Look anywhere and you'll see it noted that Donald Trump has the lowest job approval rating of any president at this stage in his term. It's true. Go down Trump's approval rating in recent surveys: CBS News 39 percent; Gallup 43; Rasmussen 52; Reuters 45; Economist 48; Quinnipiac 38, and on and on. Trump's current approval rating in the RealClearPolitics average of polls is 44.3 percent approve, 50.3 percent disapprove. Previous presidents at this point in their terms were definitely higher.

There's no denying Trump's numbers, but it is reasonable to question whether they say what they seem to say.

According to the RealClearPolitics average, Trump's personal approval rating was 58.5 percent disapprove, 37.5 percent approve on the day he was elected president. In exit polls that day, 60 percent said they had an unfavorable opinion of Trump, versus 38 percent favorable. A candidate with a disapproval rating around 60 managed to win Florida, Ohio, Pennsylvania, Michigan, Wisconsin, North Carolina, and other key states on his way to winning 306 electoral votes and the presidency.

That alone should make anyone considering Trump's current polls wonder what is going on. Consider a couple of possibilities.

The first is that there might still be a Trump Effect even though Trump is in the White House — that is, there are still supporters who are afraid to say they support the president. The second is that even though many Americans say they disapprove of Trump, they're not dead-set against him because they still believe there's a real chance he will produce on his campaign promises.

On the first, Republican pollster Neil Newhouse has done a series of four focus groups in recent weeks, in St. Louis, Cleveland, Macomb County, Mich., and Portland, Ore. Each group had ten people. Two of the groups were split evenly between Clinton and Trump voters, while two were all Trump voters.

"From the focus groups I've done, for many Trump voters, it is still socially unacceptable to admit their support for the guy," Newhouse told me Thursday. "When Trump voters are in the room with Clinton voters, Clinton voters had no reluctance to talk about their vote for Clinton and how much they dislike Trump. And in that environment, Trump voters looked around like they were wondering, Should I even say anything or keep my mouth shut?"

Newhouse gave me some quotes from the sessions. "When I told people at work, all I got was a bunch of criticism," a woman who supported Trump said at the Macomb County group.

"Being really pro-Trump in Portland, I would feel uncomfortable," said a supporter in Oregon.

"I don't feel educated enough to defend Trump," said a woman in St. Louis.

"Today is the first time I have breathed the word that I voted for him," said another woman in that same session.

From that, Newhouse concluded, "There's no reason to believe that the reluctance [to identify as a Trump supporter] that we saw last fall is not still there." Indeed, it might be more pronounced, given the intensity of the public discussion about Trump.

Newhouse doesn't believe there's a Trump Effect that would add double-digit points to the president's job approval, but he does think it might account for two or three points of support. Newhouse recently tweeted several reasons to take Trump's current poll numbers "with a grain of salt."

First, "2016 polling underestimated Trump support in key states. Nothing's changed." Second, "Focus groups indicate Trump voters don't feel 'safe' voicing their support. Shy, stealth, or camouflaged — can't poll them, but they exist." Third, "GOPers are still squarely behind Trump, with approval in mid-80s, higher than Ronald Reagan with GOPers in 1981. Trump's base solid." And finally, "We're just starting month two of the Trump administration following the most divisive and polarizing presidential election in memory." To that, Newhouse added a hashtag: #chill.

Newhouse is trying to develop poll questions that might shed more light on the nature of Trump's support and opposition. In recent polls, when he asked a standard approve-disapprove question, he then asked everyone who did not approve — that includes the disapprovers and those who said they weren't sure — whether they agreed or disagreed with this statement: "Even though I may not approve of the job he is doing, I like some of the actions that President Trump is taking." The positive responses added five or six points to Trump's rating.

Maybe that's skewing things too much; asking the question that way might have well added points to any president's rating. But Newhouse is trying to work his way through a problem. "I'm trying to figure out a question I can ask that would give these camouflaged Trump voters permission to say, yes, I support the guy," he told me.

Another Republican pollster, David Winston, stressed that the current polls probably reflect unformed opinion on the part of some Americans rather than solid opposition.

"People are making an assessment, and they're not making it quickly," Winston told me. "They're going to see what he's going to do over a period of time. My sense is we're just watching people as they think through how they're going to assess things."

Winston believes a significant number of people who do not tell pollsters they approve of the job Trump is doing — whether they outright disapprove or don't know — are eminently gettable for Trump. "He's got the opportunity because people are open," Winston said. "But that doesn't mean they're going to flip their opinion prior to anything happening."

In other words: Trump has to produce.

Winston also noted that last November, when exit pollsters asked voters which candidate quality mattered most to them, "can bring change" won with 39 percent — nearly two-to-one over any other single attribute. Among those voters, Trump demolished Clinton, 82 percent to 14 percent. The people who wanted change in November still want it now.

In last year's general election, the national horserace poll numbers were reasonably accurate; the RCP average of polls on the eve of the election had Clinton ahead by 3.2 percentage points, and she won the popular vote by 2.1. But some key state polls significantly understated Trump's appeal, and he won the Electoral College by showing amazing strength — for a Republican — in the Rust Belt.

At the time, many — actually most — political observers underestimated Trump's potential. How could a candidate who trailed in key state polls, and who had such high disapproval numbers, actually win? It's safe to say that nearly every establishment political observer got the answer to that wrong. There were also some who, to their credit, said Trump had a better chance than the conventional wisdom.

Now, the conventional wisdom is that Trump is a walking disaster at the polls. Yes, the numbers, read the conventional way, are bad for Trump. But the lesson of 2016 was that there might be something about Trump that defies conventional measurements. And now, the critics who say President Trump is beyond recovery might be making the same mistake they made just last year.