Population Growth and School Segregation

Recently my office bought new software that allows you to plot data onto a map by any demographic.  Literally, you can plot data of houses that were built before 1940 by every county in America.  I am not sure why you would need to, but the point is you can (as pointed out below).   

Number of houses built before 1940 in Washington, DC.  Darker shades of red
represent the larger number of house build before 1940.

Once I learned how to use the new software, you better believe I had some fun with it.  I have done a good amount of research looking at school segregation within the state of North Carolina among African Americans.  The current literature attributes school segregation in large part to residential segregation.  However, there is much discussion on the importance of geographic area and the number of students in a school district.  Current research shows evidence that both geographical area of a school district and the number of students in the school district have direct relationships with levels of school segregation.  However, I always believed there was a missing element that could bridge the gap between area and student enrollment.  That variable is density, and until now, it was never tested in regards to school segregation.  

I used North Carolina as a model to analyze the way density affects school segregation for a number of reasons. The primary reasons are that North Carolina contains countywide school districts (school districts and counties can be used interchangeably because the school district is countywide).  North Carolina segregated its schools based on race during the Jim Crowe Era.  After the abolition of “Separate but Equal,” North Carolina was involved in desegregation efforts via busing.  The state has a rich history of school segregation, and it is interesting to see the effectiveness of the policies.    

I ran some regressions on density and it showed that density was a significant contributor to levels of school segregation.  It could be expected, from the regression model, that as the density of a school district increases by one percent, overall school segregation within the county should decrease by .186 percent.

If you look at the first map of North Carolina, it shows density levels in each measurable countywide school district (84 in total).  The blue spots on the map represent high or low-density areas.  The second map looks at school segregation levels.  

 

As you can see from comparing both maps, school districts that are denser tend to be less segregated than less dense districts, which is also proven by the regression.  There are good reasons why this finding is important.  Demographic trends play a huge part in political outcomes.  In the time between 1990 and 2000, the population of North Carolina grew by more than 21%, which means the state has become denser.  

Map 1: Density by county.  The darker the shade of red represents denser areas.

Map 2: School segregation by County.  The darker the shade of red, the higher level of school segregation within the school district.

The purpose of this research was to analyze how dense areas (cities) and non-dense areas (rural areas) relate in terms of school segregation.  Current demographics show that cities are becoming more populated.  The problem that comes with a growing population is urban sprawl.  What happens under urban sprawl is city limits expand to accommodate the growing population.  There are a good amount of issues with this, including traffic congestion and environmental issues.  However, along with urban sprawl comes school segregation.  In order to have declining levels of school segregation, you need the number of students per square mile to increase.  If the district enrollment/square miles (which is the formula for density) has growth in the denominator, even if district enrollment grows, you will not achieve desegregation because of the growing geographical area.  The better way to alleviate a growing population is to build within the city.  Renovating depressed areas, improving transportation in downtown areas and other ways to maintain a more dense population will alleviate the problems that are associated with growing populations. 

My Choice for Baseball's AL Rookie of the Year

It is that time of the baseball season when the most important questions are answered.  Not who will win their divisions, but personal accolades.  You have to remember, baseball is a game of stats, and no matter how much of a team player a person is, they will always keep track of strikeouts they have in a season or the number of homeruns they have hit.

This year is especially exciting because of the plethora of talent that has emerged in baseball.  My favorite race is the Rookie of the Year race.  My interest in this was restored last night after seeing Yankees rookie Ivan Nova throw 7.1 innings of 5 hit, 1 run ball.  Nova has been lights out all year.  He has posted a 15-4 record with the third most innings pitched in baseball among rookies and 3.81 ERA.  Those are pretty impressive stats for any pitcher, regardless a rookie.  I normally would not consider a pitcher for an MVP award because they throw every five days if they are a starter, or an inning or so every few games.  However, the Rookie of the Year award is different because it really depends on the talent pool of the rookies.  

I found the most recent rookie stats from mlb.com.  I realized Ivan Nova has impressive stats, but there are rookies in the AL that have stats that are even more impressive.  I decided to only include those players that have at least 350 at bats or 100 innings pitched.  The rationality for this is that it is a better indicator of a player’s real talent.  The more innings pitched or the larger number of at bats, you can get a better sense of the players performance because of the larger sample size.

