It’s easy for people to have a knee-jerk reaction and support policies like the ecoEnergy Retrofit Initiative or US Production Tax Credits for Wind Energy. After all, we want people to have more efficient houses, and we want more renewable energy, don’t we? Well, yes, we probably do. Environmentalists and fiscal hawks should agree that each dollar spent on programs targeting these changes in our energy production and consumption decisions be spent effectively. Unfortunately, broad subsidies and tax credits do not meet that criterion very often.
These policies are meant to change behavior. The first problem with them is that they often end up paying people who haven’t changed their behavior from what it otherwise would have been – something called the free rider effect by many economists. As an aside, I personally don’t like the term. To me, free riders are defined as those who consume but do not pay for a public good. Take the example of the ecoEnergy Retrofit Initiative which refunds part of the cost of energy-improving renovations. There are really four types of renovations. First, there are those which would have occurred absent the policy. Funding these renovations has no effect, other than to provide a pat on the back for good behaviour, and does nothing to alter the emissions trajectory of the economy as houses are no more energy efficient than they would otherwise have been. Second, there will be renovations which occur earlier than they otherwise would have because of the policy. Funding these has a benefit on our near-term emissions trajectory since it shifts energy-efficiency of the building stock forward in time. Third, some renovators will make more energy-efficiency-improving modifications than they otherwise would have, and fourth, some renovations will happen only because of the subsidy. These types of renovation are where you really want to be, since these are real and lasting changes to the efficiency of the building stock. It would be great if we could target our money to these renovations only, but that is very difficult to do since you can’t easily observe what people would have done, and people are not often willing to report back to the Government that they were free riders, so please don’t propose a survey!
There are some policy design tools which can help. For example, the Canadian ecoEnergy for Renewable Power program contained a claw-back provision on the production subsidy if the funded project earned greater than a threshold rate of return. This is great policy at the industrial level, since those projects with a viable financial outlook without the subsidy will not bother applying to get the grant only to have to repay it in the future. That leaves more money available for those projects which have a marginal financial model and might not otherwise be built. These types of tools are harder to impose at the residential level.
The second reason you might want a policy like a subsidy or a tax credit is to mitigate commodity price risk, for example in an electricity market. Energy projects can be difficult to finance because they can find themselves underwater financially very quickly in commodity price swings – just ask anyone who has been operating a wind farm or a natural gas field in Alberta for the last couple of years. The subsidy or tax credit reduces the effective operating cost of the plant and therefore the investment is bankable under lower expected prices than would otherwise be the case, and the downside risk is muted. Subsidies or tax credits have to be set really high in order to take all of the downside risk out of an investment, and governments often end up paying out the tax credit in good times or in bad. In some research that I have done recently here, we look at the performance of different subsidy and tax credit vehicles in a commodity-risk-exposed industry. We find that the use of a capital grant up front with a claw-back in good times improves leverage ratios and reduces the risk-exposure of the facility at a cost up to 50% lower than the use of subsidies alone.
There is a subtle result in that paper which I think is often overlooked relating to the role of political risk. If you are trying to finance a facility, guaranteed future income matters and banks know that tax credits can disappear. Ontario has attacked this directly with a 20 year fixed price contract under their Feed-in tariff program, but their program is subject to the free-rider problem discussed above. An upfront capital grant allows you to go to the bank with cash-in-hand, not the promise of future tax savings, and so you are more likely to secure financing. This is critical, since facilities which cannot get financed do not get built. If the bank doesn’t believe the tax credit to be stable, then the bank won’t lend you money unless you don’t need the tax credit to be profitable – i.e. unless you are a free rider. The facilities that we want to target are those which are otherwise not bankable and thus would not otherwise be built. Political risk lands you exactly the opposite result. David Roberts (@drgrist) at Grist has a recent post on the use of up-front financing versus deferred tax credits which makes some of the same points..
Long story short – the more checks the government is writing (or the more tax they are not collecting) to reward people for things they would have done anyway, the less money is available to entice people to make the changes we want to see. Further, the more government financing is delivered through short-lived and politically risky measures in the tax code, the less people are going to be able to borrow against them to build facilities which would not otherwise have been built.
If you are one of those people who wants more renewable power, more energy efficient buildings, and more green jobs, you should want to drive changes in behaviour, not simply reward those which would have been there in the first place. If you are a fiscal hawk, you should want your government to use revenue and tax levers to achieve policy goals effectively. No matter which of these groups you belong to, you should be wary of broad tax credits and subsidies.
