen experience of government services (Welby, 2019[6]). As such, public value is a passive by-product rather than, as with later phases, something directly generated from the use or reuse of the data.

Storing, securing and processing

Once data are identified, collected and generated, they must then be stored, secured and processed. This phase of the process is highly important to the discussions covered in Chapter 4 on the role of data and public trust. With data collected, decisions must be taken about how to store it, assess its quality (including any issues around bias), catalogue it and cleanse it. These steps are essential not only in ensuring the confidence of citizens in the public sector’s capacity for the proper handling of data, but also in providing a solid basis for the subsequent phases of the cycle.

This point of the cycle focuses on the behaviours and activities of those within the public sector in addressing the architectural and infrastructural challenges of high-quality data provision discussed in Chapter 2 considering data governance models. As such, the implications of data involved at the stage of storing, securing and processing are internal to the public sector, influencing internal decisions. They do not create new public value in terms of maintaining trust in government (as covered by the discussions on data-related citizen rights in Chapter 4). Such considerations should especially be a priority for those with responsibility for personally identifiable information, whether they’re in the public or private sector.

Sharing, curating and publishing

The third phase of the government data value cycle considers the way in which data that have been stored, secured and processed are shared, curated and published. At this point, legal context and constraints may dictate how readily requests for access and agreements to share can be handled for data that are not openly available. Where there are explicit efforts to support sharing, curating and publishing, the availability of data interoperability platforms and licensing those data that are made available through open data websites should be a priority with the earlier stages of the cycle, thus ensuring the latent quality of data.

The experience of the countries whose participation in the comparative study used as a basis for this report reflects a variety of approaches in terms of how explicit the expectations might be for sharing data between government institutions, especially in order to avoid citizens providing the same information to multiple parts of government.

Figure 3.2. Does your country have an explicit requirement for public institutions to share the data they produce with other public institutions?

Source: Based on information provided by six OECD countries in response to the Questionnaire on the State of the Data-driven Public Sector in OECD Countries, Question 7: “Does your country have an explicit formal requirement (i.e. written guidance provided in an official government document: laws, directives, regulations, guidelines, action plans, executive order, other) for public institutions to share the data they produce with other public institutions?”.

In Korea and Portugal, there is a legal expectation for the proactive sharing of data. In Korea, the Electronic Government Act provides the legal basis for preventing the duplication of data collected from citizens, thereby requiring the sharing of relevant data within government as a matter of course (Government of Korea, 2017[7]). In Portugal, a similar piece of legislation implements the “once-only” principle, meaning citizens do not have to supply same document twice to the government. Portugal’s Interoperability Platform for the Public Administration facilitates the exchange of service-related information within government and, following the Resolution of the Council of Ministers 42/2015 of 19 June, extends this to private sector suppliers (Presidência do Conselho de Ministros, 2015[8]).

In Denmark and Sweden, only selected datasets are available for all public institutions to access in a proactive way. Denmark, similarly to Argentina’s data as a service approach (see Box 2.6 in Chapter 2), has produced a comprehensive set of resources that detail how to construct and publish data models.1 These resources allow for a common language for discussing Danish data and simplify its reuse by those wishing to do so. Sweden has had a long-standing legal commitment to sharing data, with legislation passed in 1998 detailing which pieces of information are to be shared and the circumstances under which to do so (Finansdepartementet S3, 1998[9]). In Ireland, different public institutions come to ad hoc agreements about sharing or accessing data. However, there are ongoing efforts to expand the legislative framework on the reuse of data to establish a more formal model for this. The United Kingdom does not have any formal requirements for data created by one organisation to be shared by another, with the implication that data-sharing happens only due to ad hoc requests rather than a strategic approach to interoperability and ease of sharing.

Almost all these countries are taking steps to ensure the publication of data through open data websites providing access to a wide range of datasets. The focus on producing good-quality data early in the government data value cycle to support its reuse later either elsewhere in government or as OGD is discussed in Chapter 2. However, the internal discoverability of these data is less sophisticated, with only Korea and the United Kingdom developing a single data inventory for government to simplify the ease with which policy makers and service teams can access these data for reuse. Denmark, Ireland and Portugal have plans for this in the near future while Sweden does not.

Internal reuse and discoverability of data is not necessarily limited by the absence of a single data inventory. Neither Denmark (Box 3.1) nor Portugal have such an inventory, but along with Korea they have taken practical steps to consider the needs and structure of base data registries. Such canonical records of information provide a foundation for policy and service delivery that simplify the acts of sourcing, and curating, vital datasets. These are also being developed by the United Kingdom, but without explicit legislation to underpin them.

At this stage in the process, the potential impacts of using data are no longer limited to internal public sector stakeholders, but begin to touch those external to government who can begin to use shared data. The first two phases of the government data value cycle shape the citizen experience and are influential in securing trust in government and maintaining public value, but this stage, and the one which follows, sees the potential to generate new public value from the activity of those with whom data are shared to explore policy impacts and service opportunities.

Use and reuse

The final phase of the government data value cycle focuses on the use and reuse of data, and offers the clearest opportunity for generating visible public value. The previous steps in the cycle are often hidden from view. Yet, they should not be neglected, because if there is not an effective approach to them, efforts to derive value from the use and reuse of data will be undermined due to poor quality data, incomplete sources, unreliable access and barriers to sharing. Therefore, although the public value from the use and reuse of data is visible, it is only because it is supported by a broader ecosystem of data governance, as discussed in Chapter 2, creating the conditions in which data can be successfully applied.

