The seed for this article was sown by a 1999 op-ed piece by this magazine’s current editor, William Watson. Writing about tax increases, Mr. Watson argued that Canada is not ”œtough on kids,” noting that the Child Tax Benefit (CTB) ”œ… is more generous to those who do get it than anything in our history.” In a similar vein, in his 1995 book The Canadian Revolution: From Deference to Defiance, Peter C. Newman pointed out that ”œWhile baby bonus payments were terminated in 1989, they were replaced by a relatively generous program of child tax cred- its for the poor.”

How could Messrs. Watson and Newman, two well- informed observers of public policy, been unaware that in fact the CTB has meant reduced support for most poor chil- dren? Perhaps because they were never told otherwise. Every CTB impact statement concentrated on winners”” those whose support was increased. No mention was made of losers””those worse off. The impression was left then that there were no losers. That simply is not the case.

This paper is about two Child Tax Benefit casualties. The first are those poor families whose support level has declined under the CTB. The second is truth””in this case the quite simple truth that such families exist. That it, too, is a casualty is reflected in the misconceptions of Messrs. Watson and Newman. Reportedly the first casualty of war, truth evidently fares no better in battles over policy.

The two gentlemen’s statements reflect a widespread impression that the CTB has increased support for ”œthe poor.” This is only partially true. While the CTB has raised support for the working poor, support for kids on welfare has declined. After adjusting for inflation, since 1985 the CTB has lowered support for 90 per cent of children. This is for children in general, but what is even more surprising for an anti-poverty program is that support is lower for the majority of poor children!

It is not my purpose here to debate the strengths and weaknesses of the CTB. For the record, I view the CTB as a partial step forward (to the extent it has increased support for at least some low-income children) but only a partial step because by excluding welfare children it excluded most poor children and indeed the neediest. The most recent CTB changes were intended to reduce the ”œwelfare wall” (the strong disincentives to work because of the finan- cial and other supports that are lost when welfare families accept employment). Yet the welfare wall could have been eliminated without excluding welfare families from increased support, which is what the CTB did. Indeed, the CTB did not tear down the wall so much as relocate it in front of the many more families who have incomes between $20,000 and $30,000.

But my central purpose here is to evaluate, not the CTB itself, but the several CTB impact statements that have led to the common mis- conception that low-income families are all assisted by the CTB initiatives, that ”œthere are no losers.” Far from being unaffected by the CTB, welfare families have seen their support levels continue to decline. They are the hidden casual- ties of the CTB initiative.

Intelligent policy debate requires at least some factual information on which participants can agreed. The impact analyses which have accompanied the various policy changes would ideally inform the public debate. Yet, given the very political context in which they are present- ed, they are also intended to influence public opinion. These twin objectives of both inform- ing and persuading make awkward and some- times conflicted companions. I will argue that persuading has usually interfered with and sometimes even eliminated informing.

An impact analysis accompanies most federal and provincial budgets, as well as major program changes, such as the replacement of Unemployment Insurance with Employment Insurance and the evolution of the CTB. Impact analyses describe the effect of program changes on Canadians. They answer questions like: How many people are better-off/worse-off? By how much? Who are they?

Factual accuracy is important in such analy- sis, but, as the examples that follow will show, accuracy certainly does not ensure clarity. Indeed, documents can at the same time be accurate and misleading, especially when the objective of truly informing the public takes a back seat to persuasion.

Impacts are inherently about change. Impact analysis involves a number of choices. What years are used for comparison? Do we adjust for inflation? Are total expenditures pre- sented in dollars or as a per cent of government spending, government revenue or GDP? If we are illustrating the impact using typical families, what family types are used and at what income levels? Is the impact shown in dollars, as a per- centage change or as a per cent of income? A single factual base can lead to a multitude of pos- sible impact statements, all accurate, yet each one telling a very different and sometimes mis- leading story.

From 1985 through 1999 federal impact statements have reflected the (perhaps natural) tendency for governments to put the CTB pro- gram in the best possible light. They have typi- cally emphasized winners while de-emphasizing or even ignoring losers. Unfortunately, over the long term this understandable political predispo- sition to accentuate the positive has produced a general misconception that the CTB assists all low-income families.