Andrew,
As you expected, I agree wholeheartedly, so I will comment on a tangential point you raised. I am not a fan of using “free-riders” in this context either. The effect of the subsidy being referred to as “free-riding” is in my opinion more akin to an example of hidden information resulting in an adverse selection problem. Although, I can see where the impetus to use “free-riding” comes from; hidden information prevents the government from practicing “subsidy discrimination”*, so the subsidy is essentially non-exclusive like an public good (still rivalrous** since the source of funds must be finite IMO).
Cheers,
Joel
*Does this term even exist? Like price discrimination except with negative prices?
**It always amazes me how many words we use in economics that spell check does not recognize as words.
Interesting take on the etimology of subsidy free riders. I tend to be pretty careful in my use of public good, because it gets misused so often. I do see where you are going with this though. Thanks!
Andrew,
Interesting post. While some policies, like the ecoEnergy for Renewable Power may be more efficient and generate greater return on the dollars invested, there are still two reasons why the ecoEnergy Retrofit Initiative (to use your example again) is a good policy.
First, although I agree with you that there are preferred groups that we want to take advantage of the ecoEnergy Retrofit Initiative, it would be premature to write off this program when we know that all public goods experience the free rider effect and yet we still maintain these public goods. To that end, the ecoEnergy Retrofit Initiiative needs a better indicator of success before it can be written off. Like you said, it’s very difficult to determine how people fit into the user categories of this program. Appealing to the economist in you, perhaps some rough numbers could be crunched to come up with a dollar value on the success of this program. Using the stated retrofits (I assume the beneficiaries have to declare what exactly they are using the money for), and average reduction in GHG emissions due to said retrofits, an economist could place come up with a value for [GHGs emissions prevented/$]. I’m not necessarily sure this would be the correct route, I am simply saying we need a better indicator of success before we condemn a program like the ecoEnergy Retrofit Initiative.
Secondly, this may not appeal to the economist in you, but their is a psychological component to removing the ecoEnergy Retrofit Initiative. The ecoEnergy programs have to encompass business and the public to project the importance of viewing Canada’s GHG emissions reduction as a collective effort. If these programs focus only on businesses, the public is given another excuse to pass the problem of GHG emissions off to another party, business in this case. If the public is aware that there are ecoEnergy programs targeting them, they will feel that their actions are both necessary and valued, and will be more inclined to reduce their GHG emissions in other direct or indirect ways. This is not a tidy way of evaluating the effectiveness of the program, like using dollars and cents. However, we do have to consider the truly unmeasurable aspects of the program that impact other decisions the beneficiaries of the ecoEnergy Retrofit Initiative will make.
I’m interested to hear your thoughts.
Hi Andrew,
Glad to have you reading and commenting.
I absolutely agree that the metric for policies should be $/ton mitigated, which is why I favor carbon pricing policy approaches in general. You could get to such a metric regarding the ecoEnergy retrofit, but you would be making some assumptions. The key assumption that you cannot get around is the question of when people would otherwise have renovated their houses. Windows, furnaces, etc. do not last forever, and so at some point in the future all of the induced renos really fall into my second and third category – renovations which are shifted forward in time and potentially augmented in their energy efficiency. As such, any figure you get is going to be an approximation. Right now, most of the figures you will see from the government have assumed that none of the renovations would have ever taken place at all, and so have attributed all future energy savings from that house relative to its state before the reno as being caused by the program. Using that method, they get to a figure of about 750kt/yr of GHG reduction, which is cumulative since each year you run the program, you renovate more houses, and the houses you renovated the year before are still in the mix. Over the 25 year lifespan of windows and furnaces and such (optimistic for furnaces, I know) you would get about $21/ton. If you suppose that half of the renos would have happened anyway, you are back to $40/ton for the renos that you caused. Then, if you assume half of the remaining renos were going to happen, just shifted earlier in time, you average cost goes up again. Finally, you have to assume that some of the energy efficiency gain would have been in the reno just due to existing technology – the worst window you can buy today is likely better than those in most houses. My guess would be that the cost of emissions reductions from this program come in at around $100/ton. Ben Dachis from the C.D. Howe Institute has higher numbers here.
Your second point is one with which I have a few issues. I have heard people say time and again that Canadians need to feel like they are part of the solution and that we need a big, collective effort to meet our GHG goals. I think that what we need are fewer policies that waste money so that people who are both environmentally-minded, and fiscally conservative can get behind the effort as well. I am certainly in favour of measuring the full effects of the policy, but if you are going to do that, you also have to take account of the impact of people who see GHG policies as a waste of taxpayer dollars in the name of the environment – I think a lot of people now feel this way about biofuels policy, for example.