It is well understood that opening data allows for external stakeholders to generate public value on the basis of these data but, in a data-driven public sector, equal weight and attention are given to the internal experience. By improving the management and application of data at each stage in this process, policy makers and public officials can increase their effectiveness by enhancing their data capabilities and ultimately generating greater public value. The second half of this chapter will discuss in more detail how the use and reuse of data can bring that value and it is suggested that for a DDPS built on the government data value cycle, the value is generated from the outcomes of its final two stages. This can include gathering insight on existing policy activity; understanding the issues facing stakeholders; foreseeing new trends and needs; delivering higher quality services; designing and adapting innovative approaches; monitoring ongoing implementation activities; and managing the resources being used to address a particular challenge.

Gross domestic product

One of the traditional ways of measuring the health and happiness of a country is to use gross domestic product (GDP). Using the size of a country’s economy as a proxy for the wealth of households and therefore the well-being of society has been a long-standing mechanism for understanding whether or not policies are proving successful at meeting political objectives. Although citizens do not necessarily directly acknowledge the impact of increased GDP, it is one of the areas in which the public value of data could be understood in terms of the financial benefits it might produce.

As discussed earlier, financial value is often the easiest way of defining the value of a tangible asset. In the case of data, that financial benefit has often been cited in terms of making the case for its greater use or release to the public. In Spain, the National Observatory of Telecommunications and the Information Society (ONTSI) publishes a yearly assessment of the “infomediary” sector of the economy, businesses that exist because of data. In 2016 ONTSI calculated that by 2015 OGD had generated an impact between EUR 600 million and EUR 800 million. In 2019, the Multisectoral Information Association reported the turnover for these businesses at EUR 1 796 million (ASEDIE, 2019[25]; OECD, 2018[26]).

The financial value of data has been a priority when it comes to making the argument for releasing OGD. In the 2017 OECD Open, Useful and Re-usable data (OURdata) Index (OECD, 2017[27]), 32 out of the 34 participating countries released government data in order to create new business opportunities for entrepreneurs and students. Certainly one of the great hopes of the OGD agenda has been the ambition to use these data to stimulate economic growth through the creation of new industries and business models and the jobs that would follow (OECD, 2018[26]). In the United Kingdom, the opening up of geospatial data through the Ordnance Survey’s OS OpenData™ initiative was identified as creating a benefit of GBP 8.1 million to GBP 18.2 million in productivity gains and GBP 4.4 million to GBP 8.3 million in additional real tax revenues (ConsultingWhere Limited and ACIL Tasman, 2013[28]). This is fairly modest compared to Canada’s 2015 assessment of the economic impact on the country’s GDP of opening geospatial data to be CAD 695 million (Hickling Arthurs Low Corporation, 2015[29]). Nevertheless, even though that figure sounds significant, it represents just 0.04%, so the GDP financial benefits of data perhaps remain somewhat elusive.

Government efficiency

The efficiency and effectiveness of government services is another means by which public value can be understood. In keeping with the definitions of public value advanced earlier, the increased efficiency of services can be directly experienced by citizens in their day to day lives as a result of digital government approaches (Welby, 2019[6]). Sometimes the citizen will directly experience the improvements to a service following the application of data. In other cases, a citizen may not be aware of the transformative impact of data on their services because government is able to proactively respond to their needs before they make a request or obviates the need for them to supply information because it has been accessed elsewhere in government. Such benefits result in time saved for citizens, but also in reduced back office processing and handling time for public servants too, resulting in increased throughput, fewer errors and greater cost efficiency from an operational point of view. Pollock (2010[30]) concludes that as well as the benefits in generating new products and services built with public sector information or those derived from developing complementary services and consultancy, data can provide indirect benefits, such as reduced transaction costs and increased efficiency of up to GBP 600 million per year.

The United States President’s Management Agenda (United States White House, 2018[31]) identifies a long-term vision of recognising government data as an asset. Public value is seen in modernising the federal government to improve the ability of agencies to deliver mission outcomes, provide better public services and steward taxpayers’ money.

Public sector efficiency gains are considered by several academics whose work has touched on the value of publishing OGD and is discussed in the OECD Open Government Data Report (OECD, 2018[26]). This aspect of public value is seen in how making OGD freely available reduces the overheads for organisations in disseminating this information as well as increases the more timely access to information. For others, efficiency is seen less as the external value of improved services and more in terms of cost savings and the improvement of internal processes. At a country level, the recognition that OGD can increase the efficiency of services was acknowledged in the 2017 OURdata Index (OECD, 2017[27]), with 31 of the 34 participating countries stating an intent to improve public sector performance through the release of data (OECD, 2018[26]). In Luxembourg, many datasets have been opened for public use, but its main users have been within the public sector itself. Historically, a lot of geodata (cadastre, aerial imagery, topographic maps, address register) were licensed at cost between different public sector bodies, but by removing this cost and opening the data, they can be more widely used, generating increased public value.

The McKinsey Global Institute has calculated potential efficiency gains at EUR 250 billion per year within the European public sector (Manyika et al., 2011[32]) and between USD 35 billion and USD 95 billion per year in the United States by 2020 (Lund et al., 2013[33]). The private sector can also provide a guide in terms of defining the potential productivity gains in a data-driven approach with McAfee and Brynjolfsson (2012[34]) finding that such companies were, on average, 5% more productive than their competitors.