We begin in 1985 with Michael Wilson’s first budget, which increased the target- ing of the child benefit system by reducing the Child Tax Exemption and increasing the refund- able Child Tax Credit. As a result, there were some losers and a few winners.

Figure 1 was included in the budget papers. It implied improved benefits for those with annual incomes less than $30,000 and losses for those over $45,000. The graph was not adjusted for inflation, however, which in 1985 was about five per cent, more than double what it is now. This failure to provide indexed numbers is iron- ic in view of the fact this is the budget that intro- duced partial indexation, which caused bracket creep and the automatic erosion of supports like the CTB that continued for 15 years until the 2000 budget restored full indexation.

In fact, had the inflation adjustment been made, the graph would have shown that the only winners were those very few families with income below $10,000 (in $1985)””and their gain was wiped out by inflation by 1992””and that losses under the plan started at $25,000.

Seven years later, the 1992 budget announced a major structural overhaul of the child ben- efit system. Family allowance, which had been clawed-back for a couple years, was eliminated. So was the non-refundable child tax credit (which had earlier been the child tax exemp- tion). The payment of the refundable child tax credit was changed from annual to monthly cheques. As well, a Working Income Supplement was introduced for poor families with labour market earnings.

For middle and higher-income families the impact of these changes depended on their cir- cumstances, their income and the number of their children. For poor families the impact was pretty clear and depended on earnings. The Working Income Supplement increased support for the working poor. For welfare families the total income support was unchanged at $1,020 per year.

The impact analysis, a 12-page ”œwhite paper,” demonstrated how an analysis can be perverted by the selective use of ”œaccurate” statements. The following quotes and chart from the 1992 paper leave the strong””but incorrect”” impression that support levels were being increased significantly for all poor fami- lies: ”œThe child benefit will be delivered month- ly and will increase assistance to families with low and middle incomes.” No mention is made that some, or indeed most, of the poor are excluded.

”œLower-income families with one child will receive a monthly payment of up to $144, sig- nificantly more than the monthly payment of $35 they now receive. Approximately two mil- lion families with incomes below $50,000 will receive larger monthly payments than they do now.” This goes well beyond selective reporting. The device of concentrating on the monthly cheque (ignoring the loss of an annual cheque), enabled Ottawa to make factually correct, but entirely misleading claims. The focus on the ”œmonthly cheque” contributes to the illusion that ”œeveryone wins.” In fact, poor families on welfare received no increase in annual support. Indeed, since their income remained at $1,020, inflation ensured that their purchasing power declined.

Figure 2 is reproduced from the ”œwhite paper.” It continues the bizarre charade of deter- mining the impact of the 1992 changes based solely on the size of the ”œmonthly” cheques. The first footnote to this chart indicates that the authors knew pertinent information was being excluded.

Thus the ”œwhite paper” presented an analysis without losers that led readers to think support for poor families had increased across the board. The Globe and Mail, for one, appears to have fall- en for the ruse: ”œMr. Mazankowski’s reforms will sharply raise support for low-income families with children,” it reported. ”œThe plan will allow Ottawa to make monthly payments of up to $144 for each child in a low-income family, four times the current family allowance. About two million families with incomes below $50,000 a year will receive larger monthly payments than they now do.”

The inner workings of government
Keep track of who’s doing what to get federal policy made. In The Functionary.
The Functionary
Our newsletter about the public service. Nominated for a Digital Publishing Award.

A publicity blitz following the ”œwhite paper” suggests that some federal officials may have lost their apolitical footing. One Finance official boasted: ”œWe have mailed out 350,000 copies of the white paper that we tabled at the time of the February Budget, which probably, in terms of white papers, makes it a runaway best seller …” One can only wonder why supposedly politically neutral officials would celebrate the popularity of such a misleading document. An even more disturbing question is how many low-income parents made purchasing decisions based on the false assurance that their income was increasing? Nowhere did budget documents published with the proposal acknowledge that on an annual basis most low-income children would receive the same funds. Casual observers of public policy can therefore be forgiven for believing that support for poor children was being increased. Only policy analysts specializ- ing in income support would recognize that support for welfare families was held constant and declined in real terms. The ”œmonthly” cheques contrivance goes beyond ”œplacing the program in the best light” and, in fact, is fraud- ulent. If it were being used to advertise appli- ances, this sort of practice very likely would lead to a court appearance.

