As Jordan Brunner has explained, the new NSPM-4 memorandum reorganizing the National Security Council has far more in it than the demotion of Steve Bannon. Most notably, from my perspective, the new structure subordinates the Homeland Security Advisor (Tom Bossert) to the National Security Advisor (Lt. Gen. McMaster). This is a return to the subordinate structure that President Obama had in place (where Lisa Monaco was nominally subordinate to Susan Rice (albeit acting with a fair degree of independence)). By contrast, the "separate structures" organization that had previously been in place (for the first 75 days or so of the Trump Administration) was equivalent to how the Bush Administration treated the two advisors.
We can begin by granting that the actual architecture matters less than the relationship between the two principals. If Bossert and McMaster can work cooperatively, the formal structure of subordination matters little. Similarly, when two or three principals clash substantively, no amount of organizational mandate can solve the problem.
But, leaving that caveat aside, the reorganization raises an interesting practical question—which structure is the more appropriate? The answer depends on how you conceive the enterprise of homeland security.
To be sure, homeland security has some significant overlap with national security. Its counter-terrorism mission and its border security mission both resonate with national security purpose. Indeed, the impetus for forming the Department of Homeland Security (and conceptualizing the idea of defending the homeland) lies in a clear national security defeat for the nation in the 9/11 attacks. Likewise, national security is often a component of homeland security—one has a strong sense that in the 2000s and early 2010s American national security engagement in Iraq and Afghanistan had the collateral positive effect (whatever else one may think of those endeavors) of restraining international terrorist activity in the US homeland. As such, there is clearly some cross-cutting aspects of the two enterprises that require close coordination and that argue for a unified structure. Only that way can trade-offs of various strategies be effectively analyzed and comprehensive choices made.
On the other hand, homeland security is much broader than national security—in much the same way, I think, that economic security is also partially about national security but also (predominantly) about domestic economic consumption, wages, taxes, education, and health. Because the non-national security aspects of the economy predominate, nobody would think it sensible to subsume the President's national economic council under the national security council.
The same is arguably true about homeland security. Many aspects of the enterprises—things like emergency preparedness and response and immigration adjudication and enforcement are less about national security than they are about principally domestic matters of the economy and government. Others, like aviation and transit safety, are mostly about counter-terrorism, but in practice focused more on domestic law enforcement matters (compare, for example, the number of domestic offenses arising from TSA inspections with the number of international terrorist incidents). Likewise, infrastructure security is a mixed issue and cybersecurity, for example, is as much about crime and theft as it is about espionage and national interests. And to the extent the terrorist threat is morphing to become one of domestic radicalization, homeland security, again, seems to resonate more with domestic law enforcement than with national security and overseas military or foreign policy operations.
On balance then, the homeland security enterprise sits in an uncomfortable middle ground—neither fish nor fowl; neither wholly national security nor wholly domestic economic and policy issues. To some degree, that realization reemphasizes the unusual (some say unholy) nature of the homeland security construct. And perhaps, if we were to do it all over again we might not start down that path in the first instance. But DHS and the homeland security enterprise are with us now and will remain with us for the foreseeable future.
In that light, and given the mixed nature of homeland security, my own assessment is that the new restructuring is an error. I think homeland security is sufficiently distinct from national security (and has sufficiently great domestic impacts) that the venue for tradeoffs and strategic considerations of homeland policy needs to be distinct from national security. Under the current structure legitimate domestic economic policy concerns risk being subsumed beneath security mandates. It is unsurprising that a security-minded President like President Trump has chosen this organizational structure, but it is, in the end, unfortunate.
To be sure, the matter is not free of debate. Indeed, very many wise analysts have cogently made the contrary argument. But for my money homeland security does not walk like the duck of national security—it's more of a cross-breed and needs its own independent organizational architecture to thrive.