There are those who seem to think that environmental policy should be held to a lower standard because it is doing something good, and the GHG reduction game is a long haul. I disagree with the first because I believe in the second. If you get people thinking that green policy just means taking a little bit from every taxpayer to give a nice payout to a few homeowners or to a few companies generating renewable fuel, then you lose the momentum. It’s nice to think about the wave of people getting behind green changes, but that wave can turn against you very easily too – notice that the Globe and Mail, champion of all things environment, is coming out regularly against the solar FIT in Ontario. I tend toward the fiscal conservative, so reading a story like this makes me very angry. I expect there are a lot of people who have the same reaction. There is a real win-win if you get more cost-effective policy, since you get more environmental bang for the buck, and I really believe you get more buy-in.
I know, for example, that Tim Weis of Pembina has a very different take. He sees things as you do – that we need people to buy into this. I contrast it with Jeff Simpon’s view that the movement to climate policy has been hurt equally by governments who over-commit and environmentalists who keep telling people that we will all save money and that it will be easy. I think we need to tell people that it will cost money, convince them of the benefits, and be very careful to spend each dollar that taxpayers give to environmental causes in a way which generates the most impact. If these dollars go further than we expect, great…it was easy. If not, we can still go back and ask for more money because we didn’t waste what we were given before.
My 2 cents.
Andrew
Andrew –
Interesting take on bankability of production subsidies vs. capital. I’m sure you’ve thought of the point I’ll raise, but I wonder if you address it formally in your paper:
Economists tend to prefer that we target our policy intervention as directly as possible at the problem we’re trying to correct. For climate change, the problem is temperature, so a temperature tax might be a good policy in theory (abstracting from time dimension). Obviously, there are lots of practical problems with this, so other potential policies, in order of how directly they address the issue, are an ambient concentrations tax (also practically problematic) and an emissions tax. To address GHG emissions, a policy that promotes renewable energy production is quite distantly connected to the goal of reducing GHG emissions – this is a key reason why most economists (myself and I assume you included) see it as a distant second best policy, only good when emission taxes are not available for political or other reasons. However, a policy that promotes renewable energy capital investment is even further disconnected from the climate goal, since the building of renewable energy capital itself does nothing to alleviate climate problems; production of renewable energy on the other hand, does have climate benefits. A capital subsidy might result in capital being built, but offers no guarantee that it will be used intensively.
There have been empirical studies on this issue. For example, although the institutional context is different, early Indian wind energy promotion programs focused on capital cost (rather than production) subsidies. Under these policies, lots of wind turbines were built, but were run at very low capacity factor. When policies were changed to production subsidies, capacity factors were much higher. (see linkinghub.elsevier.com/retrieve/pii/S0301421599000312 or ideas.repec.org/a/eee/enepol/v28y2000i3p157-168.html).
Thoughts?
Hi Nic,
I absolutely agree with you that a capital subsidy is a distant second or perhaps third best policy to a carbon price when we are interested in limiting ambient carbon concentrations. No question we should be taxing the marginal contribution to carbon concentration, i.e. emissions. Within the realm of renewable energy production, I think there are trade-offs on both types of grants with respect to the cost-effectiveness of funding. We tend to concentrate on the usual one, which you have written about a lot, in which subsidies reward the production of energy from renewable sources whether or not that energy is incremental and, in the case of biofuels, often without considering whether there is any emissions displacement impact. What we were digging at was another way to view the free-rider problem at the investment stage, where we saw the potential for a selection effect early on which would exacerbate free-ridership.
I think the point you raise about capital utilization will vary by type of energy source, but no question it’s a valid concern. Some sources, like wind power, most of the levelized cost of energy is in the capital cost, and so once you have the turbine up, there is no incentive to limit its use, but there would perhaps be incentives to build them in more marginal locations. The ethanol plant we focused on would definitely still have that problem, since any decrease in the market price for ethanol could make it optimal to idle the plant. No question that the point at which the subsidy hits will be important to the overall effect. I wish we would have been a little more clear on this in the paper, but it’s now at final proof stage.
Thanks for commenting!
Andrew
Hi Andrew,
Thank you for continuous tackling of climate, energy, and economics. It’s been great reading your posts. This one even got the neurons firing enough to tempt me into commenting.
While for some supporting retrofit policies are a knee-jerk reaction, others have been targeting these efficiency retrofits consciously, and sometimes the knee-jerk is for a reason. In 2010 the US’ National Performance Council identified 126 regional residential retrofit programs, with 86% offering incentives and overall 90% were sponsored by private or municipal utilities. A question is begs, why are all these utilities dipping into their margins to support programs? There must be some long-term financial benefit. My point, Canada isn’t the first place to be looking into these subsidies. Having a national Canadian strategy could be more effective than above, and would help to promote best practices, decrease short-term growth in energy demand, and categorize our existing residential inventory.
