The old Ben Franklin quote says just two things are certain: death and taxes. But a modern update might include a third: inflation.

Many who lived through the early 1980s were traumatized by the double-digit inflation that gnawed away at retirement funds or put mortgages out of reach by driving up interest rates. Inflation has been comparatively benign in recent years, hovering around 1.6% over the past year. But experts still tell us to work inflation assumptions into long-term financial plans, as even low inflation can raise the cost of living dramatically in the decades ahead.

And what we think about inflation affects the real world. If people expect inflation to be high, they demand higher yields on interest-paying investments like bonds. Then higher borrowing costs drive lenders to raise rates for mortgages and other loans.

But not every individual, or even the experts, will have the same inflation expectation. What happens when they disagree? Does it matter if they disagree by only a little? What if they disagree by a lot?

New research shows that disagreement feeds back into the real world much the way high and low expectations do. In periods of high disagreement, when the range of inflation projections is especially broad, bond yields tend to rise, and they go up more than when disagreement is low. This turns out to be true even if the median expectation is the same for the two periods.

The study is described in the paper “Disagreement about Inflation and the Yield Curve” by Wharton finance professor Philipp Illeditsch, Paul Ehling of the BI Norwegian Business School, Michael Gallmeyer of the University of Virginia, and Christian Heyerdahl-Larsen of the London Business School. The findings could help players like the Federal Reserve refine their inflation-control techniques, and could even be useful in setting investment strategy, making buying decisions or choosing between fixed- and variable-rate mortgages.

Though the research does not address the causes for inflation disagreement, Illeditsch says that personal experience emotions interfere with cold analysis of inflation the same way they do with other financial matters like home values or stock prices. Those thinking about inflation have imperfect and sometimes conflicting information, so forecasting is never a sure thing. Even professionals can disagree, though not usually by as much as ordinary people.

To track the range of expectations, the researchers used two surveys: The University of Michigan’s Surveys of Consumers (MSC) and the Survey of Professional Forecasters (SPF) by the Federal Reserve Bank of Philadelphia. Disagreements can be wide.

Emotions Interfere

While news reports generally focus on the median finding, the surveys also report expectations of respondents at the 25th and 75th percentiles in the range of answers. In December 2015, for example, the MSC expectations ranged from 0.9% to 4.6%, and the SPF from 1.87% to 2.25%, Illeditsch and his colleagues report. The study also found that an increase in disagreement by professionals, a gap between low and high widening by 10 basis points or 0.1 percentage points, raised the yield on two-year Treasury notes by 36 basis points. An increase in disagreement by consumers of 1 percentage point, raised the two-year yield by 94 basis points.

“When the range of inflation projections is especially broad, bond yields tend to rise, and they go up more than when disagreement is low.”

Why does more disagreement cause rates to rise?

Illeditsch illustrates with two hypothetical individuals. One expects inflation to be high, the other low. Each expects to profit from his view. The one with high expectations will lock up money in an investment he thinks will pay a high yield, while the other investor will bail out of a holding he expects to become stingy, hoping to do better in the future by avoiding poor performance.

Because they expect to be wealthier as a result of these moves, each will increase current spending by borrowing, just as someone counting on a bonus or trimming a big expense might increase credit card purchases before the additional money arrives. The extra demand for credit drives borrowing rates up, even though one of the two players will turn out to be wrong. “They both think they will be richer in the future so they both want to consume more today,” Illeditsch says. “In order to consume more today we have to borrow.”

Data on trading in bonds, bond futures and inflation swaps confirm that investors are more active when inflation disagreements are broader, he says. High disagreements also cause greater volatility in interest rates and consumer spending, the research found. History shows that forecasts from the two surveys are pretty good predictors of inflation, Illeditsch says, but what matters in this examination is how differences in expectations drive behavior and influence rates.

“The study’s findings can help regulators like central bankers better interpret expectation data, and refine their forecasting and monetary policy.”

“Both surveys are actually used by the Fed when they make decisions” about monetary policy, he says. If the Fed believes inflation will rise to unacceptable levels, it raises short-term interest rates to discourage borrowing that fuels spending and inflation. If inflation is expected to be low, the Fed cuts interest rates to encourage borrowing that stimulates spending. High inflation hurts the economy by reducing the dollar’s buying power. Low inflation and deflation are bad because they discourage people from buying things like homes that they fear will fall in value. Shrinking incomes from deflation are very damaging to borrowers, making it harder and harder to make loan payments and in extreme cases leaving them with homes worth less than they owe.

Help for Regulators

Illeditsch says the study’s findings can help regulators like central bankers better interpret expectation data, and refine their forecasting and monetary policy. By being more transparent, for example, the Federal Reserve would reduce disagreement in the marketplace, helping to reduce interest rates. And when the goal is to raise interest rates, the Fed could encourage disagreement by being less transparent. He adds: “The benefits of raising rates through move transparency have to be weighed against the cost of an increase in speculation and the resulting redistribution of wealth.”

Understanding the role of disagreement could help investors in placing bets, he adds. If high disagreement suggested that yields would go up, for example, an investor might plan on higher income in the future from interest-bearing holdings, while low disagreement might have the opposite effect.

Similarly, borrowers might steer clear of adjustable-rate mortgages if they see high disagreement signaling those loans could charge more in the future. With low disagreement, an adjustable-rate mortgage might seem like a better choice than a fixed-rate loan, because the rate would start low and not be expected to rise substantially.