Before starting, have you read my previous article on public sector analytics? If you haven’t, please see here.
As promised, I wanted to tell you briefly about the obstacles of public sector regarding the expansion of the use of advanced analytics. Honestly speaking, this can be a topic for a white paper, but I will do my best to keep it short and easy-to-read.
Many exciting services are delivered in developed countries as the utilization of public sector analytics is enhanced. However, only a few countries could get benefits by starting certain iniatives. What about the rest?
Here are the short reasons, excluding macro-economic factors:
Public sector analytics can not be achieved overnight. Certain steps need to be followed in order to reach valuable insights and foresights. Those steps are:
- Ensuring data quality: Data accuracy, completion, cleanliness, timeliness etc.
- Building an analytical model: Approach, variables, inputs, outputs, success measures, value definitions, rules & conditions and other components.
- Deriving insight & foresight: Insights are derived from descriptive analytics, whereas foresights are derived from predictive analytics. An ideal model should also contain prescriptive analytics.
Here is a well structured graph prepared by HBS:
The problem is, in most of the developing countries, many of the public entities can not even pass to the blue stage, which means those entities can not really derive insights about understanding the reasons (the answer of “why is this happening?”). You may ask “the reasons of what?” Well, the ultimate reason of an analytical study is to derive an insight from several analysis, which leads to a decision making. And the essential problem of public entities is that most of the day-to-day actions are taken using daily, weekly or monthly reports (I doubt even the accuracy of the raw data they use). Unfortunately, these reports don’t result in permanent solutions for major problems. Bad news is, you can not reach to an ultimate solution for a public-wide problem by using “reports”, you need to go far beyond… (In my next article I will give you real life examples about what I meant here)
Another essential reason behind all these maturity issues is that most of the governments have not fully realized the value of their data. This causes the late reactions of such governments to the technological improvements, as it has been always the problem of late-comer countries through the entire history.
2) Human Resources:
There is no clear data scientist career path for public sector, yet. The required capabilities that are expected from a data scientist is data management, analytical modelling and business analysis. We can hardly say that it is easy to find such people who excel in these 3 fields together, and even harder to say if those people would prefer public sector instead of private sector. Besides, big data is still considered as a new trend and most of the times it requires multi-disciplinary background.
Below is a graph of the demand for such people in UK and the problem of “job match” is revealed clearly:
3) Political dynamics:
There is a thin line between transparency and information security. Identities of citizens should be protected, as well as the secret information of the government. At the same time, certain information is expected to be shared publicly, and this increasingly transforms from favour to expectation and from expectation to obligation.
In private sector, you can study on defining micro-segments, analyzing needs, behaviors and demographics and offering campaigns in order to generate revenue by cross-sell, up-sell, retention etc.. Whereas in public sector, segmenting citizens is highly controversial. It is considered as a risky and inappropriate action; because governments, as being the rule-makers, have the tendency of satisfying certain targeted groups of citizens according to the potential benefits they can get in order to protect their power.
The concept of “equality” and “fair service” can be abandoned if governments would have the most accurate information of the “value” and the “satisfaction” of their citizens. Ideally, a government (or a public entity, it does not really matter) should have a balance (very sensitively) when customizing public services and there are some countries for which it is meaningless to argue any balance at all due to their internal political challenges & dynamics.
This sentence includes the meaning that I believe the customization of citizens to a certain extent, since it does not necessarily worsen everything for everyone. Customization can lead to a “win-win” situation for masses. The simplest example would be tax collection systems (in an ideal policy-making system), or certain public services provided for disabled people in public transportation, according to their needs. My future articles will cover several examples about this section.
On the other hand, if a government conducts a detailed cost & benefit analysis of large number of individuals, this exercise would take government into a path of “extreme profitability centric” thinking.
Extreme profitability centric thinking of government can, for instance, prevent farmers living in mountains from having access to internet. It prioritizes cost-benefit analysis and disregards the essential social necessities that lies within state’s definition. In that case, we can all forget about “equal treatment” or “equality of opportunities“.
“The Big Data Opportunity” – Policy Exchange
“Analytics for Government – The smart way to cut costs, optimise performance and deliver reform in the public sector” – SAS
Image source: Flickr