Social value

The previous two definitions of public value have tended to be rooted in financial measurements, but if public value is only defined as GDP and focused on efficiency, then it would be simply “economic value”. A 2006 study by the Center for Technology in Government (Cresswell, Burke and Pardo, 2006[35]) found that the ability of government to realise the full value of money spent on IT investment cannot be solely measured in terms of financial results. Instead, the experience they found in five case studies from Austria, Israel and the United States saw that the value of government spending existed in the broader political and social returns to the public at large rather than just the internal value to government operations (as discussed above).

In the United Kingdom, HM Treasury issues guidance to public servants on how to construct business cases. As well as setting out the expectations for how investment might be secured, The Green Book calls on public servants to focus on the generation of “social value” in its framing of any costs and benefits (Box 3.3). Together with the public value framework (HM Treasury, 2019[36]), a tool for understanding how well public money is turned in to policy outcomes, these are important contributions from a usually financially focused body to broaden the perspective on defining value.

The OECD has had a long-standing effort to develop measures for understanding both trust and well-being as part of the Better Life Initiative. The well-being framework (Figure 3.4) shifts the focus away from aggregated economic conditions towards a definition of social progress. It does so by looking to identify the impact of public policies in terms of outcomes rather than purely because of inputs and outputs (OECD, 2017[38]). This results in a blended approach that considers the objective, and subjective, aspects of life and an approach that considers the distribution of experience across a population, thereby incorporating questions of inequality and sustainability into the definition of well-being.

The New Zealand government has taken the radical step of rethinking its approach to public spending and is moving away from the pursuit of GDP growth to instead focus on well-being. The country’s 2019 budget requires all new spending to go toward five specific well-being goals: 1) bolstering mental health; 2) reducing child poverty; 3) supporting indigenous peoples; 4) moving to a low-carbon-emission economy; and 5) flourishing in a digital age. It uses a living standards framework that is built from citizen level data and tracked with concrete indicators. This builds on a long-standing interest in that country in exploring the possibilities of “social investment” (Acquah, Lisek and Jacobzone, 2019[39]).

Figure 3.4. The OECD well-being framework

Source: OECD (2017[38]), How’s Life?: 2017 Measuring Well-being, https://dx.doi.org/10.1787/how_life-2017-en.

Engagement and participation

Central to the definitions of “public value” referenced above are citizens and their understanding and participation in the value that is created. Therefore, the application of data to generate public value is not only about making the data available, whether as OGD or in its role in providing public services, but also ensuring that it can be understood and seeking the engagement and participation of citizens.

The ambition for increased transparency and greater accountability of government through the publication of OGD has been an important factor in its success (OECD, 2018[26]). The publication of OGD and the communities that have emerged around particular themes and policy sectors demonstrate that it has created public value in terms of citizen engagement and participation. Chapter 4 will go into more detail about the role of data in building public trust, and highlight ways in which governments can provide effective consent mechanisms for helping citizens to understand the specifics of how their data are being used. While this is important at a technical level, there is also a role for increasing the data literacy of the public and exploring innovative ways of incorporating their perspectives.

An emerging trend in the assessment of public value that looks at engagement and participation and is relevant to the application of data are “mini-publics”. “Mini-publics” are “a group of citizens, demographically representative of the larger population, brought together to learn and deliberate on a topic in order to inform public opinion and decision-making” and may take the form of a citizen jury, a consensus conference, deliberative poll or citizen assembly (Breckon, Hopkins and Rickey, 2019[40]).

The example of Ireland (Box 3.4) offers insight into how this process blends the application of data in evidence-based policy making with the generation of public value. Randomly selected citizens have been brought together to work through the issues and implications of politically challenging and sensitive topics. This necessitates good facilitation between citizens and experts in order to help engage with evidence and data. It also requires a commitment on the part of the public to fulfil their citizen duties in grappling with complicated issues and considering the broader, societal picture. Where such processes are understood to reflect public participation (and so fulfilling the aforementioned definitions of “public value”), people see any subsequent activity as delivering value, even if it does not follow their personal preferences (Talbot, 2011[24]). However, this means that the value of the exercise owes a lot to the extent to which governments commit to giving “mini-publics” power over the political agenda, similar to the way a judge has to behave in line with the pronouncement of a jury in court.

While “public value” can be easily understood at a conceptual level in terms of making an improvement that is understood and recognised by members of the public, it is clear that there are various ways in which that value can be generated. This section has considered the possibility of measuring “public value” through growth in GDP, government efficiency, social value, well-being, or engagement and participation.

There is no single approach for defining “public value” in the application of data, with activities under each of these categories offering different opportunities and routes to realising value. To help identify the most suitable means of defining value, it is important to clarify the reason for seeking to create value, the problem one is trying to solve and the need that needs to be met. One way of ensuring the need is understood from several perspectives is to involve the public and to bring together a diverse, multi-disciplinary team. Delivering successful outcomes and responding to the challenges facing diverse communities requires the collective involvement of external stakeholders as well as those from across policy, delivery and operational teams. This can help to avoid situations where public data are put to uses that diminish public trust and legitimacy while unlocking revolutionary approaches to how governments think about providing services to citizens and how they measure efficiency in service delivery as well as user satisfaction (Welby, 2019[6]).

Understanding, defining and measuring the value of data as an asset and its contribution to public value relies on ensuring the government data value cycle is acknowledged and each step in the process is clearly mapped and measured. Establishing the baselines and benchmarks for the use of data within, and between, government organisations is an important prerequisite for identifying value. Nevertheless, in developing business cases and securing political commitment, often this value needs to be defined on a financial basis. The United Kingdom’s approach in seeking social value, and discussed in Box 3.3, may be helpful in showing how both financial and non-financial measurements might still be defined in cash terms.

