Annex C. Data-driven human resources management: Enabling the strategic use of human resources data for a high-performing civil service

The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of international law.

At the age of the digital transformation, governments have understood the growing importance of the value of data as a foundation for improved policy making, service delivery and ongoing performance management. In this context, many OECD countries are aiming to develop a data-driven public sector (DDPS), one which recognises data as an asset, integral to policy making, service delivery, organisational management and innovation. The strategic approach for DDPS can have a positive impact on the results governments deliver through evidence-based policy making and data-informed service design (van Ooijen, Ubaldi and Welby, 2019).

In the same manner, data-driven human resources management (DDHRM) pursues strategic human resources (HR) management by using HR data. In the past, HR policy had a tendency to rely on past practices or a decision maker’s experience or intuition, with no scientific or objective evidence. Today, workforce data from multiple sources present opportunities to manage public employees through evidence-based HR policies. Governments are thus increasingly able to recruit, deploy, train, motivate and retain their employees in a scientific and analytic way based on objective HR data.

The 2019 OECD Recommendation of the Council on Public Service Leadership and Capability presents 14 principles of a fit-for-purpose public service. It includes a recommendation to develop “a long-term, strategic and systematic approach to people management … using HR and workforce data for strategic and predictive analytics, while taking all necessary steps to ensure data privacy” (OECD, 2019).

Public services are collecting more data on their public employees today than ever before. Demographic data provide a snapshot of the workforce and enable a better understanding of skill sets, workforce diversity and age. Administrative data show employment trends and patterns that can indicate organisational health through, for example, job attractiveness, the efficiency of HR processes and mobility/turnover rates. Data from employee surveys can provide rich indications of employees’ engagement and satisfaction with their work and working environment.

HR data are abundant. Today, in the era of “big data”, the amount of data available to inform strategic workforce management has exploded and thanks to the development of information technology, it can be processed and utilised more efficiently. These data can be collected from both internal (e.g. human resources information systems or employee surveys) and external (e.g. social media or labour market trends) sources. However, most countries only collect HR data, as they struggle with scientifically analysing, insightfully interpreting and proactively using them for better management decision making and HRM policy development and delivery. They are still not sure how to make sense of all these data or what to do with them; there are a lot of challenges in making DDHRM work well. Data scientist is not yet a common job profile within HR departments.

This case study will focus on how a DDHRM can be applied to strategic human resources management in order to attain organisational goals effectively and subsequently identify challenges governments may face in establishing a DDHRM.

DDHRM is also referred to as evidence-based HRM, HR analytics and workforce analytics in the literature. Evidence-based HRM is a decision-making process combining critical thinking with use of the best available scientific evidence and business information. It is composed of four elements: 1) the best available research evidence; 2) organisational facts; 3) metrics and assessments; 4) practitioner reflection and judgement and the consideration of the affected stakeholders (Rousseau and Barends, 2011). HR analytics is an HR practice enabled by information technology that uses descriptive, visual and statistical analyses of data related to HR processes, human capital, organisational performance and external economic benchmarks to establish business impact and enable data-driven decision making (Marler and Boudreau, 2017). HR analytics is the systematic identification and quantification of the people drivers of business outcomes (Heuvel and Bondarouk, 2016). This definition encapsulates HR, people, talent and human capital analytics that focus on the individual. Workforce analytics is people analytics on a larger scale. It is about scaling up the data from multiple individuals to assess trends on the broader workforce level. Sometimes it is more narrowly used to discuss workforce planning.

In this case study, DDHRM is defined as a strategic process aiming for better HR decisions and policies throughout the government by collecting, measuring and using HR data such as demographic data, administrative data (including pay data and turnover), employee perception data (employee surveys) and performance data. DDHRM is based on data and evidence instead of intuition or personal experience.

The OECD has collected and used quantitative and qualitative HR data for comparative analysis across OECD countries in the field of public sector human resources management and civil service reform strategies. The 2016 Survey on Strategic Human Resources Management in Central Governments of OECD Countries gathered data related to the broad trends of public employment and HRM across OECD countries and provided OECD countries with a better picture of where they stand compared to other countries in these fields. This included a new set of questions on the collection and use of data for HRM. The survey looked at three types of data: administrative data, employee survey data and employee performance data. In Government at a Glance 2017, the OECD presented the results of a survey which looks at the amount and type of administrative HR data collected by OECD countries, which shows a wide variation (OECD, 2017).

