Conventional wisdom, for what it’s worth, holds that once the nominating conventions are over and both sides get their bounces, most people’s votes are well along to being solidly locked down. The state of the economy matters a lot at this point as well. Typically, that would matter less in an election like this one, where the incumbent isn’t running. But this time may be different, as a key part of Trump’s case is that the economy is in terrible shape, and Hillary Clinton, like President Obama, believes that while serious structural problems exist, there are ongoing, favorable trends that can potentially be built upon.

I’d like to briefly get into one of those trends: wages. Wage growth has accelerated of late, and not just for those at the top. Meanwhile, low inflation has meant that these faster nominal gains have turned into significant real gains.

I don’t want to oversell the case; anyone who has followed this and most who’ve drawn a paycheck know that there’s a ton of catch-up to do. But I’ve long argued that the tightening job market will give working people more of the most important thing they generally lack these days: bargaining power. We’re not yet at full employment, but we’re getting closer, and when that happens, employers must bid wages up to get and keep the workers they need. Let me show you what I mean.

The first figure shows yearly changes in nominal median weekly earnings for full-time workers. It’s very noisy data so I plotted a smooth trend through it.

The important facts about this series are:

— It accelerated smartly in the late 1990s, the last time we were at truly full employment.

— It was pretty flat in the 2000s, a period of uniquely weak job growth.

— It got whacked by the last, big recession, but it’s now clearly accelerating again, though not yet at a pace similar to that of the late 1990s.

But of course what really matters to working people is the pace of real wage growth, i.e., accounting for inflation. If your pay goes up 2 percent but inflation does as well, you’re not better off in terms of buying power.

Here’s where the Trump economic take is really wrong, at least in these national data. (There’s a relevant regional story noted below.) The table below breaks real median earnings growth down into nominal growth, a positive, and inflation, a negative. That is, real earnings growth equals nominal earnings growth minus inflation — and remember, these are medians, not averages, so they’re more representative of middle-class paychecks.

The table, which uses data for the first half of the years shown, reveals that real median earnings actually fell from 2012-14, thanks to tepid nominal wage growth and faster price growth. Since then, however, wages have accelerated and inflation has decelerated, leading to solid median earnings gains in real terms.

BTW, the inflation hawks out there are using their talons to scratch their heads: Why didn’t faster nominal wage growth bleed into prices? Because, as I’ve shown in other posts, they haven’t done so for a while. The slowdown in inflation was clearly driven by tanking energy prices, with some economic slack mixed into the picture as well.

The final figure, from researchers at Goldman Sachs, tells a particularly important part of the story, bringing in both employment growth, which has also been solid of late, and a variety of different wage levels. It’s a tricky figure, but it accounts for each component of earnings growth: nominal gains, employment, hours worked and inflation. The punchline is the white, diamond dot in each bar, showing the rate of real earnings growth, accounting for all its contributing sources.

That dot shows that the group making the biggest gains is the lowest wage workers, who have benefited from particularly fast wage growth. Why? Well, one reason is the many increases in the minimum wage which have taken place across the country. And remember, while he flip-flops on the issue, Trump has argued minimum wages are too high, and the Republican platform, and more importantly, Republican members of Congress, oppose increasing the federal wage floor.

Okay, some caveats to all the above. We know that real wages have stagnated long-term for many in the workforce, particularly those beset by trade competition from low-wage countries, along with millions of low-wage workers stuck in places where policymakers, at the behest of business lobbies, ignore the minimum wage. A few quarters of solid, real wage growth don’t wipe out that record.

We also know that this recovery has been geographically diverse, and a more complete wage analysis, one I hope to get to soon, must examine regional differences, along with those of race and gender. I suspect it will show significant swaths of the country that have been left behind.

But these national data are telling an important story, one that is really happening and one that flatly contradicts Trump’s hyper-negative rhetoric. Most importantly from a substantive, economic sense, it is a story consistent with the benefits of full employment. If we want this story to continue such that the recovery has a chance to reach many more of those who draw a paycheck, we’re going to have to keep it going.

That means we’ve all got a job to do. The Federal Reserve must pursue smart, patient monetary policy; the federal government must do the same with fiscal policy, meaning no negative shocks. And the electorate must elect the president most likely to maintain and build on the positive trends shown above.

Okay, everybody. Get to work!