The Magic Bureaucrat and His Riverside Miracle

The Iron Law of Program evaluation: &quot;The expected value of any net impact assessment of any large scale social program is zero&quot;; the stainless steel law: &quot;the better designed the impact assessment of a social program, the more likely is the resulting estimate of net impact to be zero.&quot;<p><a href="http:&#x2F;&#x2F;www.gwern.net&#x2F;docs&#x2F;1987-rossi" rel="nofollow">http:&#x2F;&#x2F;www.gwern.net&#x2F;docs&#x2F;1987-rossi</a>

&gt; Neither the federal government nor the states that ran the welfare-to-work programs had the political will to rethink the programs, which they had worked already to reform.<p>This is an under-appreciated constraint on government-run welfare programs and other government-run charities.<p>The larger problem – providing welfare and other forms of charity – is, of course, huge, complex, and possibly impossible-to-solve (completely).

What is not mentioned here is that many of the welfare recipients moved to SSDI (Social Security Disability Income). SSDI is paid for by the Federal government, so the states were incentivized to encourage this move. In fact, there are companies that specialize in applying for SSDI for welfare recipients, and charge per successful application.<p>There&#x27;s the feel-good story, and then there&#x27;s the harsh reality. The reality is, SSDI enrollment has gone up by a factor of 3X since this &quot;welfare reform&quot; thing passed.<p>I have a neighbor who is reasonably fit (he can move 100lb+ stones in his side gig as a landscaper), but he is on SSDI and collects a cool $3K&#x2F;mo for doing nothing. Once you are on SSDI, it&#x27;s very difficult to kick you off.

The magic of regression to the mean - &quot;They compared these subjects to those participating in other California programs that had focused on education and training. What they found was that the effects of the &#x27;Riverside Miracle&#x27; had all but disappeared.&quot;

<p><pre><code> &gt; Before Townsend and Clinton, the prevailing philosophy &gt; around welfare was that single mothers needed it so they &gt; could afford to stay out of the workforce and raise &gt; their children. </code></pre> I wonder if it might not make sense to return to those older ways (for those parents that wish it): childcare is expensive and paying one parent not to work (or to give them the security to work part-time) may have a greater societal benefit &#x2F; societal cost savings than aiming for full employment.