In 1990, I wrote an article for Policy Options titled “Social Policy by Stealthunder the pseudonym Grattan Gray. The article was a comprehensive critique of social and tax policy under the Mulroney government.

In it, I characterized the concept of social policy by stealth:

It relies heavily on technical amendments to taxes and transfers that are as difficult to explain as they are to understand and thus largely escape media scrutiny and public attention. It camouflages regressive changes in the rhetoric of equity in an attempt to convince Canadians that tax increases are tax cuts and that benefit cuts are benefit increases. By further complicating an already complex labyrinth of taxes and benefit programs, the stealth style of policy-making confuses the electorate and so insulates itself from criticism.

The phrase “social policy by stealth” resonated with NGOs, governments, media and academia. It became part of Canada’s public-policy lexicon. In his first budget, then-finance minister Paul Martin vowed to set aside “old tactics of stealth and surprise.” In later budgets, he imposed billions of stealth-driven tax hikes and benefit cuts. Governments seem to love the stealth approach because history proves they can get away with it − for a while at least.   

One of the people intrigued by social policy by stealth was Alan Broadbent, president of the influential Maytree Foundation. In 1992, Alan and I co-founded the Caledon Institute of Social Policy. Not surprisingly, Caledon made liberal use of the concept of social policy by stealth over the years, in keeping with its tradition of using colourful language.

But social policy by stealth is more than just a catchy phrase; it is an often-overlooked dimension of how government decisions shape social policy in subtle but important ways. Social policy by stealth has two main dimensions: indexation and complexity. Understanding these dimensions allows us to better understand and design social policy.


Income support and replacement programs typically adjust benefit rates at least annually to keep up in full with the cost of living. However, sometimes governments reduce a program’s indexation to less than the full cost of living (partial indexation) or they do not index at all.

Most programs – including Employment Insurance, the Canada Workers Benefit for low-wage earners (formerly the Working Income Tax Benefit), seniors’ benefits (Old Age Security, the Guaranteed Income Supplement and the Spouse’s Allowance), and the Canada and Quebec Pension Plans – are fully indexed. Provincial/territorial minimum wages are a mixed bag; some are indexed, and some are not. Almost all welfare programs are not indexed.

Child benefits, an income program for families with children, are a good example of erosion of value by means of partial indexation or no indexation. While the programs have changed considerably over the years, they have rarely been fully indexed. Consider the formulas: family allowances (partially indexed in 1986), the refundable child tax credit’s income threshold (1986 partial), the child tax benefit (1993 partial), the Universal Child Care Benefit 2006 (not indexed).

In July 2016, the current federal government launched the excellent Canada Child Benefit, yet de-indexed its flagship program for its first five years − eroding the real value of payments, harming the poor the most. The five-year freeze on indexation brought little opposition, save criticisms from Maytree and Caledon.

The benefit started off paying a maximum $6,400 for a child under 6. With inflation running at an approximate annual rate of 2 percent, the purchasing power of the benefit probably fell to around $6,272 in 2017, declining further this year to around $5,903 in constant 2016 dollars.

Fortunately, Ottawa took steps to fix the problem. In the fall of 2017, the minister of finance announced that full indexation of the Canada Child Benefit (including its family-income threshold) would be applied in July 2018 rather than 2020.

If it were left de-indexed, the shrinking value of the Canada Child Benefit would have weakened the program. Lack of indexation steadily lowers the income threshold for maximum benefits, with fewer low-income households qualifying for the maximum benefit.

Stealth has also afflicted the personal income tax system. It is now fully indexed, but for a time was partially indexed, with devastating effects.

From 1973 to 1985, Canada’s personal income tax system was fully indexed. Personal exemptions and tax brackets were adjusted each year to the change in the rate of inflation. But starting in 1986, the government limited indexation to three per cent less than the rate of inflation, which by the 1990s effectively froze the brackets in nominal terms and lowered them in real terms. This pernicious measure lowered the federal income tax threshold from $10,505 in 1980 to $7,112 in 1998, and an estimated $6,964 in 2003 (expressed in 1980 dollars). More than 1.9 million taxpayers were pushed from the bottom to the middle tax bracket, and 600,000 from the middle to the top bracket.

Partial indexation imposed an annual hidden income-tax hike on all taxpayers, but the burden fell disproportionately on the working poor. I explored these machinations in two Caledon reports: No Taxation without Indexation (1998) and Credit Corrosion: Bracket Creep’s Evil Twin (1999).

Over the years, most income programs and taxes have been afflicted by stealth. The key exception is Old Age Security, which the Conservative government tried but failed to partially index in its ill-fated 1985 budget.

The main driver for stealth is cost. Governments have turned to partial indexation or non-indexation to silently siphon out billions of dollars in benefits and taxes. The move has been a potent, effectively secret weapon in the war on the deficit. It has boosted tax revenues from federal and provincial/territorial governments while trimming Ottawa’s expenditures on child benefits and transfer payments to the provinces/territories for health and postsecondary education.