The data governance approach discussed in Chapter 2 is fundamental to the success of DDPS efforts. Ensuring the leadership, capacity to deliver and necessary legal frameworks as well as architectural and infrastructure approaches that enhance the data value cycle should be a priority. Getting this right makes it possible to explore ways for applying data to generate public value, with the confidence that the necessary foundations are in place to achieve successful outcomes and provide a robust measurement of its value.

Evidence based policy making

The first area to consider in terms of the “anticipation and planning” phase is the use of evidence to design policy interventions that respond to forecasted challenges. This is not the same as the forecasting activities (discussed later), which attempt to predict whether or not a policy would be effective. It is about the approaches which countries take to experiment with a particular approach and consider as wide a spectrum of intervention as possible to shape future policy interventions, informed by the availability of data. This is not solely the domain of governments and public servants, as think tanks, journalists and academics are some of the actors involved in thinking through how to respond proactively to the needs of society based on data.

One of the most critical factors in securing the use of data as evidence in pursuit of policy making in the design, implementation and evaluation phases is establishing standards for that evidence. The OECD’s Expert Group on Standards of Evidence has proposed six characteristics (Box 3.6) that should be implemented in order to support evidence-based policy making.

Some countries have developed guidance for public servants on using OGD in the development of policy. According to the 2017 OURdata Index survey, Austria, Colombia, the Czech Republic, Finland, France, Japan, Korea and the United  ingdom have all developed overarching guidelines on how best to use OGD to inform policy-making processes (OECD, 2017[27]). While these countries are providing guidance for public servants, Denmark, Norway, Spain and the United States are using OGD to enhance the public debate around policy and develop co-responsibility for responding to its challenges. Moreover, Chile, Colombia, France and Israel are hoping the release of OGD will facilitate the crowdsourcing of solutions for public policy problems by tapping into the collective intelligence of the public (OECD, 2018[26]). In Mexico, the 2013-18 National Digital Strategy stresses the importance of OGD in contributing to achieving policy outcomes – land management improvement, the digital economy, and natural disaster prevention and mitigation (OECD, 2016[45]).

Korea has developed a standardised model for analysing “big data” within the public sector so that data generated in one part of the public sector can be compared with what is generated elsewhere. The use of the resulting models informs policy making in 18 areas, with 320 standardised models having so far been provided to 175 organisations. There are plans to introduce similar models in categories including citizen services, tourism, transportation, closed-circuit television (CCTV) and public housing. Such standardisation minimises local differences in the analysis which takes place between different institutions, and in particular central and local governments. This allows policy to be informed with a more accurate and comprehensive understanding of a given dataset.

In Denmark, significant effort has gone into the handling of education data. A databank and data warehouse provide institutions, regions, municipalities and the public access to a number of predefined reports and graphs with statistics. There is also the possibility for people to define their own reports on certain aspects of the data it contains.

In Ireland, the biggest use of data for policy making is in mapping. The Ordnance Survey of Ireland’s Geohive service provides easy access to publicly available geospatial data. Combined with data from other sources, these mapping data underpin analysis of housing trends and flood risk. A further development in Ireland is the creation of Pobal, a website and support service providing information on deprivation profiling in a particular area, details of local childcare services and information about other funded services available for people to access. This resource is not only for policy makers, but for citizens and community organisations as well.

One of the most powerful opportunities offered by the application of data in evidence-based policy making is in its potential to support a change in approach. In the Flemish region of Belgium, a spending review was carried out with the intent of developing policies that would create jobs in order to meet the needs of the communities. However, the analysis showed that the greatest need was in serving elderly citizens with requirements for domestic services rather than for jobs.

Anticipatory governance

A second area during the “anticipating and planning” phase in which the DDPS approach can generate public value is that of anticipatory governance. Anticipatory governance describes systematic efforts to consider the future in order to inform policy decisions today. In this context, governments respond proactively rather than reactively, based on knowledge and evidence rather than experience and protocol. The use of data leads to the better detection of societal needs as they emerge and improves predictions for future needs. This is a particularly important activity in the context of innovation. The OECD’s Observatory of Public Sector Innovation has recently initiated an “Anticipatory Innovation Governance” project to analyse the challenges and opportunities in this space.

Data-enabled prediction and modelling techniques support governments in anticipating societal, economic or natural developments that are likely to occur in the future. They may also capture early warnings and better assess the need to intervene. Additionally, governments that are built on strong data foundations are able to deliver proactive government services that anticipate the needs of citizens and obviate their need to interact with the state, without necessarily realising that has happened.

This reflects macro anticipation at a societal level in terms of understanding emerging trends and carrying out “predictive” activity to inform long-term planning. However, it also reflects the micro anticipation in terms of an individual and their needs. At that level it is not about predicting what an individual is going to need based on their historic choices and presenting them with the existing solution, but about a deeper transformation of the design of the state based on a deep understanding of life events and the sort of interactions that are associated with a given set of services.

Anticipatory governance falls into two categories: forecasting and foresight.

Designing services and policy interventions

One of the most compelling opportunities for DDPS is the way in which the application of data can reshape the opportunities for designing better policy interventions and services through a deeper and more rounded understanding of the needs of citizens and engaging citizens as co-creators of value. While this is part of “anticipation and planning” activities, there is a clear linkage between the design and the “delivery” phases.

As countries develop their data governance models (as discussed in Chapter 2) and embed a more sophisticated understanding of the government value data cycle, the use and reuse of data within government becomes more sophisticated. The flow of data around government, its cataloguing and the easing of access to it means that there is the potential for more deeply integrated service design opportunities. Furthermore, as citizens are invited to participate in the design of services and contribute to the user research establishing their needs, services begin to reflect the actual needs and circumstances of citizens.

