An interesting development is occurring in government agencies, both federal government, state and local government.
For the last twenty years, such agencies have moved from paper to digital documents. In the first phase of this movement, the key issues were the construction of data entry platforms and the storage of data for querying individual data records for administrative purposes. The second phase produced a preset group of management reports, carefully hand programmed by those who built the data entry platform. Concurrent with these developments the role of the chief information officer or CIO emerged. A typical CIO’s job was serving the user hardware needs as well as the care and feeding of software and related data storage and cybersecurity.
Built around these senior roles was an emergent set of skill requirements, often demanding knowledge from fields as diverse as computer science, engineering, project management and business process design. New degree programs in management information systems, IT management, and other areas were created to certify that aspirant CIO’s had the necessary knowledge to perform the role. Despite that, there are many current CIO’s who acquired necessary skills on-the-job, at lower levels in an IT department of an organization. This remains true many years after the CIO was common.
The use of digital data within organizations appears to go through phases. First comes the rather simple need of keeping track of basic counts of transactions or production. Second, management attempts to extract more information from the data – essentially more evaluative in nature. This is the phase in which “dashboards” are a common tool. Dashboards tend to be fixed presentations of a stock number of indicators with some way of evaluating whether things are going well or not (e.g., red, yellow, or green symbols). The third phase arises when curious users want answers to why some dashboard indicators are getting better and others are getting worse. The questions usually go beyond what is happening to why it is happening. Stock, pre-defined reports do not generally answer such questions because the questions rapidly change. Bespoke analyses are needed.
At the current time, there are two new roles that seem to be developing within organizations related to these phases. They are the Chief Data Officer and the Chief Evaluation Officer.
In some organizations, the role of chief data officer arose to address what could be viewed as a fourth phase of practical management use of data. This phase attempts to address questions that can be answered only by the merging or joint use of multiple data sets from various constituent units within an organization. Facilitating such projects requires coordination of the data resources. The needed skills involve understanding the various processes of different subunits, the nature of the data produced by the processes, and the risks and benefits of combining data from different sources. In some organizations, the chief data officer is the key point of contact for using data from organizations outside their own organization. Further, the chief data officer generally has data analytic literacy, to help guide appropriate development of analytically useful data sets.
The chief evaluation officer is yet another development. Increasingly, agencies that administer government programs are asked to provide evidence on the efficiency and effectiveness of those programs. Using the values of much scientific inquiry, the evaluation gains credibility if it uses data measuring key characteristics and outcomes, it is conducted by those independent of the designers and executors of the programs, and it provides conclusions are objective and data-based. Some organizations use part of the program budget to support such evaluations by outside contractors. The chief evaluation officer is a role to facilitate such external work or to provide it as an internal resource. The skills necessary for this role goes beyond knowledge of program processes and how to combine data, to experimental designs that would reinforce causal inference, to moving from statistical analysis to verbal conclusions about the performance.
While university degree programs and private sector certification programs map well into the CIO role, such is not yet true for chief data officers and chief evaluation officers. As these two new roles emerge, one would hope that formal educational programs preparing people for these senior roles will be developed.
I recommend the consideration of the four roles embedded in comprehensive policy analysis: (1) the descriptive analyst, (2) the predictive analyst, (3) the normative analyst, and (4) the prescriptive analyst.
Having these four roles helps to prevent the collection and storage of data for no particular purpose, and helps to make the purpose(s) rule rather than have the data rule. Form then follows function rather than convenient form circumscribes practical function!
Serving the purposes of comprehensive policy analysis, the organizational form would include (1) Chief Description Officer, (2) Chief Prediction Officer, (3) Chief Valuation Officer, and (4) Chief Prescription Officer.
Working together, they would make sure that (1) descriptive analysis is used to understand the base-line situation, (2) predictive analysis is used to envision projected eventualities ensuing from inherent trends and contemplated policy alternatives, (3) situations and eventualities revealed by descriptive analysis and predictive analysis are then judged on the basis of normative values of policy actors, affected populations and other stakeholders, and (4) economically effective and politically palatable policies are then prescribed to key leaders.
Do you know if that analytical and organizational approach is being undertaken by any public or private organizations? Are there any educational programs that are preparing people for those roles? American University’s School of International Service taught along those lines in a policy analysis course, once upon a time. Do any other programs come to mind? Any at GU?