THE BLOG

Featuring fresh takes and real-time analysis from HuffPost's signature lineup of contributors

Murtaza Haider Headshot

In a Data-Centric World, Harper's Cuts to StatsCan Are Baffling

Posted: Updated:
Print
studiocasper via Getty Images
studiocasper via Getty Images

The modern economies are all about competing on data and analytics. While smart governments and businesses are investing in collecting data and raising armies of data scientists, Statistics Canada is starving under the Harper government.

Smart planning needs robust data and sound analytics. This is the message coming from an assembly of data archivists from across the world who have gathered in Toronto for their annual conference. The Canadian government should learn from the global experts who are highlighting the advances their governments and businesses have made in data and analytics. Starving nation's statistical agency is the wrong policy in a data-centric world.

Data librarians, archivists and statisticians from across the world gathered at Ryerson University for the 40th iAssist conference. This year's theme is on aligning data and research infrastructure. In a world awash with data, cheap and ubiquitous computing, and advanced algorithms, analytics have become critical for the success of businesses and governments.

The emergence of evidence-based planning has put evidence, i.e., data, in the centre. From health sciences to economic development, policymakers and strategists rely on data to carve out their strategies. Smart firms are investing heavily in hiring dozens of skilled statisticians and data scientists. Smart governments are investing in collecting relevant data and liberating it for use by researchers. Public sector agencies must consider investing in the infrastructure that promotes data collection, free data dissemination, and analysis. Canada needs to wake up and arrest the steady decline in the nation's ability to collect data and analyze it.

iAssist is an international outfit of data specialists who support research and learning in social sciences at academic institutions, government agencies, research centres, and statistical agencies. The 40th annual conference is being hosted jointly by Toronto's three universities: Ryerson University, the University of Toronto, and York University. The conference's agenda reflects the cutting edge in data science where sessions are devoted to best practices in data archiving and dissemination, and workshops on data visualization and analytics.

The ready availability of data determines what is researched by the world's very best researchers. Readily and freely available data improves the likelihood of a phenomenon, a people, or an economy, being the focus of research. On the other hand, data secrecy could mean a lack of interest by bright researchers who may gravitate to studying behavior, challenges, and phenomenon that are well documented with rich data sets.

The Economist magazine reported on a systematic analysis of 76,000 articles published in the leading economics journals to determine what countries received the most interest by leading researchers. The result was hardly a surprise. An overwhelming majority of papers published in the leading academic journals were focused on the United States (see the graph below). "There were more papers focused on the United States than on Europe, Asia, Latin America, the Middle East and Africa combined."

2014-06-04-Economist76000papers.png

Source: The Economist, January 2014.

The question is why the United States attracts such attention by the researchers. Even India, despite having a very large number of leading expatriate economists, such as Amartya Sen (Nobel Laureate), Kaushik Basu (the World Bank), and Raghuram Rajan (formerly the IMF), hardly generated any scholarship in the leading journals. "The American Economic Review, the holy grail for many academics, published one paper on India, by some measures the world's third-largest economy, every two years," wrote The Economist.

One could explain this U.S.-centric publication frenzy for two reasons. The one obvious culprit perhaps is the bias of the editorial boards dominated by the U.S.-based academics who often see little value in scholarship that does not address challenges faced by the American economy. The other reason is the ready availability of 'troves of good data' on the U.S. economy that makes it easier to study the U.S. economy than that of others. Even star Canadian researchers in economics and management overlook Canada and focus on the U.S. in their high-profile academic publications. One big reason is the lack of readily available data on Canada.

Data-centric research has taken a big hit under the Harper government that turned the mandatory long form Census into a voluntary survey. This has grave implications for evidence-based planning in Canada. The radical change in the survey instrument implies that statistics computed on Census data from 2006 and before cannot be compared with the data captured in the voluntary survey in 2011. This discontinuity in data will hurt government planning and private sector strategies.

Even more damaging is the under-representation of marginalized communities across Canada in the revised survey in 2011. Statistics Canada is deeply concerned about the quality of the collected data. It has suppressed data for several jurisdictions where it could not certify the quality and representativeness of the surveyed data. These difficulties are a direct result of the unconsidered changes made to the Census.

Given that respondents were free not to respond to the Survey in 2011, which was mandatory in 2006, the data is now affected by the self-selection bias, which leads to biased data with limited utility. The budget cuts by the Harper government have weakened Statistics Canada to such extent that it does not have the means to analyze the data it spent millions to collect.

It is disappointing to see that while governments and businesses are embracing big data and analytics, Canada has decided to turn its back on data and evidence-based planning. Smart planning cannot take place without good data and analytics. This is also the key take away from the data scientists congress at Ryerson University.