On the last point, whether rich and poor share an ethnic/tribal identity, I don’t think Dean McHenry was assuming that at all. There are cases where it’s true that the elites share an identity with the poor, and there are cases where the elites are ethnic minorities in those regions.

One of the main points of McHenry’s talk (and huge part of the paper, he said) is that the socioeconomic analysis doesn’t show any consistent pattern, whereas the rhetoric does.

I think this also ties into what I was saying earlier about “natives” and “settlers”; in a state like Chattisgarh, the “wealthy enclaves” are almost certainly “settlers” (or descendants of the same) — whereas the separate state movements would typically have been spearheaded by those articulating “tribal”/”native” concerns. Thus to the extent that the discussion assumed that the “rich” and “poor” shared the same cultural/communal group identity, I guess I would disagree. [That does not mean I necessarily disagree with the notion that the process of new-state formation is itself open to capture by the elites. And I do not also mean to suggest that "elite" will invariably map neatly onto the "native/aboriginal" and "settler" dichotomy: for instance, affirmative action and quotas means that in all these states the government is a hugely important source of employment for "tribals" and other marginalized groups].

On states & “secession”: I am not sure I would agree that the wealthier portions of states tend to be in favor of secession: I do not believe that Chattisgarh, Jharkhand, and Uttaranchal (to name the three prominent examples from this decade) were the most prosperous parts of Madhya Pradesh, Bihar, and Uttar Pradesh, respectively. I would say the same about the ongoing Vidarbha movement in Maharashtra.

The wider point you make might nevertheless stand: i.e. each of the three I have mentioned are natural resource-rich, but the “natives” (for want of a better word Jharkhand and Chattisgarh feature very high %ages of “aboriginal” groups, loosely referred to as “tribals” or “adivasis” in the popular media, though “settlers” are likely a majority in Jharkand and certainly parts of Chattisgarh too) are dirt poor, and it cannot be doubted that “outsiders” — in the form of corporations, government and private mining operations — have been cornering the benefits of these natural resources for decades now, and the “new state” movements have grown out of that.

That’s not to suggest that I can’t think of instances where the more prosperous parts have sought to secede (think Tulunadu’s sporadic attempts to secede from Karnataka), just that I was under the impression the general rule was the opposite: the would be Telengana or Vidarbha states (currently part of Andhra Pradesh and Maharashtra, respectively), and Jharkhand and Chattisgarh, seem to me to be poorer than the larger units they are/were part of. Do you recall what examples were specifically used to illustrate this point?