Saturday, October 21, 2017

Veterans, the Living Wage, and the McNamara Fallacy

In the first post for the eighth year of the blog I was going to reflect on the Ken Burns film about Vietnam.  My first impression was how little things have changed since then with all of the protests.  The second thing that jumped out at me was President Johnson's defense secretary's obsession with collecting data (mostly body counts) to determine who was winning the war.  This is called the McNamara fallacy and is discussed below.

In my blog and my other writings I use statistics to enlighten people and to shed light on various social phenomena.  For example, for the Hill Talk, I looked at various variables which may predict the magnitude of the increase in the living wage for the cities/counties that have passed such ordinances.  

The graph below shows the strongest predictor which is the percentage of veterans in that city/county.  As the percentage of the veteran population increases by one percent, the amount of the living wage decreases by an expected 59 cents.  This relationship accounts for 28% of the variability in the amount of the living wage passed. 

The mean % veterans of the 38 living wage entities is 5% while the US as a whole has 6.2% of its population who are veterans.  Case in point Seattle, WA passed a $15/hour living wage ordinance while nearby Tacoma, WA passed a $12/hr wage.  Tacoma has 9.34% of its population as veterans while Seattle has less than half at 4.54%.  All of the cities that have passed a $15/hour wage or higher have % veterans that are lower than the US as a whole.  Six of the nine cities/counties with wages $15/hr or higher are in California.  

If one spends too much time looking at the leaves and the twigs on a tree, one can miss the surrounding forest.  This is basically what the McNamara fallacy is.  New findings with statistics can reveal important features of the forest as I believe this analysis has with regard to the forest activists must navigate to pass a living wage ordinance.  

The percentage of veterans in a city/county was the most robust variable negatively associated with the amount of the living wage increase after considering the % poverty, the % foreign born, the % change in the population, the % uninsured, the % in poverty, median household income, median housing value, and the % with a high school education or higher.  The full data set used in this analysis can be seen here.  The amounts of the minimum wage increases were found from the National Employment Law Project or NELP.  The demographic information on the cities/counties that have passed these ordinances was found from the Census Bureau at

Unlike McNamara and later Donald Rumsfeld and their ilk, I do not claim to have a full grasp of the whole forest surrounding the Fight for $15.  Further research is needed to fully understand the forest.  An argument could be made that it was the arrogance of men like McNamara and Rumsfeld that created the large population of veterans in the US.  One would think that if anyone could use a raise the veterans could.  A significant portion of the homeless population are veterans.

**Related Posts**

What do Living Wage Cities Have in Common?

Sunday, October 1, 2017

Seven Years of CSIwoDB

Usually for my anniversary posts I look at the web traffic statistics for the site.  This year I thought I would take stock of where the blog is and where it will go.

Seven years ago I began the blog by conducting data driven journalism before I had even heard of the term.  This year I have been outsourcing my writing skills to the websites Data Driven Journalism, Kolabtree, and The Hill Talk.  These opportunities are nice.  However it takes a long time to put together data driven posts.  Opinion pieces like this one are a lot less time consuming to write.  

This blog was created to fulfill three purposes.  

  • To educate people on how statistics can help people to see patterns that aren't easy to see with the naked eye.
  • To keep my statistical skills sharp.
  • To showcase my skills when I am not working.
After seven years I have had some success on these fronts.  My all time most read post was linked to on the BBC webpage for the program The Joy of Stats and still gets views on the web.  I already spoke on the outside columns that I am making.  There is more and more competition for eyeballs on the web however but at the same time there are more and more eyeballs too.  Sometimes it seems like I'm a voice crying out in the wilderness but it's still necessary to cry out.

On the second point I feel that it has kept my skills sharp for the skills I learned in school. There have been methods that have come out since then that I have not learned such as natural language processing.  Regardless of the new methods that come out it is still important to stress the basics.

I have had more work this year (though most of it unpaid) and I have less time to post so I will have less time to post here at least until my city council campaign ends on November 7. II did get the above note placed under my windshield wiper recently that I had to share with you all.

**Related Posts**

CSI senza cadaveri

Wednesday, September 27, 2017

Pennsylvania Lags Behind US in Startup Job Creation

Last Thursday I attended the Startup Alleghenies launch event to network with those interested in learning how to create one's own company.  The presentations were interesting.  I went to offer my statistical services to anyone who had such a need.  

Looking at the census bureau's website I came across this article 

Startup Firms Created Over 2 Million Jobs in 2015

The map below shows the % change in startup job creation by state.  The states with the largest increases are in the west the southeast, and Michigan.  These numbers are encouraging but the report states that "this level of startup activity is well below the pre-Great Recession average of 524,000 startup firms and 3.3 million new jobs per year for the period 2002-2006." 

The growth in startups in different states may be reflective of the business climate there. The article does not explicitly state what the definition of a startup company is.  It speaks of old (> 25 years old) and young companies (< 6 years old) but does not state how startups fit into this picture.  It also provides no information on how long startup companies tend to last. 

