This is a very interesting question, and the linked study does not address it, only "net migration" to "other parts of the country." IANAD[1] but I suspect the

This is a very interesting question, and the linked study does not address it, only "net migration" to "other parts of the country." IANAD[1] but I suspect the census estimate data is not actually capable of addressing your specific question - how could you tell who was coming from where? But as jeffbee suggested there might be a way to tease it out. The study mentions remote work being part of the reason for the shift; this also confounds "city-to-suburb" versus "city-to-Midwest" migration since it's harder to cross-check against employment statistics - again, without more detailed surveys. [1] Demographer :)