The results of the survey show that most OECD countries collect and centralise basic HR data, such as number of employees, age and gender. However, relatively fewer countries gather deeper and more meaningful HR data related to working conditions or organisational culture, such as minority status, flexible working arrangements and union membership. Data related to training, leave and mobility are often not aggregated centrally, when they are collected by ministries.

Employee surveys are another important source of HR data, and most OECD countries use them, to differing degrees. Centralised civil service-wide surveys are often run at regular intervals, and complemented by specific surveys carried out by individual agencies and ministries. OECD countries measure employee perceptions of job satisfaction, employee motivation and work/life balance through employee surveys. On the other hand, relatively fewer countries use these tools to measure inclusion, harassment and effectiveness of HRM systems.

When it comes to employee performance data, less than half of the OECD countries report collecting these data centrally. This may be due to the difficulty of objectively measuring employee performance in ways that are comparable across diverse job types and working conditions.

In most OECD countries, administrative data are used for reporting to the public and employee survey data are used for reporting to the senior civil service. Employee performance data are mainly used for assessing performance or informing organisational training plans.

The OECD framework for understanding the opportunities of a DDPS (Figure C.6) identifies three areas in which data-driven initiatives are being developed to support the decision making process across policy areas and levels of government: 1) anticipatory governance; 2) design and delivery; and 3) performance management (van Ooijen, Ubaldi and Welby, 2019).

In this case study, the framework of DDHRM will adjust the OECD framework of DDPS to suit HR functions. DDHRM creates the opportunities in two main areas: 1) forecasting and planning; and 2) monitoring and evaluation.

Based on past and present data from various sources, predictive analytics involves the development of statistical models and forecasts to help identify future workforce and talent pool trends. Anticipating such trends gives managers and organisations a critical head start in preventing, mitigating or encouraging developments, ultimately saving them costs and helping improve performance.

The possibilities for predictive HR analytics are still being explored; however, a few common applications have thus far included strategic workforce planning, improving diversity and inclusion, and retaining top talent.

Strategic HRM aligns people management with the strategic goals of public sector organisations (OECD, 2011). Strategic workforce planning for strategic HRM is crucial to predict workforce change according to change in the administrative environment – such as demographic, technology and economic situations – and prepare for recruiting needed talents. Strategic workforce planning is a core HRM process that helps to identify, develop and sustain the necessary workforce skills. In doing so, it also contributes to the career and lifestyle goals of employees and ensures the continued effective performance of organisations. Workforce planning is a dynamic process that ensures that the organisation has the right number of people with the right skills in the right place at the right time to deliver short and long-term organisational objectives. Workforce planning aims to reach an optimal combination of available personnel budget and appropriate number of human resources endowed with the required skills to bring about organisational objectives. Workforce planning not only identifies mission-critical occupations and the essential competencies to meet organisational goals, but also detects competency gaps (Huerta Melchor, 2013).

DDHRM can help to forecast a list of potential gaps in the workforce by looking at long term trends. For example, various data can be analysed and predicted when analysing the current state and supply of the workforce and forecasting the trend and needs of the future workforce. This can include, for example, workforce movement (workforce inflows such as new hires, promotions and transfers; workforce outflows such as resignations, retirements and involuntary terminations), the differences between the present available workforce level and the workforce level which will be required in the future, and gaps in competencies and numbers of employees in each job area. This information can be used to design a strategic workforce plan to close workforce gaps in each job area and strengthen organisational competitiveness in the future.

One example is the Mexican Ministry of Energy, which is using workforce planning to identify current and future skills gaps in oil and gas occupations over a ten-year horizon. “The model leverages a number of adjustable macroeconomic variables such as oil price and exchange rates that correlate strongly to the demand and supply of skilled labour. Based on an understanding of these gaps in critical skills, the ministry is able to work proactively with multiple stakeholders to address them. Building off from this initiative, the ministry has expanded the use of workforce planning and analytics to cover other sectors it is responsible for, such as renewable energy and sustainability” (Deloitte, 2016).

