Categories
Employment Financial security Governance Poverty

Why minimum wage policies are ineffective: A brief overview, plus Ontario retail sector example

Chart 1 Change in hours and wages, Retail trade, Ontario, 1983-2014*
Chart 1 Change in hours and wages, Retail trade, Ontario, 1983-2014
Source(s): CANSIM tables for Survey of Employment Payroll and Hours (SEPH), Statistics Canada. (See Note 6)

Encouraged by recent initiatives in relatively more prosperous US jurisdictions (Seattle and San Francisco), Canadian labour groups, particularly those out west (BC and Alberta), have taken to calling for dramatic minimum wage hikes. These labour groups cite research they interpret to mean that min-wages do not affect (dis)employment, and assert that mandated wage hikes will help address income inequality and/or alleviate poverty. Unfortunately, the min-wage research to date does not support this assertion. That may have more to do with the challenge of distinguishing between min-wage and other effects. Perhaps the best approach is not to try to discern the direct effect of min-wage policy at all, but rather infer it from broader labour market trends.

One of the major challenges facing researchers looking into minimum wage effects is that there really isn’t a universal minimum wage. Minimum wage policy falls under labour standards in Canada, and therefore provincial jurisdiction. Looking at the min-wage schedule of a province (like Ontario), one notices there isn’t a single min-wage in effect at any given point in time, nor does min-wage apply to all workers.

As a result, few workers at any point in time receive a form of min-wage. Those that do receive different rates given the defined rate schedules. And few continue at a given rate for any extended period, as even min-wage earners receive incremental raises over time.

This has made meaningful research and analysis of min-wage policy effects quite challenging. Out of necessity, this has spawned some rather novel approaches.

One such approach involves focusing on a specific population subgroup more likely to earn min-wage irrespective of type of work or geographic location – such as teenagers.1 Most teens who work would be between the ages of 16 and 19. They’d work either full or part-time while pursuing secondary or post-secondary education full-time – meaning they’d largely be excluded from the labour force and other related household surveys (topic for another day). Or they’d drop out of school to work full-time – a scenario more common among lower-income households. Not surprisingly, researchers using this approach have found that “a 10% increase in the minimum wage is significantly correlated with a 3%–5% drop in teen employment,” and that “a higher minimum wage may (not-so-)paradoxically result in a significant negative shock to household income among low-income families.”

Another approach is to broaden the scope to households with at least one person earning min-wage and look at the total effect on household earnings.2 This approach tends to magnify the argument against min-wage as a policy instrument for income inequality and/or poverty alleviation. It basically combines the teen disemployment argument above with the argument that most minimum wage earners live in a household with at least one individual earning above min-wage. Whether and how that’s a compelling argument against such a policy is debatable.

Because of the limited number of workers earning min-wage at any point in time and the broad range of earnings considered to be min-wage, some researchers choose to broaden the income range examined to assess policy implications.3 The argument is that since min-wage policy sets a floor, its effects would spill over to those within an earnings range above (and below) the legislated schedule. While finding some spillover effects in the US, researchers using this approach noted “the effects of the minimum wage on the wage distribution in Canada do not reach as high up the distribution as in the United States,” and that there are “relatively modest spillover effects in the Canadian data.”

Simply ignoring the complexities involved in assessing the policy implications of min-wage legislation and estimating the share of the population earning a composite real min-wage rate across all sectors nationwide is another approach.4 While making for far simpler analysis, not surprisingly this approach doesn’t shed much light on the issue. One such study found that the real min-wage in Canada hasn’t changed much since the 70’s, nor has the minimum-to-average wage ratio. Although the share of Canadians earning this composite real min-wage has apparently increased, this change is attributable to the lack of bargaining power of low-wage workers who simply find themselves earning min-wage as a result of small, incremental bumps to the legislated wage rates over time.

Then there’s the simplest correlation approach, favoured by labour unions and their ‘think’ tanks.5 The number of jobs has gone up over time, and so has min-wage, ergo raising the minimum wage has no impact on employment. Seriously.

A more interesting approach would be to look at specific sectors with relatively high concentration of min and low-wage workers in specific jurisdictions and try to assess how sensitive employment in those sectors would be to changes in wages.

