‘Suspicious’ population hampers CSO data collection

STATISTICS don’t just generate themselves. They require old-fashioned techniques like pounding the pavement to collect data from the population, that’s then analysed and published for the public record by the Central Statistical Office. Unfortunately, though, the population isn’t always willing or accessible for CSO enumerators who need to collect that valuable information. And it’s the more affluent areas that are most problematic.

“We depend on the co-operation of the public to get the data. In certain areas, in particular areas where you have gated communities, there has been a high number of refusals and non-response,” the CSO’s Director of Statistics, Sean O’Brien told Newsday in a recent interview.

People have taken to photographing and videoing enumerators, posting their images on social media and warning friends and followers to “beware” of this “suspicious person.”

“People on social media are attacking enumerators. People are calling the police, so almost daily, enumerators have to explain to the police who they are. Every morning I receive calls asking me if these people are genuine. So, with the rise of gated communities which refuse to let the enumerators even enter, with the rise in suspicion of strangers at the gates, what we have had is huge hits on our response rates,” O’Brien said.

Enumerators each have a specific area – an enumerative district – where they collect sample data of the population as part of the Continuous Sample Survey of the Population (CSSP). This survey has been going on since the 1950s, O’Brien said, and it’s from this that labour force statistics are generated.

In TT, though, labour force statistics are notoriously late, because of the difficulty to get data. “If the non-response rate is too low, we have to replace that sample because the quality of the data is unreliable. We have to replace that district with another and then start over the beginning,” he said. This invariably holds back the entire process.

Labour force statistics, for example, are supposed to come out every quarter – at most, two quarters late. In TT, they are four quarters behind.

“People think that if the labour statistics are late it’s just that the CSO isn’t bringing it out. But the CSO can’t just invent the data. We have to collect it from the source and if the source isn’t co-operating you’re going to have these delays. We can’t give you data that’s not accurate. As you can imagine accuracy and timeliness war against each other in this business.

Because of a rising non-response rate it’s more difficult to establish accuracy and in a bid to establish this accuracy in the face of non-response rates, timeliness has often suffered,” O’Brien said.

Because of the design of the survey, he said, unless all data is collated and analysed, the CSO can’t publish. “There’s certain areas that we already have 2019 data—way ahead of the curve. But then there’s certain areas where the last data we have is first quarter 2018 and we still have to look for it. So, the area that slows us down the most is the fastest we can go. So the weakest link in the chain that’s as strong as the chain is,” O’Brien said.

For more on the Director of Statistics Sean O’Brien’s interview, including need for the National Statistical Institute, see the Business Day this Thursday.


"‘Suspicious’ population hampers CSO data collection"

More in this section