While strategic workforce planning may include employee turnover as one input to its modelling, predictive analytics can delve deeper, specifically into voluntary turnover with a view to reducing this type of attrition in organisations, and particularly among top performers. Indeed, employee churn costs employers greatly in terms of lost productivity and institutional knowledge, but also in sunk costs in recruitment and learning and development. Turnover also affects citizens in terms of interrupted policies and in the quality of service delivery. The United Kingdom’s Institute for Government published a report estimating that excessive turnover in departments costs the civil service between GBP 36 million and GBP 74 million each year in recruitment, training and lost productivity (‘Moving On”, 2019). As labour markets become increasingly competitive, predictive analytics is being used to help pre-empt and advert involuntary turnover, and particularly in certain groups of employees.

Such predications, however, must be based on robust models of the drivers of voluntary turnover. Several studies have tried to identify the specific causes of resignations in public services. Based on certain “signals”, this application of predictive analytics relies on models using past/present data to identify employees at “high risk” of attrition.

Several studies have attempted to tease out valid predictors of voluntary turnover. The results have found a multitude of drivers. Indeed, a literature review of existing attempts found that the strongest predictors for voluntary turnover were age, tenure, pay, overall job satisfaction, and employee’s perceptions of fairness. Other similar research findings suggested that personal or demographic variables – specifically age, gender, ethnicity, education and marital status – were important factors in the prediction of voluntary employee turnover. Other characteristics that studies focused on are salary, working conditions, job satisfaction, supervision, advancement, recognition, growth potential and burnout (Punnoose and Pankaj, 2016).

While most studies have focused on private sector employees, public services are beginning to conduct similar studies. In the United States, researchers used a database which included information on federal civil servants from the Office of Personnel and Management, including such dimensions as age, agency type, gender, salary level, geographical location of employee, length of service, occupation type, pay plan and work plans (i.e. temporary, full-time, etc.). The results of logistic regressions revealed a significant reduction in the probability of an employee quitting as his/her service length increases; odds increasing or decreasing depending on employee age; and odds of quitting are lower if the employee is in the standard pay plan. Comparing age and length of service, we found resignations spike around 6.25 years of service, regardless of age (Frye et al., 2018).

Using such models, HRM professionals and managers can then intervene by offering salary increases, professional opportunities or changing working conditions (i.e. flexible work, telework) suited, depending on the level of granularity of the data, to teams’/individuals’ own preferences. However, many note the ethical concerns of predictive analytics for employee attrition arguing that the data can suggest dangerous and unfounded correlations that may lead managers to draw incorrect conclusions. For instance, a relationship between gender and attrition may lead unethical managers to discriminate against certain potential recruits. Additionally, some argue inclusion of certain data will confound results. For instance, self-reported data may not always be accurate, especially if employees believe their responses are being used for predictive modelling.

DDHRM can also help to meet specific future targets to develop the workforce. For example, the Public Service Commission of the state of New South Wales in Australia has adopted a data-driven approach to designing and monitoring progress on diversity and inclusion policies (OECD, 2019).

In order to monitor agencies’ expected trajectories in meeting diversity targets, the Public Service Commission has developed a model that predicted – based on current recruitment and separation behaviour across the public sector – what the proportion of women in senior leadership roles would be. This was then extended to each cluster and became the starting point to demonstrate that unless a framework of high-impact whole-of-government initiatives were in place, there would be little movement of the rate. Thanks to this predictive model, the Public Service Commission arrived at the view that to achieve 50% of women in senior roles by 2025, the public sector needed six out of every ten appointments to senior roles to shift from four out of ten. Current data have shown this rate is now at 5.5 out of 10.

These cases illustrate the potential for forecasting and planning for better HR policies through DDHRM. By focusing on a set of future goals, multiple data sources can be combined to develop insights on current challenges that may be impeding the achievement of these goals. In all of the cases listed above, data-driven scenarios can help to see various versions of the future. Another useful point is that workforce data are easily accessible since data sources are internal. Most HR offices have access to key data points around the composition of their workforce, mobility patterns and bottlenecks. The challenge is in making these data useful and investing in the skill sets needed to analyse them and drive towards insights and solutions.

Following the DDPS framework, this area shows how HR data can be used to better understand the current state of the workforce and HR service delivery. HR data can be used to address problems by reflecting the needs of various stakeholders, such as HR staff, employees and other interested parties, and improving the effectiveness and efficiency of HR policies by providing feedback to the HR decision-making process through evaluation of the impact of HR policies. When data are collected on an ongoing basis and structured effectively, DDHRM can significantly reduce the amount of time between implementation and evaluation – identifying problems as they arise and enabling policy interventions, almost in real time. These two areas are dealt with together in this case study.

