September 14, 2020

Really? Social Statistics in the Throne Speech?

By Peter Hicks

The “new normal” can lead to a better quality of life for Canadians

The new session of Parliament provides an opportunity to make major gains in the effectiveness of social policies by greatly strengthening the evidence on which those policies are based. We have many recent examples of how crucial it is for public policies to be based on solid scientific data, whether that be physical, biological, or social science.

We have a great opportunity to incorporate recent improvements in the statistical system, as well as lessons from COVID. Transformative new reforms can be introduced which can improve the quality of government.

This article summarizes the gaps that now exist, recent progress, and how important new steps could be signalled in a Speech from the Throne and operationalized by ministerial mandate letters.

The challenge

A main reason for having a national system of social statistics is to better understand trends, levels, and inequalities, in wellbeing. We need to know if things are getting better or worse overall and for different groups within the population – and for the various components of wellbeing: income, skills, health, social relationships, and happiness. We need to understand how existing government policies affect wellbeing and how alternatives to those policies could improve wellbeing.

Three large gaps would have to be filled before the system could be said to be doing a good job in meeting these policy objectives.

First, there is no capacity that is current, comparative, and authoritative, for monitoring the various dimensions of wellbeing. At present, monitoring is largely on a survey by survey basis. There is no analysis of how trends in the specific area being measured are related to other changes that affect our quality of life.

The system is even weaker in measuring wellbeing in different population groups. Most existing indicators ignore the multiple factors that affect wellbeing in different groups such as combinations of poor jobs, health, skills, living alone, housing and systemic discrimination.

Second, we lack much of the detailed information about individuals, and how people change over time, that is essential for understanding which combinations of programs and tax measures are working best. We do have good data on the immediate effects of programs that provide income. There is little data on the effects on programs that provide services such as education, health, employment, and social services. In consequence there is virtually no information about what really matters, namely the combined effects of income and service programs and related tax measures from all levels of government on the lives of individuals.

Third, we have failed to capture the benefits that could result from the use of current information technologies and new data sources in providing individuals with ‘what is likely to work best’ information related to decision-making in the social dimensions of life. These unexplored applications hold huge longer-term promise for improving the operation of labour markets, education, health care and other social programs. They hold promise of establishing new and more positive direct relationships between individual citizens and governments.

COVID 19 has highlighted many of these longstanding challenges. Post mortem studies will doubtless examine the merits of using common definitions, concepts and protocols that would allow for better sharing of lessons learned across the whole country on topics related to the spread of pandemics, to discovering risk factors for vulnerable populations, to predicting when and where outbreaks may happen and to evaluating how preventative and treatment measures are working.

Here we simply note that the pandemic has exposed the lack of a capacity for routine monitoring of progress in achieving wellbeing on a variety of social and health fronts – both overall and for different population groups, including older people in long-term care institutions, in racial communities and among indigenous people.

As well, we lack the statistical information to seriously assess policy scenarios that are now being discussed for a post-pandemic normal.

For example, some see a basic income as part of the post-pandemic package of social programs. Others see a higher payoff from using similar investments to improve targeted income support, improving the income and benefits that come from holding jobs, or improving services such as home care and the skills upgrading needed to find decent jobs in a greatly changing labour market. It is hugely frustrating that we lack the statistical information to undertake even basic comparative analysis of the longer-terms costs and benefits of moving in these different policy directions.

It is also possible to see the pandemic as providing a push to finally start implementing the long-standing, but little-acted-on, rhetoric about citizen-centred services, including patient-driven health care. However, as noted, we have done little to develop the ‘what is likely to work best’ data that is a necessary part of an effective system of services and supports that reflects the diverse needs and circumstances of real people.

Future directions

These are long-standing challenges that are not unique to Canada. Over the longer-term, we need to develop:

A conceptual framework that provides an integrative function for social statistics similar to that played by the System of National Accounts in economic statistics.
A much stronger capacity for publishing current indicators of progress and problems across all social areas in an integrated, comparative manner.
A capacity for providing user-friendly access to information from a statistical system that collects vastly more data from non-traditional sources than at present.
Partnerships with agencies which have direct responsibilities in different social and health program areas to provide tailor-made ‘what is likely to work best for you’ information directly to individuals as they make decisions in the employment, learning and health domains of life.

