Energy Probe

(April 15, 2010) Matt Horne of the Pembina Institute suggests that “careful design” will address the many real issues associated with the carbon tax about which Harvey Enchin deftly wrote a week earlier on the OpEd page of the Vancouver Sun.  Matt says: “Sweden provides and example of how governments can make carbon taxes work.”

Sweden was among the first nations to experience the serious phenomenon of the carbon tax-generated hole in government revenues. Sweden’s carbon tax shift translated directly into industrial job losses and capital investment flight. In the early years, carbon tax revenues consistently fell short of budget targets due to industrial business flight.

Between 1990 and 2008 (pre-global recession), jobs in Sweden’s goods producing sectors-agriculture, forestry, manufacturing, energy industries and construction-fell over 25%, compared to a 10.5% increase in Canada. Over that 18-year period, total employment across all sectors in Sweden grew only 2.4%, compared to 30.5% in Canada.  100% of the new jobs created in Sweden since the implementation of its carbon tax have been in the public sector.

De-industrialization explains 100% of the reduction in greenhouse gas emissions (GHGs) realized by Sweden since 1990.  In fact, given the reduction in industrial output and jobs, the national GHG reductions realized by Sweden is surprisingly small.

Sweden’s electricity and petroleum industries currently discharge 8% more GHGs within Sweden’s boundaries and 30% more globally than they did in 1990. This partly reflects that fact that in the mid-1990s, in an attempt to stem industrial job losses, Sweden ruled that all energy consumed “for the production of mineral oils, carbon fuels and petroleum coke”; in the production, transmission and distribution of electricity; by airplanes, ships and trains; and by “energy intensive business” (defined as a business for which energy costs account for 3% or more of total operating costs) is up to 100% exempt from all Swedish energy taxes not just carbon taxes.

Of course, while introducing these energy tax exemptions might have stemmed job losses in the targeted sectors, it just deepened the carbon tax hole in Sweden’s government revenue forecasts. And Sweden-home to 20% more petroleum product refining capacity than Alberta-now depends more on petroleum product exports to shore up its national economy than ever before.

When Matt Horne points us to Sweden’s careful carbon tax design as a model, is he advocating for exemptions for all of the large BC industrial consumers who would be energy tax exempt under Swedish law?

Sweden has tried to make up the carbon tax-related government revenue shortfall by disproportionately loading the new tax burden on small households and increasing payroll taxes, mandatory health and social insurance premiums, a path BC Premier Campbell has already started down.

Today, a small Swedish home that consumes less than1,000 kilowatt-hours (kWh) of electricity per year pays CAD$0.347/kWh, including taxes. But the large home that consumes over 15,000 kWh per year pays only CAD$0.176/kWh.  In other words, the large residential consumer gets almost 2,000 kWhs of service for the same price the small household pays for 1,000 kWhs. Large Swedish industrial customers who do not qualify for the tax exemption pay only CAD$0.097/kWh.

In presenting the Swedish carbon tax design as a model to British Columbians, is Matt recommending that our government disproportionately load the new tax burden on the smallest and lowest income-earning households, as Sweden has done?

Since hospitals pay carbon, excise and sales taxes on their energy consumption while power producers, oil refineries, airlines and other energy intensive Swedish businesses do not, the direct relationship between the carbon tax shift and increasing health care premiums is obvious.

Other labour tax increases are not directly related to carbon tax impacts.

But the Swedes—like many other governments—have found the only place they can mine for new government revenues, of late, has been the payroll tax accounts. Taxes on labour now combine to account for 33% of all Swedish government revenues (compared to 15% of all Canadian government revenues). Sweden’s carbon tax accounts for less than 3% of total Swedish government revenues.

Sweden’s very high payroll taxes discourage workers and dampen companies’ willingness to hire. Today, Sweden’s unemployment rate is 9.0%. This does not include the 5 – 8% of working age Swedes who are officially listed as “employed” but who do not work and collect permanent “sick leave”. In 2006, according to McKinsey and Co.,  “sick leave” payments to Swedes staying home from work accounted for 16% of all Swedish federal government expenditures.

Government spending accounts for 55.5% of Sweden’s GDP.

And this is what Matt and the Pembina Institute describes as “careful” tax system design?