In the United Kingdom, the Government Digital Service has developed a set of mandatory guidelines for assessing and assuring the quality of any public-facing government services. This Service Standard contains 14 points and is administered through a series of assessments at different stages in the implementation cycle. The role of qualitative and quantitative data in understanding the needs of citizens and their behaviours when interacting with public services is highlighted throughout. Particularly relevant to the role of data in “anticipating and planning” the design of services is the requirement to consider the role of qualitative and quantitative data in ensuring that the responsible teams have understood the needs of their users and are subsequently able to respond to them on an ongoing basis. A similar standard exists in Canada with one of its principles requiring Canadian public servants to be good data stewards and collect data from users only once and reuse wherever possible whilst also ensuring that data are managed securely, and in ways that facilitate reuse.

Another important aspect of the design phase of planning for both the “delivery” and the “evaluation and monitoring” of services and policy interventions is the initial effort required in benchmarking performance and measuring success. In order to generate the maximum public value, it is essential to clearly define the purpose of its application; but this is only half the exercise. The other half is identifying how to measure the value and judge the success, or otherwise, of a given intervention.

Improving public services that respond to citizen needs

A DDPS approach enables a closer working relationship between policy design and service delivery activities with a resulting shift from top-down implementation of public services to a user need led approach to design and delivery, based on an end-to-end understanding of a particular service journey, which can consequently increase its reach and effectiveness.

Inclusive user-driven approaches towards government service delivery can be strengthened through the forward-looking approaches highlighted in the previous section on “anticipation and planning”. The analysis of data produced by the behaviour and characteristics of existing users can facilitate predictive activities identifying potential new users and/or emerging needs and problems. While this will inform the development of new services to address unmet needs, there is also the possibility of proactively enhancing the existing approach to address issues in the existing service. As a result, government services can continue to improve, responding to the rhythm of people’s lives and continuously evolving to meet their needs and expectations, in pursuit of delivering greater public value.

Much of the public value that can be generated from developing the use of data in the “delivery” phase of improving public services may take place in ways that are not immediately obvious to an external user. Addressing the regulatory, data architecture and data infrastructure challenges discussed in Chapter 2 offers the potential to transform interoperability of data between government organisations, thereby allowing the prefilling of information and making the once-only principle of data submission a reality. Such efforts can be particularly powerful in reducing the scope of what would be involved in bringing the “long tail” of analogue processes and services on line.

Indeed, the application of data to the delivery of services can accelerate innovation cycles. This is not only because the environment is conducive to experiment with emerging technologies, implemented on a strong foundation of reliable data, but also through more agile design approaches. The Lean Startup (Ries, 2011[51]) approach of build-measure-learn encourages a minimum viable approach to be put in front of users as soon as possible in order to learn and continuously iterate and improve. The embedding of a data-driven approach with this delivery culture means it will be easier for public institutions to know how citizens use their services, or how satisfied they are with them, learn from it, and adapt the design and delivery accordingly.

A further hidden value is the capacity for public service providers to deliver in a proactive way, i.e. to deliver a service before a user has had to engage the state. The aggregation of citizen data from different public institutions allows governments to build a more complete view of the needs of their citizens and address what otherwise might be a disjointed experience if a citizen had to interact with multiple parties. Citizens may view such an approach negatively, as indicative of a “surveillance state”, even if the intentions are noble. Therefore, the discussion about data and trust, explored in detail in Chapter 4, is essential in ensuring that where DDPS activities are driven by the personal data of citizens, the necessary steps are taken to ensure it protects, and enhances, the trustworthiness of government.

A more visible release of public value from a DDPS approach to the “delivery” of services is in the experience of citizens crossing borders. Data exchange between countries can facilitate new forms of international co-operation such as that seen in the European Union’s development of the eIDAS standard and the Mi Argentina mobile application making it possible for digital credentials to be recognised in multiple countries (Box 3.9). There are also further automated benefits, such as that between Estonia and Finland, where exchange of basic business register data is expected to facilitate the development of business between the two countries (OECD, 2015[52]).

Moreover, an understanding of demand for services themselves can provide a clearer view of the service delivery landscape in a society, allowing to better target specific channels to specific segments of the population to ensure no citizens are left behind. Box 3.10 provides examples from Portugal and Mexico.

Several of the countries that participated in the comparative project that provided the baseline of insights and evidence used to back up the analysis included in this report indicated that there was an overlap in their use of data to improve public service delivery and its use in developing evidence-based policy making. This close relationship between policy design and service delivery demonstrates the importance of recognising the relationships between each phase in the application of data.

In Denmark, Korea and Portugal, legislation has been passed to address the needs and structure of base data registries. These are a valuable enabler for the delivery of data-driven services because they limit the proliferation of duplicate data, provide a reliable source for the most important datasets, and reflect a coherent approach to the internal access to, and sharing of, data. In Korea, the “Public sector big data analysis projects” have been successfully scaled to bring benefits to different population groups. These have included citizen-visible services, such as understanding where best to locate electric vehicle charging infrastructure, through to the less public applications of data in delivery, like detecting suspected evasion of military service or tackling traffic accident-related insurance fraud.

Giving public servants the capacity to focus on meeting other needs

Improving the “delivery” of public services offers a very tangible indicator of an increase in public value from the perspective of citizens and businesses whose lives may be simplified. However, alongside those benefits are those more hidden from view in terms of their implications for internal staff.