Given that human resources management is a strategic lever to achieve government objectives, most OECD countries try to innovate their approaches to HRM. The examples that follow show how HR data collected through monitoring and evaluation can be used to spark innovation in the design and delivery of people management policies and processes.

HRStat is a data-driven review process intended to improve human capital outcomes, enhance the performance capacity of agencies in achieving their strategic goals and objectives, and create a supportive culture for the use of data-driven reviews that inform agencies’ human capital decision making. The Office of Personnel Management introduced HRStat to the federal human capital community in 2013. (US Office of Personnel Management, 2017)

HRStat reviews focus on specific HR challenges, identified and explored through data analysis, monitoring and evaluation. As such, HRStat reviews do not merely present HR data on topics such as attrition rates, completion of performance evaluation plans, numbers of completed hiring decisions or training participation rates. Rather, federal agencies engage in data-driven reviews of HR areas that are in need of improvement, innovation or improved cost effectiveness.

For example, agencies may use the HRStat reviews to assess work demands, emerging mission imperatives and workforce trends likely to affect skills needs. They can also use them to evaluate HR strategies and interventions designed to reduce or eliminate competency gaps in vital positions, or to understand why certain interventions may help alleviate attrition risk among employees in high-impact positions. In this way, HRStat helps to create empirical evidence to inform HR decision making and provide agencies with a continuous means of learning and gaining insights for improving HR processes. Conducting HRStat reviews also enables agencies to evaluate progress, refine strategies and develop demonstrable quantifiable evidence of successful human capital outcomes.

Another aspect of the HRstat programme is the Maturity Model (see Figure C.7), which provides a diagnostic framework to assess the maturity level of an agency’s DDHRM. The Maturity Model serves as a practical and aspirational roadmap that will help agencies identify areas for improvement and enable them to monitor their progress over time.

The HRStat Maturity Model is conceptualised in terms of three components:

  1. 1. Scope of impact measures the degree to which metrics are integrated into the measurement of agency mission accomplishment.

  2. 2. Initiative and effort measures the degree to which an agency has developed the capacity to use HR data to inform decision making across the agency.

  3. 3. Performance of HRStat measures the degree to which an agency’s metrics are advancing to achieve targeted improvements and are validated against external benchmarks.

For each of these three components, there are four maturity levels (reactive, emerging, advanced, optimised). In describing the four maturity levels, the HRStat Maturity Model designates five domains of consideration: analytics, technology, talent/staff, collaboration, and leadership.

Korea provides another example of DDHRM for monitoring and evaluation. High demand for integrity and public confidence in the government in Korea requires transparent and accountable personnel management to better respond to public demand. The establishment of the Ministry of Personnel Management (MPM) in charge of HR innovation in 2014 has increased the demand for effective and responsive personnel management for the public. Starting in 2015, HR Innovation Diagnosis Indicators were developed and have been used to carry out objective assessments (Korea Ministry of Personnel Management, 2018).

Based on the indicators, the MPM assesses each government organisation’s HR innovations and provides feedback to enhance its innovation capability through a cycle of plan, do, see and feedback.

The indicators are composed of five fields:

  1. 1. implementation capacity measures agencies’ commitment to HR innovations and excellence in HR innovation plans

  2. 2. employment measures open and diverse recruitment

  3. 3. human resource development measures employees’ perceptions and awareness of development opportunities and organisational efforts to develop their workforce

  4. 4. expertise and performance management measures compliance and observance of various programmes to ensure professional standards, including performance management

  5. 5. improvement of working environment and conditions measures efforts to encourage the use of personal days off and a flexible work system, and attempts to fight discrimination among government workers.

In order to measure these indicators, the MPM uses a range of methods. Quantitative methods include indicators such as open positions and employment rates, and increase in employment of female managers. These are complemented by qualitative methods, for example excellence in HR innovation plans and appropriateness of education and training plans; as well as awareness and satisfaction survey methods, such as awareness of flexible working options and satisfaction with development opportunities.

As a part of collaborative innovation, the MPM sets indicators with participating government bodies and external experts after in-depth consideration. Moreover, indicators are adjusted on an annual basis, subject to the MPM’s annual innovation directions, feedback from participating bodies and changes in environments.

The MPM provides feedback and incentives (e.g. a long-term overseas training) to participating agencies and offers lagging agencies tailored consulting upon request. Furthermore, the MPM hosts quarterly workshops to spread good practices and set benchmarks.

