The merging criteria was for "area with most similar ratio of X to Y", where X is the primary variable and numerator and Y is the secondary variable and denominator. So we have popinc/pop, which is just income - we want MCDs with similar income to be grouped together.

Just a heads-up that 30k isn't really a meaningful number in the grand scheme of things. We just used it here to sort of classify "small-medium" to "medium-large" MCDs

...that are generally between 10k-50k population

...because they make up a tract, and our mortality/temperature data is at the tract level. (Let me know if you need more details)