Why Aren’t Average Incomes Rising More Rapidly?

Central bankers must continually grapple with the disconnect between economic theory and practice.

They design intricate models that try to predict how today’s policy decisions will affect future economic outcomes, but there are myriad variables involved and things often do not rarely turn out as planned. (This has led to the quip: economists exist to make the weather forecasters look good.)

Today, our policy makers can’t fully explain why average incomes aren’t rising at a much faster rate. Over the past several years, both Canada and the U.S. have seen strong job creation coincide with average income growth of about 2% and that’s barely higher than our respective rates of inflation. For a point of reference, Fed Chair Yellen has said that in the current environment, average wage growth of 4% to 5% would be a sign that the U.S. labour market has turned the corner.

Classic economic theory says that strong job creation should lead to tighter labour market conditions as the demand for labour grows in relation to its supply. When labour becomes scarcer, employers should be compelled to increase compensation in order to attract new workers. Then, as the cost of labour rises, companies would be expected to increase prices to maintain their profit margins, thereby fuelling increased inflation across the broader economy.

Central bankers are charged with maintaining overall price stability and they don’t want to wait until this process plays out fully before raising rates. Labour costs are a key component in most forms of economic output. If they rise quickly, it will create higher-than-expected inflation that will then require more severe policy actions to rein it in.

That’s why their instinct is to tighten monetary policy pre-emptively.

True to form, when the Bank of Canada (BoC) and the U.S. Federal Reserve started their current rate-hike cycles, both cited strengthening labour-market conditions as key factors in their decisions. The BoC’s policy pivot came in a speech by Deputy Governor Carolyn Wilkens where she emphasized that it takes time for monetary-policy changes to impact economic momentum and that the Bank must “anticipate the road ahead”.

But what if central bank forecasts are wrong and incomes don’t start rising more quickly? Are there other factors influencing average incomes that central bank models have difficulty accounting for?

This is the key question for anyone keeping an eye on mortgage rates today because they should move in the same direction as average incomes over time, and at a similar speed. I have read many theories about why average wages aren’t rising as expected and a recent article, Wages vs. Jobs by Gary Shilling, offers particularly valuable insight. His observations focus on the U.S. labour market, but much of what he says also applies to Canada.

Here is a summary of Shilling’s key points with my comments included as well:

• Globalization is keeping labour costs in check – Globalization has shifted manufacturing and other forms of production from developed countries to developing countries over the past three decades. This has led to a steady decline in manufacturing jobs in both the U.S. and Canada, and now an increasing number of services are being offshored as well. This long-term trend is deflationary for both prices and average incomes.
• The output gap is difficult to quantify – Policy makers focus on when the output gap will close (as a reminder, the output gap measures the gap between our economy’s actual output and its maximum potential output), but an economy’s maximum potential output isn’t a fixed number that can be reliably measured. Instead, it can adapt and expand if growth is slow and predictable (as it is now). For example, many factories have upgraded their machines but the old ones are often still there, sitting idle because they are less profitable to use. If increased demand for a factory’s outputs leads to higher prices, those older machines can be turned back on again. The same is true for potential employees who can be drawn back into the labour force if compensation becomes more attractive. In these ways, our economy’s maximum potential output can actually be quite flexible.
• The unemployment rate is a poor gauge of labour-market conditions – Policy makers focus on the unemployment rate as a key labour-market indicator but it is a poor gauge because it only counts people who are either working or actively looking for work (and not people who have given up looking for a job or won’t work for the wages they are offered). That means the unemployment rate can fall if large numbers of people withdraw from the labour force, which is hardly a healthy signal. A good portion of the drop in U.S. and Canadian unemployment rates over the past several years can be attributed to this phenomenon. The U.S. participation rate, which measures the percentage of working-age citizens who are either employed or who are actively looking for work, has fallen from 67.3% in early 2000 to 62.9% today. Canada’s participation rate has fallen from 67.6% in early 2008 to 65.7% now. So while today’s U.S. (4.3%) and Canadian (6.3%) unemployment rates appear to signal tightening labour-market conditions, our participation rates imply that there is still plenty of untapped labour available for our economies to absorb if needed.
• Corporate profit increases are coming from cost cutting – The average operating earnings per share of S&P 500 companies just hit an all-time high, but those profits have come primarily from cost cutting, not from increased sales. As such, employee incomes aren’t rising to reflect the recent increase in corporate spoils.
• New jobs have tended to be lower paying – Since 2007 the U.S. economy has added three million leisure and hospitality jobs, which pay an average of $15.43/hr, compared to 900,000 new manufacturing jobs, which pay an average of $26.51/hr. Manufacturing employees also tend to work more hours, so they earn an average of $1,081.61/wk, compared to leisure and hospitality employees, who earn an average of $402.72/wk. More broadly, Shilling observes that even in higher-paying sectors, well-paid baby boomers who retire are often replaced with new hires who come in at the lower end of the pay range, and more so than in past recessions, unemployed workers who gain full-time work and part-time workers who move to full-time have both tended to do so “at below-median wages”.
• Declining union membership – Union membership in the private sector has declined steadily from 25% of the total workforce in 1973 to 6.4% by 2016. That has a depressing effect on average incomes because U.S. union jobs pay an average of 19% more, and “over 50% more when heath care, retirement and other benefits are included”.