  • Public administrations face several barriers in implementing DDHRM:

  • technical barriers (i.e. related to IT infrastructure and resources)

  • legal barriers (i.e. privacy concerns)

  • human resources barriers (i.e. lack of skills and knowledge among HRM professionals and senior managers).

These challenges are often why HRM practitioners in public sector organisations are perceived as falling behind their private sector counterparts in embracing DDHRM more whole-heartedly. This section discusses these issues in further detail, with a view to help practitioners anticipate potential obstacles and ensure they have the right foundations in place to successfully adopt DDHRM practices going forward.

The power of DDHRM originates from the compilation and analysis of data from across entities and organisations. As described earlier, there are multiple types of valuable data (i.e. pay, tenure, perception data, information on employees’ work experiences, education and performance, and HRM metrics like churn, sick leave, etc.) that commonly feed into DDHRM platforms and analysis. The challenge is integrating the data from several individual organisations/databases when each uses different formats or indicators. Ensuring data accuracy and comparability in such cases is difficult unless some prior method of quality control and standardisation is in place. This challenge is compounded even further as data from external (non-governmental) sources, such as from social media, becomes increasingly incorporated into DDHRM exercises.

Moreover, the shift towards DDHRM requires that organisations change the ways in which they collect and store data. This entails not only changes to the IT systems themselves – including the adoption of cloud computing, DDHRM platforms and software, etc. – but also to underlying business processes. For example, maintaining payroll, timekeeping or performance data will need to adapt to the new IT systems and methods of data collection, entry and storage. Such reforms entail not only changes to processes, but also require financial resources to develop and install DDHRM tools and train staff to transition to new systems. However, fairly recent austerity measures in the public sector have limited IT spending, and many managers still remain unconvinced of the business case in favour of such investments.

In response to these challenges, most organisations have adopted a piecemeal approach to compiling relevant data, starting with what is available and slowly building more comprehensive databases in partnership with other organisations (finance, payroll, human resources, etc.). Building consensus, excellent communication and support (in the form of written guidance, personnel or IT resources) is often necessary in bringing other organisations on board. Furthermore, monetising the advantages of adopting DDHRM tools and techniques, in terms of improved performance and organisational outcomes, as well as sharing good practices and experiences from early adopters in the public administration, have proven to help improve buy-in and participation from managers.

A second major barrier to adopting DDHRM in the public sector are legal constraints around the types of information that government organisations can collect, store and analyse. Indeed, information collected and maintained on employees is sensitive – from their pay to performance and health or other personal information. In many OECD countries, there are strict regulations that protect employee privacy. The EU’s General Data Protection Regulation, for example, defines high-risk data as those which are “likely to result in a high risk for the rights and freedoms of individuals,” and that, therefore, require greater protection. Organisations that fall victim to data breaches face high penalties and fines for breaking this law. Additionally, anti-discrimination legislation in many EU countries also limits the types of information organisations can even collect on employees. For example, in many European countries, it is illegal to keep data about ethnic minorities, and people with disabilities may not want to be counted as such. Furthermore, with the capacities of big data for triangulating and reconfiguring data, there are even doubts about whether individuals’ information can remain anonymous in the first place.

The OECD Privacy Framework (OECD, 2013) recommends several principles for the handling of personal data: collection limitation principle, data quality principle, purpose specification principle, use limitation principle, security safeguards principle, openness principle, individual participation principle, accountability principle. Moreover, the European Union’s General Data Protection Regulation came into force in May 2018 and replaced the Data Protection Directive 95/46/EC not only as a means of harmonising data privacy laws across Europe, but of providing a new baseline for protecting and empowering EU citizens in accessing their own data (van Ooijen, Ubaldi and Welby, 2019).

More so than the financial risks, however, employers are additionally concerned about potential losses of employee trust should a data breach occur, or if employees perceive that their privacy has been violated (such as, for example, if the data from their employee survey are being used without their consent, or for purposes they did not agree with). Employers, including public sector organisations, may lose credibility and face difficulties in recruiting top talent, potential retention issues, as well as lower levels of employee satisfaction and engagement.

