Deutsche Bank senior U.S. economist Brett Ryan was among those who predicted a “reverse square root symbol” shaped recovery months ago.

“This speaks to what we’ve been saying for a long time, which is that this is not going to be a straight line, lock-step recovery, up and up and up. It’s going to have some bumps along the road,” Ryan tells Fortune. Indeed, Ryan told Fortune in the beginning of June that he thought a reverse square root symbol recovery was likely, “whereby the tail is below the start and gradually upward sloping,” as he described it.

True, the data has been improving. Unemployment fell from a heady 14.7% in April and 13.3% in May to 11.1% in June. Consumer spending came in much-improved in May, jumping to 8.2% (off of two months of massive drops). Ryan points out the sales were “boosted temporarily” by one-time stimulus checks.

Weekly jobless claims, meanwhile, have fallen for the past 14 consecutive weeks, but have remained stubbornly above a whopping 1 million per week for the last 16 consecutive weeks.

Now it’s the slowing pace of recovery (notable in data like weekly unemployment claims) that is worrisome to some analysts. “As the COVID-19 outbreak has recently intensified in some states, hopes for an accelerated, sustained, and successful reopening of the economy have hit roadblocks,” Bankrate.com senior economic analyst Mark Hamrick wrote Thursday. “This raises concern about the economy’s rebound.”

And when examining those reports, Deutsche’s Ryan instead looked into “exactly why perhaps [unemployment and May consumer spending] were better, and the conclusion remains that there are definitely still risks and it seems like those risks are now playing out,” he said.

What might hamper growth

One dark cloud hanging over the overall recovery: Some of the big states now reimposing restrictions amid rising case counts are among the top GDP contributors in the country: Texas, Florida, and California make up roughly one third of the nation’s GDP, according to the U.S. Bureau of Economic Analysis.

What’s more, Bank of America recently wrote in a report that while the initial bounce looked like that hopeful V, “By the time the growth in the virus peaks, we assume roughly half of the country will be impacted. With half of the country slowly opening and half slowly closing, the economy could flatten out overall,” economists at the bank wrote.

That doesn’t mean a hit to activity in some of these states will veer the economy back into negative growth and output, Ryan believes, and spiking cases in some states isn’t necessarily a death-knell for the nation’s recovery at large.

That’s because, as economists at Goldman Sachs wrote in a note Thursday, “While the risks have increased, we see several reasons why renewed broad lockdowns may not be needed,” including that early data suggests states in the Sun Belt (like Arizona and Texas) are increasing mask usage, and that “the bar for broad lockdowns appears significantly higher” than it was in March. But the firm also points out that “risks clearly remain, and further lockdowns cannot be ruled out.”

Even if the bars remained open and the restaurants welcomed guests in the coming weeks, those like Ryan note “you’re seeing consumers kind of self-regulating their behavior even absent state regulations.” That has been a key concern: That even if states do give consumers the green light to go forth and spend, activity may be diminished “on their own volition because of the fears of the virus even before states officially issue shutdown orders,” he says.

Economists like Ryan cite two other factors that might inhibit the fast rebound we’ve seen in data recently: One, “it’s becoming pretty clear that the massive fiscal support that was provided including the [stimulus] checks supported retail sales in May—that’s going to peter out over the next few months,” Ryan says. The second risk, Ryan says, is if the $600 per week enhanced unemployment, set to expire at the end of July, doesn’t get an extension. “That’s going to be a problem,” he says.

Another interesting tidbit? Ryan points out that searches for “unemployment benefits” are on the rise again on Google Trends, and he’s watching jobless claims and unemployment reports closely (specifically for temporary versus permanent layoffs). The latest unemployment report shows that while jobs were recovered in June, the number of permanent layoffs increased to 2.9 million.

Consumer spending has clearly rebounded off of its depths in the crisis (down -6.6% in March and -12.6% April) as it surged in May. But Ryan cautions that the likely positive numbers moving forward may not be boosting some of the worst-hit areas.

“On a sequential basis, it will look really good,” he says. Deutsche Bank is projecting a 10.5% annualized increase in real consumer spending in the 3rd quarter, Ryan notes, but “What you have to remember, though, is of that 10.5%, the health care contribution is going to be the biggest,” he says. “While health care is going to bounce quicker, it’s going to take a long time for restaurants and hotels to come back, and unfortunately that’s a significant part of the labor market and where a lot of damage was done.”

Still, the unprecedented uncertainty of the virus means anyone’s guess is, well, anyone’s guess. But Ryan remarks: “If you had thought that you would have faster growth in the back half of the year, you’re probably thinking, ‘Okay, it’s probably a little slower than I thought.'”