One of the arguments that has been used to justify the digital transformation of government has been the opportunity to reduce the headcount of those working in frontline roles. As more and more services are available on line, the demand in person is expected to decrease and those roles will become redundant. This perspective is based on some incorrect assumptions. First, it expects a behaviour we cannot be certain of, i.e. that all users will shift to online platforms to access the services. Second, it takes for granted that the digital transformation is simply a quicker way of carrying out the same activities as before, rather than an opportunity to rethink the way in which services are provided. Thus, in a situation where services are transformed, the role of back office and frontline staff changes too. This can happen in several ways:

• By addressing the common case interactions, public servants are able to focus more time and energy on answering more challenging questions.

• As those who are comfortable with using digital services make the transition away from face to face services, those who continue to provide in-person support are able to devote more energy to the needs of those who might otherwise be digitally excluded.

• The increase of digital submissions and the reuse of data from across government reduces the potential for errors. This reduces the prospect of having to follow up and clarify something that can delay the fulfilment of the original service.

There are also possibilities for real-time analytical information to be used across service centres and telephone support lines to predict demand and manage staff time accordingly. For example, this could extend to developing the capability for reallocating public servants from one category of enquiry to another in response to unforeseen spikes in demand.

Communication and engagement with the public

The most recent Survey on the Organisation and Functions of the Centre of Government (OECD, 2017[58]) identified the most important objective of communication strategies. Figure 3.6 shows that communicating government actions was the highest priority, with only 8% of respondents having the objective of encouraging participation or promoting transparency, and a mere 4% reporting that they were attempting to improve links with citizens. This does not reflect the OECD Recommendation of the Council on Open Government, which calls for a “culture of governance that promotes the principles of transparency, integrity, accountability and stakeholder participation in support of democracy and inclusive growth” (OECD, 2017[59]). It also contrasts markedly with the strong characteristic of the OGD and open government movements where data are understood as a valuable device for fostering civic engagement in the design and delivery of public services and policies in order to strengthen a more people-driven and participatory form of democracy (OECD, 2018[26]; Ubaldi, 2013[4]).

Figure 3.6. Main objectives of the centre of government communication strategies

Source: OECD (2017[58]), Survey on the Organisation and Functions of the Centre of Government.

The potential for OGD to stimulate greater engagement and opportunities to communicate with the public is clear. In Chile, the Ministry General Secretariat of the Presidency put out a call to artists to develop original digital content with OGD and showcase the possibilities. In every corner of the globe, countries are hosting hackathons that encourage teams to come together and work with OGD in response to particular challenges; these include the Accountability Hackathon in the Netherlands, the CANDEV data challenge in Canada, the OpenGovDataHack and Hack the Plan in India, and GovHack in Australia, among others.

However, the experience of Sweden and HackforSweden (OECD, 2019[60]) is particularly inspirational. HackforSweden currently works with almost 100 Swedish government agencies and companies to promote innovation through open data with the explicit intent to create a platform for citizen-driven responses to societal needs and seeking collaboration to co-create public value. From humble beginnings in 2014, HackforSweden is now moving away from annual events to seek ongoing engagement and creativity from the digital ecosystem in the country with the mission to swap 200 people at an event for 2 million people connected through a platform in just 2 years (Hack for Sweden, n.d.[61]).

Outside of OGD, the potential for DDPS approaches to be utilised by governments to communicate with the public is currently underexploited. According to the comparative research carried out with the six OECD countries, only Denmark and Sweden could provide concrete examples of how policy initiatives were in place to encourage the use of data to engage societal stakeholders. This means that Korea, Ireland, Portugal and the United Kingdom have not made this a priority.

This is unfortunate because the use of data can support the process of debating policy options with visualisations offering accessible ways of presenting detailed information about particular problems. Concepts that are hard to communicate verbally or dense sets of numbers or co-ordinates can be transformed into tools that facilitate valuable discussion and engagement exercises. In the context of generating public value, Morabito (2015[62]) argues that communicating how data have been used in accessible ways can have a positive effect on citizens’ sense of achievement and their satisfaction with government services.

In Denmark, the approach to property tax was redesigned in order to provide citizens with a more transparent view of the system based on better data and the use of improved statistical methods. In handling the transition from old to new, data visualisations were used to provide clarity for citizens. By prioritising transparency and openness in the change, citizens were not only able to understand the policy in general terms, but the government invested in the capabilities and data literacy of citizens to understand how it affected them specifically.

Responding to emergencies, crises and developing situations

The final area of opportunity for adding public value associated with the “delivery” phase of the DDPS is the way in which public sector organisations are able to respond to emergencies, crises and developing situations.

Emergencies and crises by their very nature cannot be predicted. While their impacts can be minimised, it is impossible to be prepared for every eventuality at all times. This underscores the importance of the “anticipation and planning” stages in terms of the design and foresight activities for thinking through what might be needed in the event something went wrong. Indeed, the OECD Recommendation of the Council on the Governance of Critical Risks emphasises the importance of building preparedness through foresight analysis, risk assessments and financing frameworks, to better anticipate complex and wide-ranging impacts (OECD, 2014[63]).

In Seoul, Korea, the mayor has invested a significant amount of money in developing a 3.5 metre wide dashboard on the wall of the office providing a real-time feed of interesting data about the health and vitality of the city and its inhabitants, including the cost of living, demand for housing and the real-time budget position. This is helpful operational insight that can shape the political priorities and ongoing service delivery in the city. However, the effort has been made to consider how this resource might add value in the event of an emergency. During an emergency, the dashboard comes into its own, providing a live feed of the situation unfolding and giving the mayor direct contact with those on the ground (Crawford, 2017[64]).