Turn now to the CTB announced in the 1997 budget. It added another layer of complexi- ty: the co-ordination of federal and provincial support levels. The federal government increased its support for low-income children on condition that provincial social assistance spending be clawed back dollar for dollar. Provinces agreed to ”œreinvest” the money thus freed up in a range of supports for low-income children””which may or may not include chil- dren whose families are on social assistance. Provinces also promised that no welfare child would be worse off after the change. (Newfoundland and New Brunswick did not take part in the claw-back scheme.)

Most of the controversy surrounding the CTB has concerned this agreement that provinces would ”œclaw back” the increased fed- eral support. Information about the clawback’s impact is clearly critical to the policy debate about the CTB and its impact as an anti-poverty measure. Unfortunately, the impact analyses which accompanied the changes failed to inform the debate about the clawback.

The analyses published with the 1997 budg- et included the following statement, which overstates the number of those benefiting from the CTB by simply ignoring the provincial claw- back. ”œSome 1.4 million families with 2.5 mil- lion children would receive higher federal bene- fits.” Use of the word ”œfederal” is critical to the truth of this statement since the clawback ensures that welfare families will lose their feder- al increase to the province. But if the number assisted federally is worth documenting, then why wouldn’t the number who are still better off after the clawback not be just as significant and newsworthy?

The following government statement from that year’s Budget Plan suggests the program benefits the ”œneediest children.” [The program will] … provide better support for low-income families with children … Over time, this initia- tive will build a more secure base of benefits and services for Canada’s neediest children.” But while most observers would argue that welfare children are Canada’s neediest their base of ben- efits and services is far from secure. Welfare is now the program of last resort. To receive it, families must have even exhausted their finan- cial assets. In reality, welfare families are not assisted by the Child Tax Benefit but instead are further excluded from society by its provisions. In fact, the CTB’s treatment of welfare families has been the its most contentious aspect. It has been opposed by municipal governments, the UN committee on economic, social and cultural rights, and virtually all social policy research and advocacy groups.

The impact analysis of the CTB’s anti-poverty value might have included, at the very least, the number of CTB recipients being clawed-back. Children in such low-income families receive no direct benefit from the CTB initiative. But so far, the only data available on the number clawed back have been published by the National Council of Welfare, a citizen body which advises the Minister of Human Resources Development. In a 1998 paper, it estimated that because of the clawback 64 per cent of poor children and 83 per cent of poor children of single parents do not receive any increased income support from the CTB. Getting a rough estimate of these figures is not difficult. One only needs to know the propor- tion of poor children whose parents rely on wel- fare income. The Council’s estimate was roughly comparable to available data on social assistance: The nation’s one million children relying on wel- fare are about 70 per cent of its per children.

When these numbers were reported in the Toronto Star federal officials were said to ”œdispute the figures.” This is difficult to reconcile with my own Access to Information request for these data. The reply to my request was in part: ”œPlease be informed that no reports, analyses, deck or other presentations, studies [sic] were done or commissioned by Human Resources Development Canada on the impact of changes to the child benefit system and the Canada Child Tax Benefit on Canadian low-income fam- ilies.” On what possible basis then did HRDC ”œdispute” the Council of Welfare’s figures?

The confusion is further increased by the fol- lowing government statement in a submission to the United Nations ”œ… a new National Child Benefit (NCB) system was implemented on July 1, 1998 as a collaborative initiative by federal, provincial and territorial governments; about half of the beneficiaries are single-parent families headed by women.” Most readers of the UN docu- ment presumably accepted the statement at face value and concluded that this ”œcollaborative ini- tiative” is targeted to assist female single parents.

But the statement directly contradicts the Council of Welfare estimate that very few single parents are assisted by the CTB. Can both statements be true? The statement to the UN must simply ignore the clawback. After the clawback, the Council of Welfare points out, the vast majority of single- parents see no increase in support. So the only public data on the number of poor children sub- ject to the clawback are produced by one federal government body yet are disputed by HRDC offi- cials who have published none of their own, having informed me they’ve done no calcula- tions of their own. But HRDC has sufficient data to make the UN statement, which only confuses the legitimate policy debate.

