Even without the proof of research, most people would agree that when an individual stops working, whether they’ve been made redundant or fired, their habits and behaviour start to alter. This could mean their daily routine changes, i.e. when they wake up, when they leave the house, when they go to bed or even when they use their mobile phone.
Recent research has shown that such shifts in people’s behaviour can be recognised in the study of mobile phone and/or social media usage, in what’s known more broadly as big data analysis; the study of which has been widely discussed in the media.
More recently there has been a particular focus on how this data can provide insight into the levels and effects of unemployment in the UK.
The use of mobile and social media data
Two papers, discussing exactly this topic have recently been published; the first in the Journal of the Royal Society Interface and the second in PLOS One. Both papers detail how important the data from people’s mobile and social media use are because this data is immediate and this results in speedier analysis, tracking and action than traditional data collections methods allow for.
Jameson Toole, one of the authors of the Royal Society Interface paper explains the older, traditional method governments currently use, stating that statistics are collected based on “a paradigm of data collection and analysis begun in the 1930s. Most economic statistics are constructed from either survey data or administrative records.”
The issue that many have with these traditional methods is that they are often very time-consuming and expensive to carry out. Data is collected from a number of different sources, and published at a number of different times throughout the year, which means the published results can often be rather out of date. It also takes a long time to collate the data from all the different resources it is collected through.
Advocates of big data focus on how much quicker results can be seen and analysed when used instead of more traditional methods, in this particular case, to analyse the effects of unemployment. Both of the published papers mentioned above explore the benefits of big data in understanding UK unemployment, when this data has been collected from mobile phone and social media.
Social media analysis can give us real-time insight into people’s locations and therefore provide a detailed understanding into how much people move about, as well as the times of day they’re most active. The report analysts even studied how grammatically correct people’s social media posts were, which fed into their research papers. This research showed there was a direct relationship between areas of high employment and social media posts where grammar is correct, and where people were more mobile and active earlier in the day.
Researchers believe that big data can and should be used over and above the more traditional, and lengthy research methods, not only because big data can produce accurate estimates of unemployment in particular areas, but because it can do so at a speedier rate. Speedier analysis and prediction would obviously be of great benefit to the government and could, potentially help when politicians have to make important economic or employment policy decisions.
Concerns about privacy
Individuals often express a lot of concern at the amount of data that is accessed by large organisations and the government. Most, quite rightly, assume that their private data is exactly that, private, and may therefore be concerned to realise companies and governments are able to make such strong deductions about people’s states of mind, employment and general behaviour simply by accessing their social media and phone records.
Unfortunately, there is a lot of potential for this kind of information to be mistreated and obviously the standard argument by the public is that they don’t want companies and the government accessing what is deemed as their private information. Especially when it involves their changing behaviour following losing their job, or after a long spate of unemployment, i.e. if data from calls to a suicide hotline were studied in a known area of unemployment.
This is an issue that David Lazer, Northeastern University computational social scientist and report author recognises, however, he emphasises that such research processes have to follow certain rules and regulations in order to pass Institutional Review Board (IRB) ethical standards. This means that things like anonymity have to be taken into account during data collection by a third party, where customers have not given their express permission for their data to be used.
To further protect the privacy of those studied, Lazer’s researchers had to sign non-disclosure agreements and the names of the mobile carriers, which provided the mobile data, were not disclosed. He believe the opportunities provided by big data outweigh the privacy issue, stating: “There are considerable opportunities for intrusion by corporate actors and government, but there are also powerful uses of these data for public good, in ways that should not intrude on privacy if done properly.”
The future of big data and government statistics
Even though there are many advantages to using big data, it seems we are not quite ready to switch over to it completely and leave the traditional data collection methods behind.
Both methods have their own strengths and weaknesses, for instance social media data is limited in its usefulness because of the age groups using it, i.e. the data collected would probably not reflect the 60+ age range. However, it produces speedier results, whilst traditional methods are a lot slower but extremely comprehensive, accurate and include all age ranges.
It seems, for now, the best route forward to ensure data collection and analysis on unemployment rates are quick, as well as comprehensive, may well be to combine the two methods, creating a data collection hybrid.