In Singapore, the challenges of overcrowded and ageing transport infrastructure coupled with the country’s “smart nation” ambitions prompted the development of the Fusion Analytics for Public Transport Emergency Response (FASTER) initiative. With FASTER, anonymised location-based information is collected and combined with fare card transactions and video feeds from stations to identify overcrowding. When crowds are detected, additional buses are deployed and messages are relayed to customers allowing them to plan accordingly. These data provide detailed models of how users move through the city, helping the government to understand traffic patterns, how the transport network is used and issues to address with the network.

Policy evaluation

The first area in which the “evaluation and monitoring” of data can generate public value is in the process of evaluating the success, or otherwise, of policy interventions. In a well-functioning democratic society, the implementation of policy is scrutinised by a variety of actors. There are those who wish to ensure government resources have been effectively managed, those motivated by a desire to understand the impact of the intervention and those who may be looking for political opportunity to exploit. These actors have competing priorities, but the reporting of progress, particularly if done in public, offers an evidence-generating activity that can hold the tension of the spectrum between politically motivated and ideological claims at one end, and theoretically unbiased and rational evidence-based policy making at the other.

In this sense, the evaluation of policy after its implementation complements the use of evidence in the initial design and development of a policy intervention. The insights generated from evaluating policy are critical for iterating and developing new policy approaches and it is in responding to the evaluation of what has been done that governments can generate public value. Nevertheless, beyond the direct application of data in shaping government activity, “evaluation and monitoring” data serve an important role when shared in being used, and reused, to inform and equip politicians, journalists, academics and the wider public.

Increasing the amount of data associated with the outcome of a given policy allows for agile policy adjustments in the short term, but more importantly will generate better insights into the policy process in terms of accountability and learning in the mid- to long term. Those responsible for a given policy can establish whether their policies have had the desired effect or not and, if those data are published as OGD, so can other stakeholders. As a result, policy evaluation can turn into an open, inclusive and ongoing process rather than an internal, snapshot moment. The ability to reduce the lag between the design of a policy, its implementation and insights into its performance should not just have theoretical and conceptual value, but should provide the basis for rapidly informing “delivery” activity and remedying any unintended adverse effects (Höchtl, Parycek and Schöllhammer, 2016[47]). While the monitoring of performance might be prompted by a top-down desire for oversight and reporting on delivery, a DDPS is interested in how those insights can be analysed and, crucially, applied in improving performance based on a deeper understanding of the needs of the organisation and its users.

Carrying out retrospective evaluation and analysis is helpful in keeping an open mind as it encourages an ongoing learning from experience and stimulates efforts to adapt future policy as a result. Putting in place mechanisms to gather, and apply, new insights set an expectation that lessons will be learnt and new insights gained. Taking this approach enhances the ongoing quality of outcomes. The OECD/Bloomberg (2019[49]) report on Enhancing Innovation Capacity in Cities establishes that “those cities which evaluate … are better positioned to scale up innovative projects that improve operations, and less likely to engage in practices or projects that offer little return on investment”. Being open to the “evaluation and measurement” of impact is indicative of a more sophisticated appreciation for the role of data that will manifest itself throughout the policy and delivery life cycle.

It is therefore encouraging that Figure 3.7 shows five of the six researched countries (Korea, Denmark, Ireland, Portugal and the United Kingdom) implementing some steps to strengthen policy monitoring and evaluation throughout the government data value cycle. Denmark, Ireland, Korea and Portugal are particularly focused on how they might strengthen policy monitoring and evaluation by investing in the capacity to analyse the data. As discussed in Chapter 2 in the context of identifying the necessary conditions for successfully implementing the DDPS, the skills of public servants, and the public, are critical for maximising public value. All five of those countries have taken steps to make the sharing of data a focus, which as discussed at the beginning of this chapter is a critical factor in successfully creating a whole-of-government approach to generating public value through data.

Figure 3.7. In what areas has your country implemented policy initiatives to strengthen policy monitoring and evaluation through better data management and use?

Source: Based on information provided by six OECD countries in response to the Questionnaire on the State of the Data-driven Public Sector in OECD Countries, Question 16: “Has your country implemented policy initiatives at the central/federal level to strengthen policy monitoring and evaluation through better data management and use? (e.g. using data to enable continuous and/or automated policy monitoring and inform agile policy adjustments)”.

The previous section discussed the role of data in enhancing communication and engagement between the public sector and the public. Those examples discussed how the government to citizen relationship might be facilitated, but there are opportunities and possibilities for government to develop a greater understanding of the impact of policy on citizens. In a DDPS, organisations can analyse a wealth of input to evaluate policies from crowdsourcing initiatives, but they may also be able to gather these insights without necessarily asking for them directly, as discussed in Box 3.11. Similar to some of the practices mentioned in the earlier discussion on forecasting, data from a myriad of digital channels can be accessed and analysed, revealing the opinions and behaviours of citizens, and subsequently used to evaluate policy. When exploring these issues, governments must act with sensitivity and in ways that are informed by the framing of the debate in Chapter 4.

Operational performance

A second area for considering the “evaluation and monitoring” activity of the DDPS is in how data can transform the operational performance of government, leading to both more productive and effective use of public resources and developing a culture of continuous performance improvement. Measuring performance is an “evaluation and monitoring” activity that has to happen retrospectively. Nevertheless, applying its insights, it is tightly coupled with the quality of government “delivery”. This relationship once again highlights how important it is for clarity about the purpose behind the application of any data.