Until recently, the HR field did not emphasise quantitative skills, and the majority of HR practitioners did not receive training in HR analytics. Pertinent skills are not only IT-related (i.e. sifting through data, developing and maintaining dashboards, etc.), or statistical (i.e. running regressions), but most importantly around “story-telling”. That is, HR analysts should be able to ask the right questions and use data in ways that directly respond to business problems and improved performance. This includes also the ability to develop visually impactful representations of data.

The approach adopted by many organisations until now has been two-fold: intensify training in HR analytics, and recruit data scientists to create cross-functional teams/groups around HR analytics. For example, people management decisions at Google are guided by the powerful “people analytics team”, which is made up of social scientists who conduct experimental, survey and archival research to inform people-related business decisions.

In June 2019, the US Office of Personnel Management issued a memorandum recognising the job title for data scientists. In a similar vein, Global Affairs Canada has developed a data analytics training pilot programme as part of its overall data strategy to increase data capacity among employees to make greater use of data in evidence-informed policy making. The UK Civil Service has also initiated a Digital, and Technology Fast Stream to attract and develop personnel with digital skills – including digital scientists – into the public service.

References

[8] Deloitte (2016), People Analytics in HR, Deloitte.

[6] Frye, A. et al. (2018), “Employee attrition: What makes an employee quit?”, SMU Data Science Review, Vol. 1/1, https://scholar.smu.edu/cgi/viewcontent.cgi?article=1010&context=datasciencereview.

[11] Heuvel, S. and T. Bondarouk (2016), “The rise (and fall?) of HR analytics: The future application, value, structure, and system support”, Academy of Management Proceedings, Vol. 2016/1, http://dx.doi.org/10.1108/JOEPP-03-2017-0022.

[16] Huerta Melchor, O. (2013), “The Government Workforce of the Future: Innovation in Strategic Workforce Planning in OECD Countries”, OECD Working Papers on Public Governance, No. 21, OECD Publishing, Paris, https://dx.doi.org/10.1787/5k487727gwvb-en.

[1] Korea Ministry of Personnel Management (2018), Measurement of Innovation with Personnel Management Innovation Diagnosis Indicators.

[4] Marler, J. and J. Boudreau (2017), “An evidence-based review of HR analytics”, The International Journal of Human Resource Management, Vol. 28/1, pp. 3-26, http://dx.doi.org/10.1080/09585192.2016.1244699.

[9] OECD (2019), Inclusive Leadership in the Public Service of New South Wales, Australia, OECD, Paris.

[2] OECD (2019), OECD Recommendation of the Council on Public Service Leadership and Capability, OECD, Paris, https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0445.

[12] OECD (2017), Government at a Glance 2017, OECD Publishing, Paris, https://dx.doi.org/10.1787/gov_glance-2017-en.

[14] OECD (2016), Engaging Public Employees for a High-Performing Civil Service, OECD Public Governance Reviews, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264267190-en.

[7] OECD (2013), The OECD Privacy Guidelines, OECD, Paris, http://www.oecd.org/sti/ieconomy/oecd_privacy_framework.pdf.

[15] OECD (2011), “Strategic human resources management”, in Government at a Glance 2011, OECD Publishing, Paris, https://dx.doi.org/10.1787/gov_glance-2011-22-en.

[5] Punnoose, R. and A. Pankaj (2016), “Prediction of employee turnover in organizations using machine earning algorithms”, International Journal of Advanced Research in Artificial Intelligence, Vol. 5/9, pp. 22-26, https://pdfs.semanticscholar.org/fa49/19810eaee67e851ad13775b78c94217a7908.pdf.

[3] Rousseau, D. and E. Barends (2011), “Becoming an evidence-based HR practitioner”, Human Resource Management Journal, Vol. 21/3, pp. 221-235, http://dx.doi.org/10.1111/j.1748-8583.2011.00173.x.

[10] US Office of Personnel Management (2017), HRStat Guidance, US Office of Personnel Management, Washington, DC, https://www.dcpas.osd.mil/content/documents/PA/4_OPM%20HRStat%20Guidance%20June%20%202017.pdf.

[13] van Ooijen, C., B. Ubaldi and B. Welby (2019), “A data-driven public sector: Enabling the strategic use of data for productive, inclusive and trustworthy governance”, OECD Working Papers on Public Governance, No. 33, OECD Publishing, Paris, https://dx.doi.org/10.1787/09ab162c-en.

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

https://doi.org/10.1787/059814a7-en

© OECD 2019

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at http://www.oecd.org/termsandconditions.