Chart 1 illustrates a fairly crude example of this approach. It shows payroll hours in the retail sector adjusted for population growth along with estimated wages adjusted for inflation (using all-items CPI) for the province of Ontario between 1983 and 2014.6  Retail sales probably more than anything else strongly correlate with population growth. Retail payroll hours could be affected by a number of factors, but the two most likely sources are wages and technology (which this crude approach obviously doesn’t factor for – a measure of labour productivity could work, provided correlation could be factored for) . Using hours instead of employment makes more sense as it eliminates change in hours per worker over time; as such, hourly wages including over-time for employees paid by the hour are used to even things out.

While historical population and inflation trends in Ontario have to be kept in perspective (for example, relatively much higher inflation and population growth in the 1980s vs the 1990s), this crude approach nevertheless appears to suggest a relatively strong negative correlation between changes in payroll hours and changes in wages in a sector with relatively high concentration of min and low-wage earners; As real wages in the Ontario retail sector rise, relative payroll hours tend to decline, and vice-versa.

This isn’t just another anecdote suggesting minimum wage hikes have potential disemployment effects. As a review of the research from the early 1990s onward succinctly summarised it: “First, we see very few – if any – studies that provide convincing evidence of positive employment effects of minimum wages, especially from those studies that focus on the broader groups (rather than a narrow industry) for which the competitive model predicts disemployment effects. Second, the studies that focus on the least-skilled groups provide relatively overwhelming evidence of stronger disemployment effects for these groups.”7

Canadian labour unions’ perspective aside, there seems to be little evidence to suggest boosting the minimum wage even incrementally over time has resulted in tangible benefit in terms of addressing income inequality or poverty, and plenty of evidence to suggest it’s had disemployment effects, particularly on low-income households. How labour groups anticipate hiking the minimum wage by as much as fifty percent will somehow have the opposite effect is an interesting question, to put it kindly.

This isn’t to suggest that simply doing nothing is a solution. Far too many Canadian households are falling behind, and sound policy proposals to address rising income inequality and persistent poverty are as necessary as ever. Quite simply, boosting the minimum wage to address these very real problems is a case of misallocated political capital.

Notes:

1. Teen employment, poverty, and the minimum wage: Evidence from Canada, Anindya Sen, Kathleen Rybczynski, Corey Van De Waal, Labour Economics Volume 18, Issue 1, January 2011.
http://dx.doi.org/10.1016/j.labeco.2010.06.003

2. The (Non) Impact of Minimum Wages on Poverty: Regression and Simulation Evidence for Canada, Michele Campolieti, Morley Gunderson, Byron Lee, Journal of Labor Research, Volume 33, Issue 3, September 2012.
http://dx.doi.org/10.1007/s12122-012-9139-8

3. Minimum Wages and Wage Spillovers in Canada, Michele Campolieti, Canadian Public Policy, Volume 41, Issue 1, March 2015.
http://dx.doi.org/10.3138/cpp.2013-060

4. The ups and downs of minimum wage, Diane Galarneau, Eric Fecteau, Insights on Canadian Society, Statistics Canada, July 2014.
http://www.statcan.gc.ca/pub/75-006-x/2014001/article/14035-eng.pdf

5. Dispelling Minimum Wage Mythology: The Minimum Wage and the Impact on Jobs in Canada, 1983–2012, Jordan Brennan, Jim Stanford, Canadian Centre for Policy Alternatives, October 2014.
https://www.policyalternatives.ca/sites/default/files/uploads/publications/National%20Office/2014/10/Dispelling_Minimum_Wage_Mythology.pdf

6. These are really two separate series, one covering changes between 1983 and 2000, and the other covering changes between 2001 and 2014. The available Survey of Employment Payroll and Hours (SEPH) CANSIM tables cover the first reference period based on 1980 Standard Industrial Classification (SIC), and the second based on 2012 North American Industry Classification System (NAICS). There is a significant difference between the two classifications for Retail trade industries. There’s a third SEPH series that covers 1991 to 2000 which appears incompatible with the two series illustrated. While a concordance between 1980 SIC and 2012 NAICS is possible, it’s beyond the scope of this simple post.
http://www.statcan.gc.ca/eng/concepts/concordances-classifications

7. Minimum Wages and Employment: A Review of Evidence from the New Minimum Wage Research, David Neumark, William Wascher, NBER Working Paper No. 12663, November 2006
http://www.nber.org/papers/w12663

Leave a Reply

Your email address will not be published. Required fields are marked *