Public value can be unlocked in terms of the direct impact it has on the activity of government, as discussed in the “delivery” section, but also in the more hidden, structural activities of government. For example, data about the real-time status of physical assets can be used to better plan maintenance with the least disturbance to service continuity as possible (OECD, 2016[66]). The complementary case study “Data-Driven Human Resource Management” focusing on public employment and management highlights how people analytics might change approaches to recruitment, retention and the ongoing performance management of staff, as well as understanding how to improve the environment in which they work to ensure their productivity and well-being.

While an increasing amount of our transactional relationships are moving on line, safeguarding the built environment and our material conditions continues to be an important function of government. The application of data can allow for targeting scarce resources, such as focusing policing on areas with high criminal activity (Höchtl, Parycek and Schöllhammer, 2016[47]).

In Korea and Portugal, entrants to competitions focused on the use of data have been encouraged to target improvements to the efficiency of government operations. One of the success stories from Korea is the development of a data model for identifying businesses most likely to present a risk. Twelve months after its introduction, there had been an increase in the proportion of inspected workplaces violating the Labour Standards Act from 40.9% to 77.7%, almost doubling the impact of inspection staff in terms of public value. A similar impact was achieved in the United States, where the city of New York’s Department of Buildings pooled data with 19 other city agencies. The outcome of that collaboration was a five-fold return on the value of an inspector’s time, with the rate of detecting high-risk property conditions increasing from 13% of inspections to above 70% (OECD, 2018[26]).

Demonstrating return on investment

Because public sector organisations are accountable for the money they spend and there are always more demands for funding than there are available resources, sophisticated investment and procurement processes have been developed. The OECD Recommendation of the Council on Public Procurement (2015[67]) promotes a strategic approach throughout the different sectors and levels of the public sector with guiding principles addressing the entire procurement life cycle. Specifically within the context of digital government, Recommendation 9 of the OECD Recommendation of the Council on Digital Government Strategies (2014[68]) emphasises the importance of business cases to reinforce digital policy. The ongoing work of the OECD Working Party of Senior Digital Government Officials (E-Leaders) has included a focus on developing business cases and transforming ICT commissioning.

This third aspect of “evaluation and monitoring” looks at the way in which a DDPS approach can support arguments for funding and measurement of the impact of that spend. Public value in this case is then much more closely aligned with financial accountability over spend, improved procurement decisions and the ability to describe the outcomes that were made possible. In Denmark, several initiatives within the national digitisation strategy are informed and supported by an analysis of the value of data sharing in the transformation of internal activities within government.

Demonstrating return on investment is another activity that relies strongly on the clarity of intent for any DDPS activity. Measuring the impact of a given set of activities requires setting a baseline and developing a methodology to understand costs and benefits. These need to be considered when a public sector organisation is deciding to invest money in tackling a problem or designing a policy intervention during the “anticipation and planning” stage. They then need to have been implemented during the “delivery” phase. This ensures that there is always recognition of the value of an investment, as the example in Box 3.12 shows.

In the context of data-related spending, and particularly OGD, the focus has sometimes been on measuring the act of publication as a policy outcome, resulting in the objective to increase the number of datasets available for public access. This can mean efforts are made to releasing data rather than for its reuse. While release is a prerequisite for creating value in the use of data, any measurement for calculating the public value returned through investment on data-related efforts needs to consider defining benchmarks for both the sharing, curating and publishing stage and the use and reuse stage of the government data value cycle.

Accountability through audit trails and transparency

This final area of the “evaluation and monitoring” activity within the DDPS concerns accountability, and in particular the role of data as a tool for providing access to government records and encouraging a culture of transparency. The role of integrity actors in government relies heavily on the successful implementation of the DDPS. The case study on “Data-Driven Integrity” highlights the opportunities that present themselves in an associated but different approach to accountability, namely the prevention of fraud and corruption.

The importance of auditing activities and the desire for transparency in government should not be underestimated as the basis for making the argument about increasing the profile of data-related activity within government. In the context of OGD, that agenda has been a significant driver in seeing data published in line with the OECD Recommendation of the Council on Open Government (OECD, 2017[59]; Ubaldi, 2013[4]).

More broadly, public sector organisations should be interested in exploring how the impact of a policy can be demonstrated in ways that build accountability and stimulate trust with the public. This echoes some of the discussion earlier in the chapter around the analysis of “delivery” activities. Box 3.13 presents the example of Japan, whose whole-of-government approach to evidencing the impact of policy comes from the very top of the government (supporting the importance of leadership and vision discussed in Chapter 2). Within that context, the analysis and evaluation of a policy intervention is not only seen as promoting greater accountability, but provides the underpinning to ensure that a particular policy intervention is having the desired effect. Adopting robust methods like these to measure, and publish, the evidence about the reality of a policy or service not only during a pilot or alpha but throughout, can form the basis of arguing for future funding and, over time, demonstrate return on investment as well as helping to create a culture of accountability.

In Portugal, the Health Service Transparency Portal,3 and the justice transparency portal4 provide open access to performance data on a sectoral basis. Moreover, at the municipal level, data and dashboards are provided by the Municipal Transparency Portal,5 detailing the local use of public money and the performance of the services they provide. In Denmark, Statistics Denmark6 provides a common platform for reporting data relating to the performance of the country. These performance insights include the presentation of data about the effectiveness of policy. In the United Kingdom, all services are required to publish their performance data as part of the Service Standard, but this does not always identify the success, or otherwise, of a service in relation to its policy intent.