Since 2009, the Canadian Labour Congress has demanded an expansion of the Canadian Pension Plan (CPP). With the election of the Liberal Party—which campaigned on enhancing the CPP—to a majority government in 2015, that expansion may finally be on its way.

The argument is in its favor: The C$282.6 billion (US$223 billion) already under management by the Canada Pension Plan Investment Board (CPPIB) enables the investment team to access the highest-returning asset classes at low prices, while also retaining significant expertise in-house.

Growing the fund could only make it that much more efficient. Right?

“The answer is unequivocally no,” said Philip Cross, a freelance researcher and member of the Business Cycle Dating Committee at the CD Howe Institute.

Cross examined the costs of major Canadian public pension funds with fellow researcher Joel Emes as part of a study for the Fraser Institute, a leading Canadian think tank. The report included the CPP, plus the five largest plans based in Ontario: The Ontario Public Service Employees Union Pension Trust (OPTrust), the Ontario Municipal Employees Retirement System, the Ontario Teachers’ Pension Plan, the Ontario Pension Board and the Healthcare of Ontario Pension Plan (HOOPP).

“There’s an untested assumption among many people that when it comes to big public pensions, the larger they are the more efficient they are—that there’s economies of scale,” Cross told CIO. “This actually hasn’t been tested.”

“To suggest there are no economies of scale in the pension business is dead wrong.”Cross and Emes tested the theory by collecting cost data from annual reports published by each of the six pension funds and comparing them as a percentage of plan assets. CPP—the largest by more than C$100 billion—was also the most expensive, with an expense ratio of 1.07%. But overall, the report found no systematic relationship between costs and size among these six pension plans. OPTrust, for example, was the second most expensive fund to run despite being the smallest in the study with C$18.4 billion under management.

HOOPP and its C$63.9 billion in assets had the lowest expense ratio—but it was in the middle of the pack in terms of fund size.

Source: The Fraser’s Institute’s “Comparing the Costs of the Canada Pension Plan with Public Pension Plans in Ontario“.

“There’s no evidence there are economies of scale,” Cross said. “The head of HOOPP [CIO Jim Keohane]—he himself has said, ‘I don’t see any reason why a pension plan would become any more efficient after about C$75 billion of assets.’ And the data seems to bear out exactly that—we didn’t find any economies of scale beyond C$75 billion.”

But with a sample size of just six funds, the Fraser Institute study is far from “unequivocal,” argued Tom Scheibelhut, managing principal at CEM Benchmarking.

While Scheibelhut acknowledged that economies of scale are logarithmic—as funds get larger, more assets are required to shave off each additional basis point of expenses—that doesn’t mean they cease to exist entirely.

“To suggest there are no economies of scale in the pension business—it’s dead wrong,” Scheibelhut said. “Not only do bigger funds have a cost advantage, they do better on average than smaller funds.”

The results of the Fraser Institute study, Scheibelhut said, are misleading for two reasons: the small sample size and the reliance on annual reports. The costs disclosed in annual reports, he argued, are often inconsistent and incomplete, making it difficult to draw accurate comparisons between funds.

Karen Danylak, a spokesperson for OPTrust, said that she “can’t speak to the report’s methodology,” which calculated the fund’s expense ratio as 1.02%. However, the pension plan’s internal costs for 2015 were just 43 basis points, she added. External costs, meanwhile, are taken into account as part of the investment strategy, with returns reported net of fees.

“Although we always work to ensure that the costs our organization incurs are reasonable, we also note that looking at costs alone as a measure for a pension plan ignores the results a given strategy has achieved—in OPTrust’s case, a fully funded pension plan,” Danylak said.

At CEM, a data firm co-founded by KPA Advisory’s Keith Ambachtsheer, Scheibelhut studies a standardized database of more than 500 pension organizations around the world, representing over $8 trillion in assets. Based on these observations, he said fund size is one of the top predictors of high returns net of fees.

“Partly it’s due to style—for example [large plans are] less likely to use funds-of-funds, which have on average done dreadfully,” Scheibelhut continued. “The other reason why bigger funds do better is they start to be able to go into more expensive asset classes in a cost-effective way—things like real estate and private equity.”

Another primary factor in cost savings, according to CEM? Internal management—something large funds with large staffs are more capable of implementing.

Even if the expense ratios given by the Fraser Institute study are based on inconsistent data, Scheibelhut said it would make sense that HOOPP—a fund that manages investments entirely in-house—would be among the cheapest to run. CPPIB, meanwhile, would likely pay more by employing external managers.

“HOOPP has shown that if it is your goal to keep costs as low as possible, it can be done,” said Cross. “It is possible to keep costs much lower if that is what management wants to achieve.”

Keeping costs as low as possible is a noble goal—so long as it is balanced out by an effort to achieve the highest possible returns.

“The answer isn’t just low cost,” Scehibelhut added. “The lowest cost strategy is to be 100% indexed. You have to play the game to win.”

As for whether higher returns will justify more expensive strategies, Cross argued it’s too soon to tell.

“[Asset owners have] been investing in a unique environment of ultra-low interest rates and very high asset prices,” he said. “Everybody’s investment plan has looked pretty good over the last 10 years. We’ll see in a difficult investing environment whether that holds up.”

Despite Cross’ skepticism, OPTrust remains confident in its investment program, which has resulted in a net funding surplus since 2009.

“Looking at the costs of these plans actually demonstrates their efficiency and cost-effectiveness in delivering a retirement benefit when compared to typical retail vehicles used for retirement savings, such as mutual funds,” Danylak said. “A sophisticated and diversified investment portfolio that uses alternative asset classes has been one of the key drivers of [OPTrust’s] funding results.”

Related: In Canada Pension Plan Fee Controversy, Critics Go for Jugular