The Greatness of Public Infrastructure

Today, Statistics Canada released an excellent paper concluding that the rate of return on investment in Canadian public capital is around 17%. This paper builds on another Statistics Canada paper from a few years ago that Marc cited a few days ago on this blog.

If anything, the paper’s text somewhat understates its mathematical findings. The author’s hypothesis was that “the long-term government bond rate could reasonably be employed as the rate of return on public capital.” In his words, “the paper concludes that the long-run government bond rate could be used as a conservative estimate for the rate of return for public infrastructure.” The opening line on Statistics Canada’s website is that public infrastructure “provided a rate of return to public capital at least as high as the government long-term bond yield over the period from 1961 to 2005.”

Put another way, the 17% figure suggests a rate of return appreciably higher than long-term government bond yields. Since the long-term economic benefits of infrastructure significantly exceed the financing costs, Canadian governments should be undertaking more public investment. The short-term economic stimulus of such investment would be an added bonus.

UPDATE (April 15): Eric Beauchesne has written a story on today’s Statistics Canada paper and FCM release (which Toby mentioned below):

Infrastructure pays off, StatsCan says

Eric Beauchesne

Canwest News Service

Tuesday, April 15, 2008

OTTAWA – The rate of return to businesses and individuals of government investment in infrastructure, such as roads, bridges and sewers, is at least as great as the government’s cost of raising the funds for that investment, a new Statistics Canada study suggests.

“Public infrastructure, the roads and water and sewer systems that comprise the foundation of Canada’s economy, provided a rate of return to public capital at least as high as the government long-term bond yield over the period from 1961 to 2005,” according to a summary of the study, which estimated that return “centres” on an annual average of 17 per cent.

The findings support the case for more such investment, a labour economist argued.

“Since the long-term economic benefits of infrastructure significantly exceed the financing costs, Canadian governments should be undertaking more public investment,” said Erin Weir, an economist with the United Steelworkers.

But there would be an “added bonus” to such investment at a time when economic growth is slowing, as is occurring now, Weir added.

“Although we also need interest rate cuts, there is legitimate doubt about how quickly lower rates from the Bank of Canada will translate into lower borrowing rates for consumers and businesses,” Weir said. “In this context of financial uncertainty, direct public investment becomes even more important as a potential source of economic stimulus.

“Infrastructure investment would be a particularly effective type of stimulus because it would necessarily create employment in Canada,” he added, noting the construction work would have to be done here and that many of the materials used in construction, such as concrete, would be produced here as well because it is impractical to import them.

Further, such investment would help reduce what the Federation of Canadian Municipalities has said is a $123-billion infrastructure deficit that has left Canada’s water-treatment facilities, roads and public infrastructure is on the verge of collapse.

The federation released poll results Tuesday that it said “shows Canadians overwhelmingly want the federal government to provide greater financial support to municipal governments.”

“More than 90 per cent say the federal government should help municipal governments deal with infrastructure issues, a view shared across the country, including in Quebec and Alberta,” the federation said.

. . .


  • “This paper uses the Cobb-Douglas function to generate parameter estimates of MFP(t)F(K, L,G).”


  • As Erin points out, the summary in the Daily does understate the results in the paper, which left me wondering about possible spinning by Statscan, which would be a concern.

    The paper does indeed appear to rigorously adjust for possible uncertainties and variances in results from using different estimattion methods. Still, it confirms an estimated rate of return of 17% for investments in public infrastructure — higher than the long-run private rate of return on capital and the long-term government bond rate (as noted on page 42 of the paper). But a 95% confidence interval of plus/minus 12% spans government bond rates and the rate of return of capital, hence the qualifier in the Statscan Daily summary.

    This report comes out on the same day that the Federation of Canadian Municipalities released a major poll showing that 64% of Canadians polled actually think the federal government should raise the GST back up to 6% and use these funds to invest in local public infrastructure.

    See page 46 of poll results linked to:

  • These are trickly calculations to make. After the steam engine, global productivity increased 2% annually, upward to maybe 4% annual now, assuming optimistic environmental accounting. In the future, this should be even higher. So when you pay down the debt, you are giving maybe the top 20% tax-bracket of citizens more money to spend (since it debt payment savings go to tax-cuts now) on a cutting-edge 2030 cancer treatment in the USA, vs. paying debt interest to maybe the top 5% tax-bracket. For paying down debt, you are actually getting a slice of the world’s future productivity gains. It doesn’t amount to much until compounding interest takes hold after a few decades…basically, the world will innovate at a faster rate than Canada, so it makes sense to keep the coffers full when considering extremely long-term purchases; to have a dividend to constantly purchase the world’s techno-junk.

    I wouldn’t think the 17% infrastructure ROI applies to incremental infrastructure investments in general. Certainly our $1 trillion economy is not the basketcase a $123 billion deficit would suggest. I’d suggest the needed investments are things like getting fire response wait times down to 4 mins from 5 mins, rather than building a firehall where none exists. I haven’t read the paper, but would be curious to know what the incremental ROIs are. The way it is presented, suggesting all $500 billion annual public expenditures should be increased 25%, doesn’t make sense when considering firehall construction yet is a significant underinvestment in considering daycare.
    We’re gonna look silly foraging for ants and squirrels 40 years from now, if we now build a water treatment plant addition meant to last 60 years, and global warming kills two years of crop harvests and kills half of us.

Leave a Reply

Your email address will not be published. Required fields are marked *