November 20, 2009

Significant HST Errors

Is the Ministry of Finance more likely to make an error in how much revenue the harmonized sales tax (HST) will return to government or in how much that tax will cost BC families? The Ministry's estimates on the cost to families and on government revenues are inconsistent.

This is not to question the government's intent to be accurate on both sets of estimates, but when it comes to estimating HST revenue all it has to do is take the most recent data on GST collected in BC (available from the federal government to the province, although not available to the general public), multiply by the ratio of the federal to the provincial tax (7/5) and adjust for the few items that BC excludes from the tax. For the impact on BC families, it is necessary to make more assumptions and to use data obtained through a Statistics Canada survey sample. When the two don't match, it is important to keep in mind that the tax data are more accurate.

In order to understand the basis for statements made by the government about the tax shift from business to BC families, I submitted freedom of information requests for details of how the impact on a typical family was calculated and for details of how benefits for the construction industry were determined.

Occasionally it is appropriate to break the mold and say something nice about the government, or at least about the public service. Analysts in the Ministry of Finance appreciated that available documents would be inadequate to answer my questions so through a team leader from the Ministry of Citizens' Services (now responsible for handling FOI requests) a telephone briefing was arranged to answer my questions. The briefing was extremely useful, demonstrating that when there is a will there is a way to comply with the original spirit of the freedom of information legislation.

In the briefing I was told the estimate of how much the HST would cost a hypothetical family of four with a $60,000 income was calculated by reducing the $60,000 to the amount of after tax disposable income, subtracting items like food and shelter that are non-taxable and then applying information from Statistics Canada's Survey of Household Expenditures to calculate the difference between what would be paid with the current PST compared to what would be paid with the proposed HST. That method seems reasonable, and using it I produce results similar to those published by the Ministry of Finance; however, that result appears to be inconsistent with what the Ministry has published with respect to the amount of revenue it expects from the HST and with the benefits to business. Furthermore that method yields an estimate of $278 per year in extra costs for the hypothetical family, while simply dividing the amount of the tax shift by the size of BC's population yields a substantially higher estimate.

Table 3 on page 85 of the September budget documents, titled "Estimated Net Fiscal Impact of Harmonization" (reproduced below), shows the province expects to receive $6.5 billion from the HST before rebates, compared to $4.9 billion from the PST, and that it intends to provide various rebates and personal income tax adjustments to mitigate some of the impact of the HST, resulting in a net loss of $40 million with the HST compared to the PST. In order to fund those mitigations it must first collect the estimated $6.5 billion, or $6.2 billion after instant point-of-sale rebates are deducted. That makes it look like an extra $1.3 billion in HST payments has to come from somewhere in order to make Table 3 accurate, but it is worse than that because businesses won't be paying the HST since they receive "input-tax credits" and $1.9 billion worth of those credits currently is counted in PST revenue. In other words, the HST has to raise $3.2 billion more than it collects from the tax base it shares with the PST.

If Table 3 is correct, that $3.2 billion has to come from taxing things like restaurant meals, funerals and other goods and services that currently attract a 5% GST and no PST but will attract a 12% HST. The question is who will pay that extra $3.2 billion?

We know that business won't pay any significant part of the $3.2 billion because any business that is registered to collect HST will claim what it pays as an input credit (all businesses that provide over $30,000 in taxable goods and services). The only entities left to pay the $3.2 billion are BC families, municipalities, charities and non-profit organizations. Table 3 shows that municipalities, charities and non-profits will receive a rebate estimated to total $285 million so that they are not significantly worse off as a result of the HST compared to their financial position with the PST. This means that the HST must raise an extra $2.9 billion from BC families in order for Table 3 to be accurate. That works out to an average of $644 for every person in BC, or an average of about $1,500 per family (of any type), or $2,578 for a family of four. The Ministry of Finance argues that a family of four with a $60,000 income will only pay an additional $278.

Is it possible that higher income families will pay enough to close the gap between expected HST revenue and what has been calculated as the increase in HST costs for the hypothetical family of four? Statistics Canada's Survey of Household Expenditures for 2007 provides average expenditure by quintile. For BC, the averages from lowest to highest quintile are: $25,189; $44,661; $65,059; $86,149; and $142,048. If we assume there are 400,000 families in each quintile, that the amount of increased tax they pay due to the HST is proportional to what is paid by those in second quintile, and that the second quintile pays an extra $278 due to the HST, then the total extra tax paid due to the HST is "only" $900 million, not $2.9 billion. (Using the simpler method of just multiplying average expenditure per household for items taxed by the HST but not by the PST by 7% and by the number of families yields lower estimates for government revenue.) Perhaps the Household Survey isn't adequate for this task since using it produces a revenue estimate that can't be reconciled with more reliable estimates of government's HST revenue.

It appears that there is either a serious error in Table 3, and possibly a $2 billion shortfall in government revenue following the implementation of the HST, or there is a serious error in the calculation of how much additional tax various families will pay as a result of the HST. One way or the other, the apparent discrepancy needs to be resolved so everyone can understand the impact of the HST before it is implemented.

If government has underestimated of the impact of the HST on families, it is not likely to provide adequate relief and to realize the full consequences of its HST decision.