I agree with you. Effective use of funds is required on all government expenditures (based on GHG mitigated/$). Though most sources I’ve seen point to energy efficiency as being on the ‘quick’ end of the payback scale. If these are expensive, then what programs are cheaper at this scale with this measure? We should definitely do that first. Period. Then come back to this. In addition, energy efficient homes are the first essential step towards introducing any renewable energy retrofits. Without these upgrades the payback periods for renewable energy would be currently untenable. A long-term strategy could really help Canada!
You mentioned four main categories for the program, and put a lot of emphasis on the last point (Reno’s only because of the subsidy). While I’d agree that yes one wants this category to be the majority of users, but the other three categories are not inherently bad either. Renovators that will make more efficiency upgrades due to the program should be highlighted as well, since these upgrades might never have been made in the first place. These retrofits also require an initial energy audit that would shift behavior into the higher payback regimes for the building residents, freeing up generation capacity and lowering emissions.
I think at the same time homeowners should be required to put an Energy Rating on their residence. To properly supply energy to a growing market we need to define that market, for centralized (or decentralized) generators and for the tenants/homeowners. This should be mandated for whenever a house is renovated, sold, or rented. We don’t have a system akin to this now, so we’re not able to target subsidies more effectively. It’s less messy than a survey too!
On the point of targeting, these programs should really be focused on Alberta and Ontario, the two provinces with >200Mt GHG/year. I know that’s not very Canadian like, but if we want to have the highest GHG mitigated/$, it’s these two provinces that need help. Considering Alberta’s thermal-coal capacity too, this province needs the most help!
I’d be interested in your comments, as that’s just my 2 cents.
Cheers,
Ryan
Thanks for commenting Ryan. You raise some very good questions.
Many of the programs you speak about are in regulated utilities, so the costs of these programs would be deemed as eligible for a rate of return. An example I know of is Gaz Metro in Quebec. In their case, the economic incentives should drive them to grow their market as much as possible, since they earn a regulated rate of return on all of their assets. To provide the correct incentives for demand side management, the regulatory board has indemnified them from the costs of the programs and from any displaced demand…in other words, the programs are paid for, and they are generally profit-indifferent or close to it as to whether the program succeeds or not. These programs are naturally popular, since they are a small cost to all tax/ratepayers, and can provide a big benefit to some. It’s also important to point out that some are much better designed than others.
Home renos are among the cheapest GHG reductions, particularly in provinces that still use a lot of fossil-fuel-based heating, so what you really need is a policy that induces new renos without rewarding too many that would have happened anyway. It’s really hard to do with incentives. To me, the only way you get cost-effective GHG policy is with pricing of emissions, since that will make retrofits pay for people where the value of the GHG emissions reductions are sufficient. You then put some complementary programs in place for rental housing and low income people to make sure you are not putting a disproportionate penalty on some.
I agree 100% with the energy rating. This would be a very small price to pay to drive competition in the resale and new home market on energy use. As Mike Holmes is fond of saying, no one looks at the bones of a house, they look at the countertops. Having a homeowner or the utility provide information on energy use in the house would be a close second. It would cost less than $200 at the time of a house sale to have a rating done. The home Energuide rating does this for new houses, but it’s hard to do a good analysis of a reno’d house without having energy use and behavior data. You could come close though. Sounds like an engineering problem!
I disagree with targeting to Alberta and Ontario. We need a national approach that targets low cost emissions reductions, not emissions in high emissions locations. The more low-cost emissions reductions we can generate, if we insist on playing a target game, the better. Again, that is most easily done through a carobn price, but I guess I am beating a dead horse here.
Thanks again for reading!
Andrew
[…] On retrofit subsidies, tax credits and other expensive policies via @andrew_leach (You should probably just subscribe to his blog’s RSS cause these will keep showing up.) […]
Back to one of Joel’s initial comments referring subsidy free riders:
“Does this term even exist? Like price discrimination except with negative prices?”
I agree with you both that free rider is misused with regards to subsidy abuse.
Is there also a term for a person who would have purchased these vehicles without the ecoEnergy Retrofit initiative (ie. a subsidy)?
I am currently having this same discussion in my office…albeit with regards to a different service.
Thanks everyone for the interesting comments and thread.
– R
That’s how people use the term free-rider with respect to subsidy. I guess you could also use an infra-marginal consumer…i.e. one for whom the subsidy decreases the price they have to pay (net) but who would have purchased the product at the original market price.
Cheers,
Andrew