HRDC’s second policy assertion is that sup- port for welfare children will not decline with the CTB. We are told that provinces have agreed that children will receive at least as much income support as they did previously. As their social union website puts it: ”œUnder the National Child Benefit, families who remain on social assistance will not see a reduction in their com- bined income support.” Or, as they put it in a July 1999 statement ”œFamilies on social assis- tance will continue to receive at least the same level of income support from governments as under the previous arrangement.”

There are a number of problems with these assertions. Along with the federal income tax system, the CTB has been subject to only partial indexation since 1985. For the income tax system, the result has been the well-publicized ”œbracket creep.” For the CTB, it has been declin- ing purchasing power. This has been true for all recipients of the CTB, but the working poor, at least, have enjoyed real increases in support which have more than offset the erosion due to inflation; not so welfare families. Until February of this year, the last budget that increased CTB supports for welfare families was 1985 (with the increases phased in between 1986 and 1989).

Support payments per child on welfare peaked in 1989 at $1,241 (in 2000$). By 2000 it had dropped to $1,020””a decline of more than $200 per child or about 20 per cent. To argue that no welfare child is worse off requires wilful blind- ness to the amounts eroded away by inflation.

As noted, full indexation of the CTB was restored in the 2000 federal budget, which also provided a small increase in the actual support to children on welfare (about $4 per month after inflation). But that still doesn’t end problems with the assertion that welfare children are ”œno worse off.” Recent announcements of cuts to social assistance in Ontario and Nova Scotia would appear to violate the federal-provincial agreement that no child would be ”œworse off.” Ontario’s welfare reductions are in STEP (Supports to Employment Program). The pro- gram cuts involved will reduce incentives for welfare recipients who find employment. Amazingly, Ontario is rebuilding the welfare wall that the CTB was designed to knock down! Nova Scotia has also announced reductions in welfare rates that would seem to violate the agreement that no children on social assistance see a reduc- tion in income support.

That the Child Tax Benefit is being marketed as an anti-poverty program may explain why many people assume it benefits poor children. In a recent paper for the Caledon Institute, Tom Kent decried the use of marketing by political parties. ”œNationally, the parties have become lit- tle more than advertising machines, competing for public support by much the same techniques of imagery as businesses with almost identical products use to compete for market share of detergents or banking services or whatever.”

Mr. Kent’s comments predate the recent del- uge of political advertising masquerading as pol- icy information, and the ludicrous display of the federal and Ontario governments using taxpay- ers dollars to pitch their competing versions of the federal share of health spending. (Because the CHST covers health, post-secondary education and social assistance, this number is in fact unknowable.) Mr. Kent’s critique applies equally to many government docu- ments, including the ones referred to here. We should also be concerned that a traditionally apolitical civil service is regularly preparing documents that use data selectively to help market whoever happens to be the governing political policy.

A related issue concerns the openness reflected in the Access to Information Act. The response to my Access request was that no data on the poverty impact of the CTB existed. Yet HRDC officials publicly disputed data from the Council of Welfare, while offering none of their own. Government is often the only source of much data perti- nent to policy analysis and debate. In my experience, gov- ernment practices fall far short of anything that could be called openness. The only data HRDC has made available places the program in a positive light and the Ministry reportedly ”œdisputes” the only data that exist on those poor families neglected by the program.

The need to inform and the urge to persuade arise in two other fields: commercial advertising and accounting. In advertising, industry standards have evolved to ban mis- leading statements. The goal is not only accuracy but fairness””perhaps honesty is a better word. The current situa- tion is far from perfect. Firms continue to make court appearances over misrepresentations. But honesty is at least the declared goal of industry self-regulation.

The accounting profession has also developed a set of accepted standards that facilitate the presentation of com- panies’ financial data, thus making cross-company compar- isons possible. While not perfect, financial statements are light-years ahead of statements made in public policy debates where groups pick and choose data, assumptions and indicators more to persuade than inform. This is true not only of governments but think-tanks and advocacy groups as well. It is time we established standards for the publication of impact analyses. If we take seriously the idea of a democracy based on informed policy debate, then we should require that data released by government attain a much higher standard than now exists.

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

Creative Commons License