The new integrative conceptual framework will set out the concepts, definitions, and protocols to be used across the whole system of social statistics, as opposed to the present focus on separate collection vehicles such as surveys and administrative data. This new function will be supported by advice from a broadly-based consultative network of users and experts. This network, in turn, will help break down the silos that now impede policy action that crosses professional and jurisdictional boundaries.

The new capacity for releasing current indicators of wellbeing that are comparative and integrated across different social and health areas will show indicators for different parts of the country and for different groups in the population, as well as facilitating national and international comparisons. The most recent data in any particular social area will be released in a context that shows how those data fit into the larger picture. There will be ready access to the supporting data that is needed to assess cause and effect. This function will eventually become the main access point for all the data in the social statistics system. Only Statistic Canada can perform this function well since only it has the mandate and expertise to develop the needed standards and common definitions needed to ensure consistency across the various social areas.

New approaches are needed to provide easy access to the detailed data about individuals and their life histories that will be collected. This includes data showing people’s interaction with social programs. It will also include new ‘synthetic’ information that is based on calculations that make use of the overlapping information about individuals found on different data sources. Administrative data and other non-traditional sources of information will play an increasingly large role. Surveys will increasing be used to fill gaps left by other data sources.

Partnerships are needed to develop the ‘what is likely to work best’ data since key responsibilities, say for labour market information or health promotion, are in the mandates of different federal departments and of the provinces and indigenous governments. Yet in all cases there is a need for overlapping data about individuals that comes from multiple administrative and survey sources. Statistics Canada is the only agency that can provide consistent data from these multiple sources while ensuring quality and protecting privacy.

Next steps

We cannot move from where we are today to the more ideal world described above in one easy step. A gradual implementation strategy is needed to ensure that the merits of the existing system are reinforced and not put at risk. Indeed, Canada is already a leader in filling some of the gaps that were identified. Statistics Canada is undertaking an impressive modernization exercise that will result in a radical increase in the use of administrative data and data from non-traditional services.

Statistical agencies can take a lead on the technical side – how the needed statistics are produced. However, it is up to the centre of government to identify priorities for the kind of statistics that should be produced and the resources that should be allocated to their development. And here, as well, there has been much progress. Much emphasis has recently been placed on shifting to policies that are based on statistical evidence. For example, the last mandate letters for key federal Ministers called for the incorporation of quality of life measurements into government decision-making and budgeting, drawing on lessons from other jurisdictions such as New Zealand and Scotland. In those jurisdictions, efforts are underway to formulate policy priorities whose success can be measured by indicators of wellbeing of the sort that would be routinely produced by the new social indicators function referred to above.

The time seems ripe to take more concrete steps at the government-wide level.

This could include, for example, using Ministerial mandate letters that call for Statistics Canada to quickly develop a capacity for publishing indicators of social wellbeing that are current, comparative, authoritative and that facilitate in-depth follow-up analysis.

Statistics Canada could be mandated to begin work with other departments and with provinces to develop a database that describes, using consistent definitions, the operation of all social programs, including service programs.

The government might re-enforce its commitment to evidence-based policy by highlighting the role of Statistics Canada as the authoritative source of social data. This might involve explicit support for a continuation of recent work in building a modern system of social statistics based on fuller use of administrative data and a repositioning of traditional survey data – including the development of an appropriate new balance among issues related to response burden, trust, privacy and open access.

Responsible ministers could be mandated to identify lessons for the statistical system that were highlighted by COVID and to take remedial action.

In many social and health areas it will take considerable time to develop reliable ‘what is likely to work best’ information directed to individual users. However, much of that development work has already taken place in labour market programming. Employment and Social Development Canada has for many years used statistical techniques to determine what kinds of training and other interventions have worked best in the labour market in its evaluations of past programs; these techniques can be adapted to measure the likely success of the options that are open to today’s job seekers. The shift to a new way of doing business could be launched by giving Employment and Social Development Canada the mandate to work with Statistics Canada and selected provinces in developing pilot studies and demonstrations.

The steps above appear small but they would represent a milestone shift towards more effective social policies in Canada, towards the creation of evidence that will allow the multiple social programs of all orders of government to work together efficiently and harmoniously in supporting the wellbeing of Canadians. A Throne Speech commitment to these reforms might not have high drama, but would signal a solid commitment to better government planning and accountability.

Peter Hicks, who worked for many years as an assistant deputy minIster in social departments and central agencies in Ottawa and at the OECD in Paris, writes about and advises on matters related to social policy.