There are plenty of other theories about why incomes aren’t responding to the recent string of strong headline job growth numbers.

Statisticians in both Canada and the U.S. use “plugs” in their employment models to account for variable factors like seasonality, birth/death rates, weather and so on. Economist David Rosenberg recently estimated that “more than half” of the U.S. economy’s estimated non-farm payroll job creation this year “has come from the magic wand of the BLS [Bureau of Labor Statistics] birth-death model”. In other words, we should also be questioning whether U.S. and Canadian employment growth is actually as strong as is being estimated in the monthly employment reports because much of the data therein are comprised of guesses that are based on past trends which may no longer be relevant.

In many ways, the employment data we receive each month are an evolving theory about what is happening with the labour market, whereas the wage and income data offer more quantifiable proof in the pudding.

Thus far, the BoC and the U.S. Fed have pre-emptively raised their policy rates in the belief that higher labour costs are around the corner and that they must tighten monetary-policy now to stay in front of soon-to-increase inflationary forces. But their assessments are made with imperfect information and they are trying to forecast against an ever-shifting backdrop where influencing factors are difficult to quantify.

At a time when central-bank policy language contains the consensus warning that materially higher rates are inexorable, it is worth remembering Yogi Berra’s adage that “it’s tough to make predictions, especially about the future.”

Five-year Government of Canada (GoC) bond yields rose two basis points last week, closing at 1.49% on Friday. Five-year fixed-rate mortgages are still available at rates as low as 2.64%, and at rates as low as 2.79% for low-ratio buyers, depending on the size of their down payment and the purchase price of the property. Meanwhile, borrowers who are looking to refinance can find five-year fixed rates in the 3.04% to 3.09% range.

Five-year variable-rate mortgage discounts remain largely unchanged and are still available at rates as low as prime minus 0.90% (2.05% today) for high-ratio buyers, and at rates as low as prime minus 0.75% (2.20% today) for low-ratio buyers, again depending on the size of their down payment and the purchase price of the property. Borrowers who are looking to refinance should be able to find five-year variable rates around the prime minus 0.45% to 0.70% range, which works out to between 2.25% and 2.50% using today’s prime rate of 2.95%.

The Bottom Line: The BoC and the U.S. Fed have pre-emptively raised their policy rates, in part because they believe that tightening labour-market conditions will soon increase inflationary pressures on both sides of the 49^th parallel. But that outcome is far from certain, and if average incomes continue to rise at a level that barely outpaces inflation for the reasons outlined above, warnings about higher rates on the horizon will once again prove premature.