Aldyen Donnelly, April 15, 2010

(Apr. 05, 2010) US regulators appear to be planning/budgeting for a high level of non-compliance with the proposed new CAFE standard. This planned non-compliance will generate significant new tax revenues for the US Treasury. Canadian manufacturers will take a hit, as they’ll account for more than 15% of the new US Treasury revenues from fines.

The attached “Regulatory Impact Analysis” (“RIA”, mandatory part of US rule-making) identifies two categories of regulatory cost impact:

1. the cost of compliance with the new regulations, and

2. the cost of fines for non-compliance. The attached RIA forecasts that the incremental US technology costs—to be born by US new car purchasers and not including non-compliance costs paid by manufacturers—will be $17.6 billion in 2016.  The RIA acknowledges that the cost of fines paid will also likely be passed through to new car purchasers, but the regulatory impact analysts elect not to include the cost of fines in their estimates of the impact of the cost of the regulation per vehicle sold.

The logic built into the RIA modelling exercise is explained as follows:

And:

I found the remaining tables in the RIA report, given the explanation on page 296 under “VII. Cost Impacts” rather confusing. So I phoned a NHTS (the US regulator) contact to ask for the proper interpretation.

The bottom line is that this cost analysis starts by setting an “Adjusted Baseline” for each car and light truck manufacturers’ vehicle sales fleet average fuel efficiency and tailpipe emissions performance. This Adjusted Baseline assumes all cars sold in 2011 will all be in full compliance with the existing US CAFE standard for 2011.

Then, to estimate the cost impact of the proposed new CAFE standards for 2012 through 2016, the regulator forecasts the costs each manufacturer will bear to increase fleet average fuel efficiency and reduce tailpipe emissions to the CAFE targets for 2012 through 2016 from this Adjusted Baseline for 2011.

There are two significant loopholes in the regulator’s analysis:

1. Many of the manufacturers, I am told, will fail to comply with the 2011 CAFE standard—the Adjusted Baseline. They will pay fines to comply with the 2011 standard (or “Get To the Adjusted Baseline” in Tables VII-1a and 1b), instead of delivering vehicles to the market that physically comply with the regulated model year 2011 performance standards. In Tables VII-1a and VII-1b (pages 297 and 298) the US regulator estimates the continuing costs to the manufacturer—costs, which I am told, will be largely incurred  in the form of non-compliance fines, not new technology costs—to cover the difference between their actual forecast 2011 model year sales fleet average energy efficiency and emissions performance and the Adjusted Baseline.

2. Then, in tables VII-2a to 2o and tables VII-3a to 3o the US regulator shows us only the technology costs of compliance with the new 2012 through 2016 CAFE standards.  The RIA report acknowledges that a number of manufacturers will pay additional fines for non-compliance, over and above the technology costs shown in these tables. The RIA report also discloses that the regulator has estimated what the US government’s revenues from incremental fines will be. But the RIA does not disclose those estimates, arguing that any fines paid are “transfer payments” which will translate into consumer price increases, but which do not need to be reported because they will have no net social or GDP impact.

I am told that, for purposes of forecasting actual new car emissions, the analysis also treats payment of the fine for non-compliance as equal to achieving physical emission reductions. But there is no commitment on the part of government to spend the revenues collected in the form of non-compliance penalties on other measures that will directly cut vehicle emissions. So this procedure results in a forecast that could significantly overstate the actual emission reductions, overstate fuel cost savings and understate the consumer price increases that will arise from the implementation of this new CAFE regulation in the US.

My argument is that the US regulator’s decision not to disclose the extent to which the regulator assumes the manufacturers will simply pay penalties and fail to comply with the regulated performance standards is too cute by miles. A legitimate and transparent analysis would discretely disclose:

• 100% of the estimated incremental costs per vehicle that will be passed on to customers to cover manufacturers’ investments in new technology;
• 100% of the estimated incremental costs per vehicle that will be passed on to customers to cover manufacturers’ non-compliance penalty payments, which costs are effectively a new carbon tax that generates new revenues for the Federal Treasury;

• the regulator’s estimate of actual new vehicle sales emission reductions, given their estimates of manufacturers’ non-compliance rates.

Obviously, in the absence of this full disclosure, any government would be highly motivated to maximize government revenues—and minimize compliance with the stated environmental objectives, by:

• regulating impractical(y) high new CAFE standards, and

• setting non-compliance penalties below the estimated sector average cost of physical compliance with those high performance standards.

