America's Welfare State fails

[Admiral Yi]

However, none of the papers seems to analyze the pure existence of an unemployment insurance, merely the effect on when people search for a new job.

You lost me.  That's what we've been discussing: the effect of unemployment insurance on job searches.

[Jacob]

However, none of the papers seems to analyze the pure existence of an unemployment insurance, merely the effect on when people search for a new job.

You lost me.  That's what we've been discussing: the effect of unemployment insurance on job searches.

I thought we were discussing whether unemployment insurance helped the poor or not.

[Admiral Yi]

Have you considered reading the thread Jacob? 

[Jacob]

Have you considered reading the thread Jacob? 

I did.

Here's the exchange that I've been reading:

Except neither article actually shows anywhere that it isn't helping the poor.

Sure they do.  Raising the minimum wage causes unemployment and unemployment insurance disincentivizes job searching.

While it may be true that unemployment insurance is a disincentive for job searching, the conclusion that it creates poverty does not necessary follow. It could just as well be that without unemployment insurance you would just have more destitute people that cannot find employment at all. In recent years the rate of employment in the US has dropped drastically. Searching and finding a job is not the same...

Zanza asked if there is any proof that UI harms the poor. You claimed it disincentives job searches. That claim, if true, does no show that the poor are being harmed; merely that job searches take longer.

The original question that Zanza asked, and which you responded to, asked about harming the poor. You may - or may not - be correct that job searches take longer, but you will not have answered the question that I - and I presume you - read earlier in the thread.

[Admiral Yi]

You ignored the part about the minimum wage increasing unemployment, but leaving that to one side, your point is taken.

However, we do have reams of articles in the other long thread about the perniciousness of long term unemployment.  A dollar of unemploment insurance may be interchangeable with a dollar of income, but the UI comes with the negative of decreasing the recipient's employability.

[DontSayBanana]

You ignored the part about the minimum wage increasing unemployment, but leaving that to one side, your point is taken.

However, we do have reams of articles in the other long thread about the perniciousness of long term unemployment.  A dollar of unemploment insurance may be interchangeable with a dollar of income, but the UI comes with the negative of decreasing the recipient's employability.

Bull.  You're conflating the unemployment insurance with the period of unemployment itself- I've only ever had one employer ask directly whether I received unemployment during a specific period, and that was in a personal conversation, not as part of an interview process.

[Admiral Yi]

Bull.  You're conflating the unemployment insurance with the period of unemployment itself- I've only ever had one employer ask directly whether I received unemployment during a specific period, and that was in a personal conversation, not as part of an interview process.

I'm not conflating anything  I'm including 2nd order effects.  UI prolongs spells of unemployment.  Longer periods of unemployment reduce employability.

[The Minsky Moment]

Is there a way we could actually test something like this?  It is hard to find testable and repeatable conditions for economics theories.

Econometrics!

Which is its own methogological battlefield.

[DontSayBanana]

I'm not conflating anything  I'm including 2nd order effects.  UI prolongs spells of unemployment.  Longer periods of unemployment reduce employability.

OK, so you're addressing a correlation without establishing a causation.  You haven't proven UI is not necessarily the primary cause of longer periods of unemployment, since you can't prove it's a cause of the unemployment, you can't prove it's a cause of the stigma.

[Admiral Yi]

Which is its own methogological battlefield.

The only real methodological battle to be fought is over omitted variable bias.

[Admiral Yi]

OK, so you're addressing a correlation without establishing a causation.  You haven't proven UI is not necessarily the primary cause of longer periods of unemployment, since you can't prove it's a cause of the unemployment, you can't prove it's a cause of the stigma.

Whenever I hear this old bromide repeated it makes me think people believe their is a test for causation, like you dip litmus paper in and if it comes out blue you have causation.

Quote unquote proving causation is just the act of attaching a plausible mechanism to an observed correlation and ruling out other plausible mechanisms with correlate similarly with the outcome.

[The Minsky Moment]

Which is its own methogological battlefield.

The only real methodological battle to be fought is over omitted variable bias.

It's broader than that.  Models can be specified in many different ways - including variable choice, sample period, dynamic assumptions.  It is a notorious problem that different specifications can give rise to different significance results - aside from the obvious risk of cherry-picking, it raises the difficult question of how one objectively determines the "right" choice or how or whether to incorporate a proper senstivity analysis.  On top of that is the old Lucas Critique - the claim that even if one can demonstate a robust historical empirical economic relationship, one cannot assume that the relationship will continue to hold if policy changes.

These aren't irrelevant concerns - there are plenty of studies out there concluding that there is little negative employment effect from raising the minimum wage; similarly it is not difficult to find studies finding only small impacts on UI on employment.

[Admiral Yi]

Actually we've been over the minimum wage issue pretty exhaustively Joan.  The majority of studies show an elasticity of -.1 to -.3.  The only real controversy there is if you spin that as "close to zero" or as "overwhelmingly showing a negative impact on employment."

[The Minsky Moment]

The only real controversy there is if you spin that as "close to zero" or as "overwhelmingly showing a negative impact on employment."

Both.  The studies tend to show a negative effect that is very small.  Some studies have shown no statistically significant negative effect.

[DGuller]

However, none of the papers seems to analyze the pure existence of an unemployment insurance, merely the effect on when people search for a new job.

You lost me.  That's what we've been discussing: the effect of unemployment insurance on job searches.

Are we actually discussing this?  I think we're discussing whether unemployment insurance helps or hurts the poor.  You're the one who discusses the side effects of the policy without even acknowledging the fact that they need to be compared to the benefits.