Logical Syllogism Showing That Corbyn Proves the Left Is Correct

I. Premise
Zack Beauchamp clearly established this as the proper test:

“The reason we are losing ground to the right today is because the message of what socialism is and what it can achieve in people’s daily lives has been steadily diluted,” Corbyn said in a March 2016 speech. “Unless progressive parties and movements break with that failed economic and political establishment it is the siren voices of the populist far-right that will fill the gap.”

Corbyn’s year-plus of Labour leadership has been something of a test case for this theory. So far, it has failed utterly.

II. Fact
In the 2017 election, Jeremy Corbyn’s party stunned the world by picking up 30 seats, and increasing its vote share by 9.5 points.

III. Proof

  1. If Corbyn does well, then that proves the theory that leftist economic populism can reverse the trend of losing ground to the right (Beauchamp 2017).
  2. Corbyn did well (BBC 2017).
  3. Therefore, leftist economic populism can reverse the trend of losing ground to the right.

Why would rich disabled parents spoof their kid’s disability?

Terrence McCoy has a long piece in the Washington Post about multi-generational disability. Or, more accurately, a piece about one family McCoy spent a few days with. The only parts of the piece that try to quantify multi-generational disability make very little sense.

Households with multiple family members on disability

From the article:

As the number of working-age Americans receiving disability rose from 7.7 million in 1996 to 13 million in 2015, so did the number of households with multiple family members on disability, climbing from an estimated 525,000 in 2000 to an estimated 850,000 in 2015, according to a Post analysis of census data. The analysis is probably an undercount.

The first problem is that “multiple family members” is not the same thing as “family members from different generations.” Two spouses would count. Two siblings would count. And so on.

The second problem is that you can be “on disability” without being disabled if you have a disabled parent. When a parent is receiving SSDI benefits, each child they have will receive a benefit equal to 50 percent of their parent’s benefit, subject to an overall family benefit limit. Children do not need to be disabled to receive this benefit. They receive it simply because their parent is disabled.

The third problem is that the household counts for 2000 and 2015 should be divided by the number of households in the country to convey the real magnitude of the change. In 2000, there were 104.7 million households, meaning that 0.5 percent of households had multiple family members on disability. In 2015, there were 124.6 million households, meaning that 0.7 percent fell into this category. That’s the crisis, apparently.

Rich disabled kids

The other data element of the piece uses the American Community Survey to determine how often disabled children have a disabled parent. This data does not show how often children and adults in the same household both receive disability benefits, only how often children and adults in the same household tell survey workers that they have one of the five serious disabilities tracked by the Census.

I used the same ACS file to produce a similar calculation as the one made in the piece. For my calculation, I kept things really simple. I isolated two types of households: 1) those with disabled children in them but no disabled adults, and 2) those with disabled children and disabled adults in them. Then I divided the number of households described by (1) and (2) by the total number of households in the file.

Screen Shot 2017-06-05 at 10.24.18 PM

Here is the same calculation but broken up by income level, again in a similar way as the piece in question did:

Screen Shot 2017-06-05 at 10.26.27 PM

Here is the same graph, but I divide the red bars by the blue bars to get a ratio:

Screen Shot 2017-06-05 at 10.32.26 PM

These figures do not substantially differ in their trend from what is presented in the piece, though they are more detailed and use all households for their denominator rather than whatever denominator the piece decided to use.

The data element in the piece is presented without much commentary, but the patterns shown in the above graphs are actually at odds with the welfare cheating story that the piece is trying to tell.

The only disability benefit children can get for being disabled themselves is from the Supplemental Security Income (SSI) program. That program is means-tested, meaning that only children in lower-income families can receive it. For those with incomes over the quite low limits, claiming your child is disabled is of no use to you as far as welfare income goes. Yet, child disability is related to adult disability up and down the entire income distribution. In fact, the odds of a disabled child also having a disabled parent gets larger as you move up the income scale.

The welfare-cheat model suggested by the piece is supposed to be that you have an adult who is on disability; they then realize that if they claim their kid is disabled, they can get more benefits; so they do that. But this model would predict the opposite of what you see in the last graph. The likelihood that a “disabled” child is in the same household as a disabled adult should go down as you climb the income ladder because there is no way to cheat welfare money when your household income puts you above the SSI eligibility thresholds. Instead, the likelihood goes up and up and up even as the incentives to spoof your kid’s disability go down and down and down.

California Single Payer Costs Look Very Doable

In a prior post, I wrote about a report from the California Senate Appropriations Committee that estimated the cost of a proposed single payer system in the state. The report concluded that, under the single payer plan, health expenditures in the state would be around $400 billion per year, or 15% of California’s GDP.

Although many covered this as a ridiculously high figure, it is actually quite cheap relative to the US as a whole, which currently spends 18% of its GDP on health care. For those not reflexively afraid of large numbers, it is clear that this cost is definitely doable and is quite a bargain for the benefits the single-payer plan is said to provide.

After that report came out, a new, more detailed cost estimate was provided by economists at UMass Amherst. The methods in the UMass paper are mercifully easy to follow. The cost estimate basically works like this:

  1. Determine the current level of health expenditures in the state using the national health expenditure database. This includes spending on hospitals, physicians/clinics, pharmaceuticals, dental care, nursing homes, home health care, and insurance administration. This figure is $368.5 billion.
  2. From there, they determine how much health care utilization would increase under single payer. They assume that uninsured people would double their health care utilization and that underinsured people would increase their health care utilization by 15%. They inflate the spending of those groups of people by these percentages (adjusted slightly for the age composition of the two groups) and arrive at a new figure for total health expenditures in the state: $404.1 billion.
  3. From there, they determine how much savings a single-payer system could deliver over the current system. They estimate that savings on administrative costs would reduce total spending by 6.7%, that savings on pharmaceutical drugs would reduce total spending by 3.4%, that savings from switching to Medicare reimbursement rates would reduce total spending by 2.9%, and that savings on unnecessary services, inefficiently delivered services, prevention, and fraud would reduce total spending by 5%. In total, then, the cost savings trim 18% off the figure in (2), giving a final estimate of $331 billion, or 12.5% of GDP.

Needless to say, this cost is quite a bit lower than the one estimated previously by the Senate Appropriations Committee. If the costs really would just be 12.5% of GDP, then that means California’s health sector size would shrink to around the OECD average.

If the authors’ estimates are too rosy, it is probably because they underestimate how much utilization would increase.

They estimate that the uninsured would double their utilization based on data showing that uninsured people consume about half as much health care as insured people consume, which seems reasonable enough. Then they estimate that underinsured people are 36% of the insured population and that their utilization would increase by 15%, which was the high-end estimate from a paper about what happens when people are switched from a no-cost-sharing plan to a high-deductible plan. And that’s it. That’s all they did on utilization, which may not have been enough.

But in any case, even if their utilization estimates were massively too low, the plan is still affordable. The authors could be off by 20% and the plan would still be eminently doable.