People love to quote how Apple computer was founded in a garage but often they are the exception to the rule.  Would there be differences between the states in how long startups last?  Usually one discovery raises more questions than it answers.  I'm sure the data exists elsewhere.

On a side note I have begun to write columns for the The Hill Talk online publication.  I have written two posts there so far.  My next post will be the seventh anniversary post where I take stock of the blog.

**Related Posts**

Personal Bankruptcy Rates as a Measure of Underinsurance (Cross Post with PUSH)

Tuesday, September 19, 2017

Medicare for All Forum in Johnstown, PA

Bernie Sanders has introduced a Medicare for All (formerly single payer) bill to the Senate for consideration.  He garnered 16 cosponsors in the Senate including Elizabeth Warren, Kirsten Gillibrand, and Cory Booker.  It has no chance of passing the current congress as they are now considering yet another repeal of the Affordable Care Act (aka Obamacare) but it can be the catalyst for a better healthcare system in the future.  We must pick up the torch (not a tiki torch or a literal torch) and speak out.  Here is just such an opportunity:

Medicare for All Forum
Cambria County Library Community Room
Thursday, September, 28

A discussion on how universal health care that is not profit driven can save money and lives. 

**Related Posts**

Parallels between 2010 and Today's Healthcare Debate

US Life Expectancy Decreases and Those Who Want the Affordable Care Act Expanded Increases


Santorum: No One Has Ever Died Because They Didn’t Have Health Care |The New Civil Rights Movement

Saturday, September 9, 2017

A Successful Campaign Kickoff Rally for Labor Day

We had a successful campaign kickoff rally for my City Council run.  It was on Labor Day in the afternoon at Central Park in Johnstown.  16 people came including Rachel Logan from Our Town Johnstown who wrote a nice article on the event and took the photo of the event seen below.  You can see my campaign platform on my website.  Other photos that I took are also embedded.

Thanks to Woody Weaver for making a video of my Campaign Speech at my Campaign which can be seen at the top of this post.  

**Related Posts**

Hillary Clinton Won Johnstown, PA -- By 83 Votes

A Successful Rally for Women's Rights

Thursday, August 31, 2017

PodCamp is Back for 2017

PodCamp Pittsburgh is back for 2017 at the Point Park University Library.  It is a social media unconference where practitioners share the tricks of the trade.  It is free to attend and if you like you can propose your own topic to share your wisdom with the masses.  

I have presented at three previous PodCamps.  The most recent presentation can be seen below.  Because I'm running for city council I cannot present this year but I encourage you to learn valuable social media skills.  David DeAngelo of 2political junkies will present on fact checking Donald Trump.

Major figures from the Pittsburgh media go as well as business figures and activists from across the political spectrum.  The event will be live streamed for those who can't go.

**Related Posts**

Podcamp Pittsburgh XI Video Now Posted

My Raging Chicken Press Interview 

Podcamp Session Feedback Part 2-The Video of My Session Has Been Posted

A New Science YouTube Channel


Wednesday, August 23, 2017

Trump Approval Rating in Pennsylvania Now at 35%, What is it for Cambria County?


Last Sunday Meet The Press (the worlds longest running television series, not The Simpsons) announced the results of a poll they conducted in the pivotal states of Michigan, Pennsylvania, and Wisconsin of President Trump's current approval ratings in those states.  His approval ratings in these states are 36%, 35%, and 34% respectively with a margin of error of +/- 3.5%.  In 2016, Trump barely won Pennsylvania with 48.8% of the vote.  In Cambria County, Trump won with 67% of the vote.  

Assuming that the differences still hold between the county's approval for Trump and the State's approval still hold, we could extrapolate these results to Trump's approval rating in the county.   This extrapolation rests on another assumption that his % of the vote is close to his overall approval rating.  With his current PA % approval being 13% lower than his % of the vote last year, his approval in Cambria County may be closer to 54% now.  

You may think that these assumptions are shaky at best.   Yes it would be better to conduct a survey here to obtain a more precise estimate of his approval rating in Cambria County but these surveys are expensive to conduct.  Often times extrapolation is the best we can do.  It is safe to assume that his approval rating in the County is lower than it was last November.

Last Thursday here in Johnstown, PA (the largest city here in Cambria County), there was a unity rally in response to the terrible events in Charlottesville, VA.  A large crowd was present as the panoramic photo I took shows.  It was larger than the Love Trump's hate rally the Unity coalition had when Trump visited last year.  On the Tribune Democrat Facebook page the post for the event received 115 responses on their response buttons:  three of which were angry and the post was shared 20 times.

  **Related Posts** 

Hillary Clinton Won Johnstown, PA -- By 83 Votes

Cambria State Senate and House Races

Facebook Followings Predict The Outcomes of Elections in PA