African American teenage birthrate and reliance on welfare

The following is posted to support a discussion I am having elsewhere. The Political Scientist is arguing, or seems to be, that teenage birth and recourse to welfare shouldn't be characterised as a problem or 'poor choice' and offering long-acting contraception is the wrong approach.

I have already responded to one claim he cites - "there was likely no correlation between the level of welfare ben­efits and the incidence of out- of-wedlock births" -  at his blog.

Below I address two more:

Claim:

 "An examination of the rate of babies born to unwed African American teenagers remains virtually unchanged from 1920 through 1990.[39]"

The "rate of babies born" means the number of births per 1,000 15-19 year-old African Americans.

The graphs below show plummeting unwed teenage births to African Americans since the early 1990s. Many individual states had their own reforms well before 1996 - the year when federal reform occurred.

The claim implies welfare benefits had nothing to do with the rate of African American illegitimacy 1920 to 1990 because of its constancy (ignores the steep rise during the 80s).

In which case you would have to conclude that reforms have nothing to do with the tumbling rates after 1990. I don't accept that. Neither do I rule out other factors at play.

The rate has continued to fall by the way with a slight blip somewhere around 2007-08 (from memory).

Source

Claim:

"...the myth of widespread permanent welfare status proved equally impossible to correct. Numerous reports and studies demonstrated that welfare was a transient state for most recipients and that it served primarily as a temporary safety net during financial crisis caused by job loss or family crisis.[42]”

 The Political Scientist reiterates this in a comment to me:

"I’m also aware that you have spent much time and energy arguing for ‘welfare reform’ on the assumption that it creates something that you, and many others, refer to as ‘welfare dependency’.

It is, of course, mistaken but, given the correlational nature of much of the research that is perhaps understandable. We are all prone to confirmation bias."

The statistical phenomenon of short-term reliance hiding long-term is explained below via NZ research into how long people rely on the DPB:

On average, sole parents receiving main benefits had more disadvantaged backgrounds than might have been expected:

·          just over half had spent at least 80% of the history period observed (the previous 10 years in most cases) supported by main benefits

·          a third appeared to have become parents in their teenage years.

This reflects the over-representation of sole parents with long stays on benefit among those in receipt at any point in time, and the longer than average stays on benefit for those who become parents as teenagers.

Had the research considered all people granted benefit as a sole parent, or all people who received benefit as a sole parent over a window of time rather than at a point in time, the overall profile of the group would have appeared less disadvantaged.

 And another explanation I have used before from The Poverty of Welfare:

8 out of 10 years reliant on the DPB is a good enough definition of 'welfare dependency' for me.