Still More Evidence that COBS is a Flawed Concept

Since I recently posted a piece by an Arizona reader about that state’s efforts to radically increases in child support, and the failure of those efforts, I thought I’d gild the lily just a bit.

I linked to a couple of pieces I did back in July criticizing the new COBS (Child Outcome Based Support) guidelines and its intellectual support provided by Professor Ira Ellman.  Ellman’s paper, written along with his wife, purports to analyze the policies promoted by child support and to provide a better model for achieving them, i.e. COBS.

There are a multitude of problems with Ellman’s paper, only a few of which I was able to address back in July.  So here’s another one.

Ellman’s approach includes the idea that, because children tend to do better in families with more income than in those with less, if we’re to improve children’s outcomes, then we need to increase incomes to their custodial parents.  He provides a number of limitations to that basic approach, but the upshot is that he mostly pays lip service to the limitations.  The “bottom line” can be readily seen in the proposed Arizona guidelines that would increase child support obligations enormously – in many cases over 50%.

Now, as I pointed out in July, if income is the main factor in child outcomes, then the clear, simple public policy should be to grant primary custody to the parent with the larger income.  That obvious conclusion to Ellman’s argument escaped him entirely, for what I suspect should be obvious reasons. 

But since I first critiqued Ellman’s work and the proposed guidelines, I’ve also reported on an important book by Canadian academic, Paul Millar called “The Best Interests of Children: An Evidence-Based Approach.”  And, not coincidentally, Millar reports some findings that strike directly at the heart of Ellman’s work.

Millar of course is Canadian and he’s dealing with data from that country’s National Longitudinal Survey of Children and Youth.  So, to the extent that U.S. figures are radically different from Canada’s, our conclusions should differ as well, but I suspect any differences are slight.

As I’ve said in previous posts, Millar is concerned with determining, in a rigorous, scientific way, what actually benefits children and what doesn’t.  After all, our courts routinely announce that their decisions are in the best interests of the child, so maybe it would be a good idea to find out what actually promotes those interests and what doesn’t.

For example, given that in Canada, some 90% of primary custody is granted to mothers, Millar asks the obvious question: is the sex of one parent correlated with better child outcomes than that of the other.  And the answer is a resounding “no.”  There is essentially no way we can analyze the data that allows us to conclude that children do better with mothers than with fathers, and yet public policy, enacted every day, many times a day assumes they do.

But we know that income is important to child wellbeing, so Millar analyzed the data to find out to what extent.  What he found sinks Ellman’s notions faster than a submarine with screen doors.

First, “the impact of income (on child wellbeing) is considerably less than that of parenting.” (parenthetical mine)  That is, the most important thing to children is good parenting.  Of course greater income often means parents have more time and resources to be good parents, but Millar accounts for that.  The main point is that children of families at any income level can have good parents and that far outweighs any improvement increased income might provide.

More to the point though is the question how much extra income does it take to produce measurable improvements in children’s welbeing?  After all, that’s what Ellman is arguing for – that we should increase child support because, by doing so, we’ll make life better for kids.  And it’s precisely that notion that Millar shoots down.  In discussing the improvements that attend increases in income, he pointedly writes, 

“[I]t should be noted, however, that these effects require a tenfold increase in household income, something not easily achievable by social policy.”

(Emphasis mine.)  In other words, in order to achieve Ellman’s desired outcomes for children, noncustodial parents would have to pay enough in child support to increase the custodial parent’s income ten times.  Millar’s remark that such a result is “not easily achievable by social policy” is putting it mildly.  In fact, in all but the most exceptional cases, it would be impossible.

The point being that what Ellman proposes and what the State of Arizona almost agreed on would not only have bankrupted many noncustodial parents, particularly in a time of economic hardship, but into the bargain, it wouldn’t have benefited children.  

The more one looks at Ellman’s proposal, the more one concludes that it’s a stalking horse for another policy altogether – the wholesale transfer of fathers’ incomes to mothers.  It’s a policy that Ellman never discloses, but it’s hard to see it any other way.

Leave a Reply

Your email address will not be published. Required fields are marked *