Who is to say the economy can’t avoid another recession many experts expect to arrive between 2019 and 2021?

In May of last year, Goldman Sachs said the current expansion had a two-thirds chance of being the longest in U.S. history. Furthermore, other countries like England and Ireland have enjoyed far longer expansions dating from the early 1990s to the Great Recession.

But measuring this cycle against previous ones in terms of months may be a mistake, according to Jim McDonald of Northern Trust. The downturn and destruction of wealth during the Great Recession were so severe that a meaningful recovery was delayed for several years after 2009.

Many American homes remained under water until recently, and workers have not received hefty raises. Contractions in economic activity are typically caused by excesses in the system and, aside from Bitcoin, those are hard to find. It’s possible that a weak recovery could last a long time.

For this scenario to materialize, a lot of things would have to go right—or viewed another way, not get too much better.

Jim Paulsen, chief investment strategist at the Leuthold Group, notes that at 2.4%, the 10-year Treasury is within 25 basis points of its three-year high. The markets love 2% wage inflation, but they might be thrilled if that figure hit 3% in 2018. Oil prices also are approaching a three-year high. Finally, unemployment could soon approach a 50-year low while the S&P 500’s price-to-sales ratio nears the all-time record set in 2000.

If these and other indicators start “to breach important milestones,” Paulsen wonders whether they will also alter Wall Street’s complacent mind-set and cause a market correction at the very least.

That’s uncertain, but demographic developments over the next five years are crystal clear, and they will provide a sobering influence. Between now and 2023, peak earners and spenders aged 45 to 54 years old will decline from 13% to 12% of the population, according to David Rosenberg, the chief economist with Gluskin Sheff.

Meanwhile, the population over 70, which experiences the sharpest declines in income and spending, will rise from 10.5% to 12.5%. “Looking at data going back to 1920, it will be the first time there are more people over 70 than peak breadwinners aged 45 to 54,” Rosenberg wrote in early December. “This is likely to have a more powerful impact on spending, savings and inflation than any government attempt at tax reform.”

It may not reflect positively on CEOs or the long-term sustainability of their companies, but many are making capital allocation decisions based on the demographics of their investor base, which explains their emphasis on dividends and buybacks. “Since the bull market began in the spring of 2009, the S&P 500 has advanced 285%, but total return has surged 363%, with dividends accounting for 20% of the total,” Rosenberg says.

This cycle has been characterized by low growth, low interest rates and low inflation. Maybe it just could stay lower for longer.

Or maybe not. Call it whatever you want but no economic condition is permanent. What if Secular Stagnation, the New Normal or the New Mediocre is about to end and we are about to enter a new phase.

In a recent interview, Mohamed El-Erian declared that within two years the lower-for-longer world would come to an end and the global economy would either transition into a period high, inclusive growth or a "recession with renewed instability."

El-Erian noted that Brexit had driven Europe to focus on the problems that has caused it to experience a sclerotic economy for nearly 40 years. Under new leadership, France of all places is addressing the issues that have calcified its labor market—namely, that when the government makes it painful to fire workers, employers become afraid to hire anyone.

In America, tax reform should should be able to sustain and possibly accelerate the expansion for 2018. But a question investors rarely ask but need to keep in mind is: What could go right?

The economy is running low on workers but robotics and artificial intelligence are picking up the slack. The media is obsessed with how new technologies like driverless cars could displace tens of millions of workers. But what if these new innovations also create millions of new jobs. It's happened again and again in the past. Will this time really be that different?