If and when governments pursue this strategy, they are doing little more than disguising a new tax as an environmental measure—a measure that will inevitably fail to achieve its stated environmental objectives. Among other things, it will extend on-road vehicle stock turnover rates, damage the economic health of all vehicle manufacturers and put the most innovative of manufacturers at greatest risk.

What Does the Existing US Regulator’s RIA Tell Us, if Anything, About US CAFE for 2012 through 2016?

This document does show us the regulator’s estimate of the cost that the manufacturers will pay on a continuing basis to cover the difference between the actual performance of the vehicles they plan to sell in 2011 and the Adjusted Baseline. My NHTS contact suggested that most if not all of this cost will likely be covered in the form of non-compliance penalties.

For now, we can add cost estimates appearing in Tables VII-1a and VII-b in the attached report to the technology cost estimates in the VII-2 and VII-3 tables, then divide the VII-1 tables’ values into those sums, to estimate a minimum apparent non-compliance fine share of the RIA estimated average new car price increase. This procedure is far from perfect, and what we really require is full disclosure of the regulator’s estimates of fines versus their estimates of new technology costs. But in the absence of full disclosure, let’s use the available data and verbal input from the NHTS contact to generate a preliminary signal.

When we complete this little analysis, it appears that the US regulator could be planning that non-compliance fines will represent, at a minimum, the following %s of the reported incremental new single passenger vehicle cost increases from 2012 through 2016, by manufacturer:

I also show the implied value of non-compliance payments to the US regulator, by model year and manufacturer, for single passenger vehicles. At the bottom of the table, I add in the total estimated costs of non-compliance penalties for the light truck fleet. This less-than-perfect analysis suggests that the US Treasury could collect as much as $19 billion in non-compliance penalties from vehicle manufacturers who supply product to the US market over the 5-year period of 2012 through 2016.

If non-compliance penalties prove to be only 50% of the cost “to get to the Adjusted Baseline”, we are still talking about $8 to $10 billion in new taxes on US car sales to be collected by the US Treasury form 2012 through 2016.

Please note that the regulations stipulate that non-compliance penalties will be collected at the point of sale. So, under the regulatory strategy that is currently proposed, the government of the US will collect the lion’s share of Canadian manufacturers’ non-compliance penalties, not the government of Canada.

Also note that historical market experience suggests that most if not all of the non-compliance penalties paid to the US Treasury by Canadian vehicle exporters will not be passed through to US consumers, but will be born as a combination of extra price increases for Canadian car buyers and reductions in export sales margins for Canadian car manufacturing plants.

What Can Minister Prentice Add as an Amendment to Canada’s Version of the NA CAFE Rule to Mitigate Financial Leakage from the Canadian Auto Manufacturers to the US Treasury?

Minister Prentice can maintain his commitment to implement a CAFE regulation in Canada that is “harmonized” with US CAFE, and address the risk of financial leakage described above by:

• adding an alternative CAFE compliance option for Canadian manufacturers that enables, but does not oblige, them to cover any new car sales fleet compliance shortfall by acquiring and scrapping older, high-emitting used Canadian vehicles, and

• ensuring that the Canadian CAFE penalty for compliance failure is high enough  to ensure that this measure is truly an environmental measure and not a new tax revenue measure in disguise.

And, of course, Canadian auto manufacturers can only survive a high relative non-compliance penalty if the government of Canada commits that any non-compliance penalty revenues would be used to partially finance, say, a 1-year ACCA depreciation rate for new Canadian clean vehicle purchases by Canadian corporate fleet operators.

Please be assured that I do not present the need for these amendments to the US CAFE regulation for its Canadian application out of distrust of the intentions of Minister Prentice or the current government of Canada. I am concerned, however, that a literal adoption of the US CAFE regulation exposes Canadian auto makers and governments to a potential significant wealth transfer to the US, which could, in turn, dampen private sector investment in Canada’s auto sector.

As in the softwood lumber context, Canada’s final CAFE standard must ensure that any non-compliance penalties associated with Canadian export sales are invested in Canada—not outside Canada—either in the form of emission-reducing car scrapping activities or direct non-compliance penalties.

Aldyen Donnelly, April 05, 2010