The other dimension of stealth is complexity, specifically the structure of social programs.

The child benefit system is particularly prone to counteracting changes in programs and taxes whereby progressive reforms are undone by regressive measures. Confusion arises through complexity, and there is a succession of modifications in program design that only a handful of experts can decipher.

Canadian social policy enjoyed a watershed year in 1993. The federal government launched a major structural reform of federal child benefits, which replaced Family Allowances and the refundable and non-refundable tax credits with a single, geared-to-income Child Tax Benefit. The provinces and territories joined in by reforming their welfare-embedded child benefits.

The new Child Tax Benefit replaced three programs:

  1. Family Allowances: Roughly progressive and nominally universal (though with an income test that effectively ended their universality).
  2. The refundable Child Tax Credit: Very progressive and geared to poor families.
  3. The non-refundable Child Tax Credit: Replaced the regressive Children’s Tax Exemption, was a quasi-universal grant that delivered the same federal and provincial/territorial income tax savings to all non-poor families. Because the credit was non-refundable, poor families without taxable income got nothing.

This mismatched trio of child benefits was inequitable and surely incomprehensible to most parents and probably most politicians and media. The combined distribution of benefits bore no logical or defensible relationship to need as measured by family income. Families with the same income received different amounts of child benefits depending on each parent’s share of family income. The core aims of child benefits — reducing poverty and easing childrearing costs — were in tension.

The Child Tax Benefit was a non-stigmatizing, inclusive program that delivered its benefits to the large majority of Canadians through the same vehicle. It was portable, providing a stable and assured supplement to income no matter where families lived or moved. It was progressive, meaning benefits declined as incomes rose. What you saw was what you got because benefits were not subjected to income tax.

Matters got worse under the Harper government, which wanted to put its own stamp on child benefits. In 2006, it brought in the Universal Child Care Benefit, similar to the archaic Family Allowance. Benefits were subject to income tax, so most families did not receive the full payment. To confuse matters, Ottawa later exempted Universal Child Care Benefits from taxation in single-parent families, though in families with two parents, benefits remained taxable.

As a result, families of different types (single parents, one-earner and two-earner couples) with the same income received different amounts. Because their tax rates varied, families with the same income received different amounts of after-tax payments depending on their province or territory.

The Harper government added another social-policy zombie in 2007. The non-refundable Child Tax Credit was a rehash of the program of the same name that operated between 1988 and 1992. The “new $2,000 child tax credit” was neither new nor worth $2,000; its maximum actual value, in federal income taxes, was $300 per child. All non-poor families received $300 per child, including the rich. Some lower-income families with a small tax liability ($1 to $299) received a smaller amount, while the poorest got nothing because they do not pay tax. In 2015, the Conservatives eliminated the program to help pay for an enriched Universal Child Care Benefit.

The long list of child benefits grew to include the Family Tax Cut, a program so hard to fathom that leading economist Jonathan Rhys Kesselman wrote a trenchant critique. It took four pages to explain in the Notice to Amend the Income Tax Act. The federal government tried to explain the measure in plain language, with convoluted results that would be amusing were they not concerning.

The incoming Trudeau government got rid of the preceding flawed programs with the creation of its single, strong, income-tested Canada Child Benefit in 2016, building on the advantages of the earlier income-tested Child Tax Benefit.

Still Stealthy After All These Years?

Today, indexation stacks up pretty well. Most of Canada’s income programs and taxes are fully indexed: Old Age Security, the Guaranteed Supplement and the Spouse’s Allowance, Employment Insurance, the Canada/Quebec Pension Plan and the new Canada Workers Benefit and some minimum wages.

However, other programs are still complex. Employment Insurance has a multiple-rule structure based on unemployment rates across Canada. Recent reforms to the Canada Pension Plan are difficult to follow, especially because key features will be phased in over time. The Canada Pension Plan’s maximum $2,500 death benefit is frozen, so it falls in value each year. Welfare remains a labyrinth that seems impervious to reform. The majority of welfare systems remain un-indexed.

Social advocates should remain vigilant.

This article was written for Maytree as a follow-up to the event, “Celebrating 25 Years of the Caledon Institute of Social Policy.” The full Caledon archive, including all papers by Ken Battle, is here.

Photo: A crowd gathers at the Alberta Legislature to take part in Canada day celebrations in Edmonton Alta, on July 1, 2018.THE CANADIAN PRESS/Jason Franson

Do you have something to say about the article you just read? Be part of the Policy Options discussion, and send in your own submission. Here is a link on how to do it. | Souhaitez-vous réagir à cet article ? Joignez-vous aux débats d’Options politiques et soumettez-nous votre texte en suivant ces directives.

Ken Battle
Ken Battle is one of Canada’s leading social policy thinkers. He co-founded the Caledon Institute of Social Policy in 1992 with Alan Broadbent and served as its president until 2017.

You are welcome to republish this Policy Options article online or in print periodicals, under a Creative Commons/No Derivatives licence.

Creative Commons License

More like this