Wiping the glass on data transparency

By Flickr

THE MIDDLE GROUND | December 15, 2015

Lee Kuan Yew School of Public Policy academics Prof Asit Biswas and Mr Kris Hartley said in an ST commentary last Friday that, to quote the headline, “S’pore could afford to be more open to sharing data”, after analysing the 2014 Global Open Data Index collated by Open Knowledge, a global non-profit network aimed at increasing global data transparency.

Today, there is a riposte from the Ministry of Finance (MOF) in a letter headlined, “Govt more open in sharing data“. The ministry hit out at the academics’ use of outdated data. The 2015 index, for example, showed that Singapore was placed 23rd out of 122 countries in terms of data transparency, not 63rd as the earlier index had shown. And if the academics had checked with the ministry or looked at the Budget website, they would have found more details on G spending.

“It is regrettable that they had drawn the wrong conclusions without exercising the care and rigour one expects of academics from the Lee Kuan Yew School of Public Policy.”

Now, it is a very serious thing to cast these sort of aspersions on academics, which is why, as expected, the academics bristled: “The insinuation that academics like us used outdated data is not only unwarranted but is demonstrably unfair and erroneous.” They said the 2015 index appeared just last Wednesday, well after they had done their analysis and submitted their commentary. And while Singapore’s scores improved, with G spending scoring 10 per cent in transparency, “it is difficult to understand why most developing countries share such low scores”.

A storm in a teacup? Or are there deeper undercurrents?

From a reading of the ministry’s reply, it would appear that Mr Lim Yuin Chien was more intent on criticising the judgment of the two academics rather than addressing the state of data transparency in Singapore. It would have been helpful, for example, if the spokesman had explained what he described as “flaws on the measurement for Singapore in the old index” – rather than use the line “if you had asked us, we’d have told you so”, and then use concrete examples to show that Singapore does not deserve its low rankings in some aspects of data transparency.

We took a look at the latest 2015 index and it generally yielded similar conclusions, even as Singapore made some strides towards improving its overall data transparency, as seen by the significant increase in data transparency ranking from 63rd in 2014, to 23rd in 2015.

MOF pointed out that the academics could have referred to data regarding G budget and spending at the Singapore Budget 2015 website. The academics did not refer to this, and it can only be inferred that they were looking at data comparisons across countries rather than methodological errors. Prof Biswas and Mr Hartley also said that they planned to examine the methodological errors of the index in another academic paper.

So we took a look at the Budget website. The ‘Past Budget Archives’ section containing Budget reports from previous years is available on the site as well, but only go back to 2002. Detailed breakdowns of the Budget are only available for the years after 2008. Most of the data available on the Budget website are also in Portable Document Format (PDF), which means that the data cannot be processed by a machine or a computer (unlike data in an Excel spreadsheet, for instance) – one of the criteria for data openness set by the Index.

Leaving aside the Budget data, the MOF reply raises more questions about the state of data transparency in Singapore, rather than filling in the gaps. Perhaps, it is the nature of what and how the G shares its data with the public that we should be concerned with here, rather than argue about whether the G is being transparent with its data or not, in absolute terms.

On Nov 27, TMG wrote about the number of public consultations that the G conducted between GE2011 and GE2015. The Ministry of the Environment and Water Resources (MEWR) wrote to us to say that it had conducted 27 public consultations within that period. Our search on the Reaching everyone for active citizenry @ home (REACH) portal (the G’s feedback-gathering unit) for consultations conducted by MEWR yielded three entries. We thank MEWR for this information.

Some public consultations can be found by searching within the ministry’s websites, while other public consultations are reflected on the REACH website. Certainly more coordination is to be expected, especially since REACH touts itself as a facilitator of a “whole-of-government efforts to engage and connect with Singaporeans on national and social issues”. That there is no one-stop, or even up-to-date platform for the public to find out exactly how many and what kind of public consultations each ministry conducts suggests that the G should look at improving the way it releases data intended for public knowledge.

Data transparency is not merely about making the data available for public viewing, it includes removing barriers that prevent members of the public from accessing those data, including erecting paywalls and releasing certain data only in hardcopy format. Ensuring that datasets are consistent among ministries and agencies is also another important aspect when it comes to making governmental data more transparent.

The two academics said that data should be readily available to the public and academics for independent scrutiny and research. No one would disagree with this. Not many would disagree with this other statement: “In recent years, we have seen perceptible improvements in data availability.”

The G’s new portal, for example, is a treasure trove of information. The beta version of the new Open Data portal, which was released earlier this year in July, has “more than 12,000 data sets available”, featuring some datasets that would definitely interest the public. The types housed within the data portal ranges from social statistics on marriage rates by gender, to transport statistics on the number of driving licences in Singapore by class. The purpose of the new portal, according to the portal administrators, is to “show how data can be relevant to the layman”.

But while there are whole matrices and sub-classifications that would drive the nerdy nuts, there is still some data which is apparently still off (public) limits. We’re not referring to data concerning the armed forces or internal security but on immigration and the foreign workforce, for example. It is disconcerting that the G can sort migrant workers by work-pass classification, but does not do so by nationality (or just won’t say so). Back in 2008, The Online Citizen sent an enquiry to the Immigration and Checkpoints Authority (ICA) and the Ministry of Manpower (MOM) regarding statistics on Burmese workers in Singapore. Both ministry and agency clarified that the data existed, but was classified. We wonder why the G is so unwilling to divulge such statistics. Even when asked in Parliament back in 2014, MOM Minister Tan Chuan-Jin said that data on foreign workers categorised by nationality was unavailable.

At least, that’s better than the Media Development Authority (MDA) which wouldn’t tell us which books had been lifted from its ban list of objectionable material when we asked last month. TMG came up with its own list which others have said could be wrong. Except that it doesn’t look as if MDA is about to enlighten us either way. Not in the name of transparency.

This article was published by THE MIDDLE GROUND, December 15, 2015.