So much for the notion that Trump is just riding a wave and has nothing to do with it. If hate crimes went up by that much in counties where he made speeches, it's tough to argue otherwise. It's nigh impossible to have correlation that extreme and with that much geographic specificity without causation.
Like with any such claim, you should look into it before drawing conclusions. There are lies, damn lies, and statistics.
Unfortunately, the referenced paper does not appear in Google, and clicking the "our research" hyperlink in the article leads to a page stating that they will present the paper at some polisci conference next month.
But it's possible to at least vet the claim a bit. The article states that they used the ADL's extremist heatmap data to create their statistical analysis.
You can find that here:
https://www.adl.org/education-and-resources/resource-knowledge-base/adl-heat-map
And you only need to interact with it for a little bit to see that there is the potential for serious problems, and it's probably not good data to test their hypothesis in the first place. It doesn't track hate crimes, for example, despite all the mention of hate crimes in their op-ed. Only the top three categories would appear to potentially fall under that label. Of which there were 69, and then one should also subtract the 16 attributed to Islamic or left wing ideology. And not all of those remaining even sound like hate crimes. For example:
"Paul Ridgeway, an associate of white supremacist prison gangs, was fatally shot after firing on a Contra Costa Sheriff's deputy who was attempting to arrest him for outstanding warrants."
But even being overly generous and saying that there are 50 or so significant events that may be valid to test against the effect of Trump rallies, that's a very low sample size and I doubt you'd be able to say anything with confidence from that.
But I'm guessing that they used more data than that. There were 3,787 total incidents in 2017 (the latest available) EDIT: (Actually I had 2017 and 2018 selected). But when you just do a cursory scroll the events, you can see that a really sizeable portion are fairly benign.
For example:
"Patriot Front, an alt right group, posted flyers and stickers that read: "Keep America American," "Not stolen conquered," "Money does not rule you," and "Not here not ever."
and
"Identity Evropa, an alt right group, distributed flyers that read: "Protect your heritage." They also posted stickers featuring their group logo."
In fact, stickers put up by these two groups look like they account for about a third and at least a quarter of the reported incidents. So, in the event you have some of their members showing up at Trump rallies and putting up those stickers, that alone could theoretically explain the results of the study. But in that case you can hardly say that Trump's rallies are driving hate crimes (ignoring for the moment that these aren't actually hate crimes) in those communities, since it would actually be the confounding variable of people following him around and putting up those sticks.
It's possible that the researchers used a methodology that accounted for some of these potential issues, and maybe we can read more about that next month, but... I'm not hopeful. Almost certainly looks like a nothing burger.