Right to Work" snark hunt

Oct 28, 2012 8 Replies

I noticed that the "Right to Work" controversy has come up again in another thread. As some of you know, one of my "hobby horses" is econometrics, or the application of statistics to discussions of this type where objective data is available. [FWIW -- if objective data is *NOT* available, it is theology.]



For those of us that may be interested, I downloaded a list of the states, and dummy/binary coded the "right to work status" as 1 = "right to work", 0 = no "right to work, and from other sources downloaded family poverty rates, median household income and a measure of income distribution equality called the gini coefficient.



To download a copy of the data set click on

formatting link



Some quick spreadsheet work shows the un-nuanced data, i.e. "out of the box," not adjusted for anything, i.e. not population weighted, no logit/probit transformations etc. indicates the following:



Statement: Right to work states have higher family poverty rates. For RTW v family poverty rate in percent R square = 0.1153072619 F-RATIO = 1.42732643240265E-046 F-dist = 1.0000 Conclusion. NO SALE



Statement: Right to Work states have lower family incomes. For RTW v median family income R square = 0.1569634784 F-RATIO = 9.39707857939395E-192 F-dist = 1.0000 Conclusion: NO SALE



Statement: Right to Work states have higher disparity in income distribution as measured by the gini coefficient. For RTW v state gini coefficient R square = 1.02125374946014E-005 F-RATIO = 5.96807933525614E-058 F-dist = 1.0000 Conclusion: NO SALE



Cautions: As the calculations were not population weighted, the larger states such as California may have "swamped" the results, but this is highly doubtful as the split is c. 40% of the population lives in RTW states and 60% do not. D.C., P.R., the territories, etc. were omitted from the calculations. Not adjusted for family size. Other statistical tests such Kruskal?Wallis as may give different results. [Feel free to download data and "knock yourself out."]

formatting link


(you may find my LFTR01 proposal {pdf and odt format} in the drop box of interest also -- feel free to browse website)

formatting link



File uses the free OpenOffice suite which you can download at

formatting link
{current version



3.4.1} Many versions of MS Excel will also open the file.

For more info on R-squared to start see

formatting link



For info on the gini coefficient see

formatting link



Google on either or both terms for more than you want to know.



==>Now lets move on to something than means more than a p*** hole in the snow before the politicians/banksters bring the economy down around our ears.


Interesting but actual poverty is what you can buy with what you have. Take someone from NY City and transplant them in MS or NM, they might be living high on the hog. Cost of living matters. I guess my question is, "Do high prices follow high wages? Or vice versa?"

...

...

The most important thing to remember about R^2 is that "correlation does _not_ imply causation"...

For such a simplistic measure of such a complex question, the number of confounding factors one can postulate is essentially unlimited...

What it demonstrates is that "right to work" laws do not reduce poverty levels.

Indeed, but this only applies ==>if a correlation is detected.

On 10/29/2012 11:47 AM, PrecisionmachinisT wrote: ...

It also does not demonstrate that right to work laws have any causative bearing in preventing rise from poverty...

causative

I never said it did.

However, the evidence strongly suggests it to be the case.

===============

While this seems plausible on its face, where is the data ==>and the analysis?

Except, upon analysis, there are some troubling facts..

For instance, one of the main purposes of collective bargaining is that of extracting higher wages from of an employer.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required