No national policy can be comprehensive and durable unless backed by robust data derived from reliable sources and made available in z timely manner to policy managers. Pertinent to mention that that David Spiegelhalter, President of Royal Statistical Society in the UK, in his presidential address in 2017 observed that there is need for rectifying declining trust in numbers and figures. Stated differently, instead of relying on esoteric statistical techniques, there is need for improving trust in quality of data. He recommended to the statistical community that the best way of inspiring trust was to be trustworthy by demonstrating competence, reliability and honesty.
In the Pakistani context, a national policy on official statistics, besides acquisition of reliable data, should lay the groundwork for ethical data collection, highlighting importance of data quality, addressing the need for documentation and durable data storage. Sample surveys, the bedrock of Pakistani statistical systems, must make explicit choices on who, what, and how to ask various questions.
Unsurprisingly, in a statistical system developed by renowned statisticians and econometricians, much attention has been directed towards identifying the universe of respondents and sample selection, which is only a part of the challenge. Importantly, it should involve scholars from various disciplines – given the increasing need for statistics in diverse areas. If for example, sociologists and psychologists interested in occupations were involved in overseeing this change, the results would be more refined and credible.
Designing of surveys and developing questions into a science needs to transcend the skill set usually employed by statistical systems. In a country, surveys repeatedly document extremely low mathematical skills, and how reliable is the data when individuals are asked to compare their expectations of inflation rates over the coming year with that in the future? So, there is little understanding of reliability and validity of the data and yet they form the bedrock of policies.
Here, experiments designed by cognitive anthropologists, educational assessment experts and survey design specialists are needed to frame the right questions to arrive at right answers. And even then, estimating uncertainty surrounding these results has to be adequately catered for with healthy scepticism.
Honest and integrity in data collection and analysis are thus important steps as many reports pay greater attention to data collection is increasingly being done by contractual employees and for-profit organizations. So, supervising them and ensuring their honesty remains an issue. While improved technology for monitoring fieldwork such as random segments, audio recording of interviews and real-time checks for detecting frauds and errors may help in honesty, there is no substitute for empathy and experience.
One expects interviewers to work under challenging circumstances and they are often sent out to collect data with little training, preparation and logistical support. Hence a smart survey management structure that understands the difficulties of on-the-ground data collectors and responds appropriately to find ensuring quality and honesty is essential the cornerstone of good data collection.
The official statistics engage with these challenges only tangentially by focusing largely on shoddy coordination with different ministries at the Centre and between Centre and provinces. Moreover, a tendency to centralize authority and decision-making within well-defined structures forms the core of the policy statement. Therefore, a registered society under an oversight set up with endowment fund needs to be tasked with all government data collection and statistical analyses.
By creating a statistical data ecosystem that harnesses the energy of diverse institutions and disciplines in which innovative thinking on data on and analysis is necessary. However, collaboration between academics and official agencies has weakened substantially in recent years.
In order to remedy revitalizing the statistical infrastructure efforts to be made to harness diverse energies from academic and research institutions- government, semi-government, NGOs, think tanks, research centers and technology-savvy private sector organizations to garner technology-driven data collection.
As an example, China, South Africa, Brazil, the UK and the US, the statistical ecosystems rely on universities, research institutions and government agencies working synergistically. Creative thinking about building synergies with diverse communities such as academic and research institutions would strengthen it and reduce the burden on the government statistical bureau leaving it free to devote greater attention to developing quality control parameters and to play an oversight and coordination role.
In post-statistics society, the phrase ‘figures don’t lie, but liars figure’ seems to sum up the motif. A report in The Guardian in 2017 noted that declining trust in official statistics around the world has become the ‘new normal’ and argues that it damages democracy by jeopardizing quality of public knowledge and public discourse.
Hopefully with the new government in office after in July 2018 general elections, a draft national policy on official statistics could offer a fresh start by fostering trust in statistics and enhancing inclusiveness through competence, reliability and honesty in public statistics. Needless to reiterate, trust in official statistics is vital for democracy and any new policy must avoid centralization and political interference by fudging or obfuscating genuine data.Dr. Maqsudul Hasan, "Data services," Business Recorder. 2018-07-05.
Keywords: Political science , Reliable sources , Random segments , Management structure , Official statistics , Statistical bureau , Coordination role , Statistical data , UK