His default assumption seems predicated on the idea that our unfunded liabilities are vastly underestimated (or we're being lied to about their size). He thinks they're around $86 Trillion - which, yeah, okay, that's way beyond the debt ceiling.
What's clear is that he's taking medicare and social security liabilities into account - what's NOT clear is (a) how far into the future he's looking in order to get that $86T estimate, or (b) what methods he's using to assess the size of the liabilities in future years.
In other words, we can estimate what our outlays for social security in 2030 are going to look like today, but are they accurate, are they based on worst-case scenarios in terms of inflation, revenue, GDP, employment, population growth? Are they based on best-case scenarios? And so on - and, more to the point, based on how close we can accurately assess those future potential deficits (because they are potential deficits, they aren't real until they occur), should we actually implement policy today? If there's a confidence interval for that $86T of, say, $150T, well, maybe we shouldn't implement any policy today.
Here's the odd thing about right-wing policy; the same people who fervently deny the overwhelming scientific evidence of climate change (that our actions today will have aggregated, substantially detrimental and irreversible effects in the decades to come) seem to be the same people who apply the absolute *inverse* of this logic to fuzzy estimates of future financial outlays in those same decades to come. In short, it doesn't matter if we're irreversibly
ing the climate up, but we better do something today to sure that we don't overspend in 30 years.
What should we do today? Let's raise the social security age!
Why? So we don't have to cut social security benefits in the future!