I decided on two candidates for the AL Rookie of the Year, and Nova was not in consideration.  The first is the Angel’s Trumbo who has 27 homeruns with 82 RBI’s and .256 batting average.  He also leads AL rookies with 500 at bats.  He is also second in the AL with 128 hits (Homser from KC is first with 134).  His batting average is a little low, but among AL rookies, he is ranked third.  He has also played in the most games (139) than all AL rookies.  His downside is he has a .295 on base percentage (which is awful) and 108 strikeouts.  He does not get on base a lot, and strikes out a good amount, which is a big problem.    

Trumbo hitting his 15th homerun of the year

The other possibility is Tampa’s Hellickson.  Hellickson is having an unreal season.  He has a 2.96 ERA with 170.3 innings pitched and 109 strikeouts.  He also leads all rookies with 2 complete games and is the only rookie to throw a complete game shutout.  However, he has a 12-10 record, which is his only real downside.  If we use statistics derived from sabermetrics (that was compounded by ESPN, definitely not by me), he has 6 tough loses and no cheap wins.  This shows that sixty percent of his losses came in games where he threw at least six innings while giving up less than three runs.  He also has a very respectable WHIP of 1.14.   

In my opinion, the AL Rookie of the Year has to be Jeremy Hellickson.  There are rookies that have impressive stats, but are weak in important areas.  Hellickson has strong stats from a team that lost its bullpen and best players in the off-season.  Hellickson also pitches in the very competitive American League East where he has to go against the best hitters in all of baseball.           

Jobs, Education and the Economy for the Month of August

The Department of Labor released its August job numbers, and the aggregate unemployment rate remained the same from July.  If you are an optimist, this means the economy has not gotten worse; but if you are like me, this is a sign that economy is still struggling.  But, I always like to break down numbers into its roots to see what exactly the unemployment is derived from. 

I looked at the numbers and here are some results:

• The unemployment rate for those 25 years or older, and have at least a bachelors degree is 4.3%.  That is not a joke, the unemployment rate is half that of the national average.
• The telecommunications industry lost more than 47,000 jobs in the month of August.  

But the news that there was no net job gain in the month of August made headlines this morning on cable news and really shook the confidence in the stock market.  However, the numbers show a perfect tradeoff between the job separation rate of private and public sector workers.  In August, 17,000 new private sector workers were hired whereas 17,000 public sector workers were laid off.  This stat can be interpreted by some as when the government becomes smaller, it allows for more room in private industry to grow and absorb the unemployed government workers.  At face value, you can come to that conclusion.  However, the areas where the private sector is growing is different from where the public sector is shrinking.

34,000 new jobs were created in the health service industries throughout the month of August.  Professional and business services saw 28,000 jobs added (accounting, architecture, administration services, computer IT work, etc.).  The government subtracted 17,000 jobs throughout August.  15,100 educational workers lost their job in the month of August.  Out of that number, 13,700 were at the local level (most likely to be teachers), and 1,400 at the state level.  Almost 89% of all laid off government workers in the month of August were in the education sector, with almost 81% coming at the local school level.  

It is a problem that the private sector areas that are growing and the public sector fields that are seeing job losses are completely different.  Those working in the field of education cannot be absorbed as an architect by the private sector because the worker has a different skill set than the job entails.  Due to the discrepancies in skills, the job search will be a painful one.   There is a mismatch in the job market that will cause prolonged unemployment. 

The Case for A.J. Burnett at Home

September 1^st is always the best time of year for a sports fan.  College and professional football is starting, and the Yankees are always right there (always alongside the Red Sox) to win the American League pennant.  During this time of year, everything in baseball is thrown on its head.  Teams are trading players in last minute efforts to boost their rosters for the playoffs (even after the trade deadline), rosters expand to 40 players, and the divisions are starting to shape up for the post season. 

During this time, managers are trying to set in stone their pitching rotation for the playoffs.  This couldn’t be more true than for Joe Girardi and the New York Yankees.  The obvious ace of the staff is CC Sabathia.  There is no question that he will be starting game one of the playoffs.  But who will be number two?  I get a lot of grief for this, but I say, if the Yankees are at home, go with A.J. Burnett.

You are probably shaking your head right now, but let’s look at his numbers. 

This year, his number of walks and hits per innings pitched (WHIP) is 1.44.  There are 37 starting pitchers in baseball that posted better overall WHIP’s than A.J.  However, in my opinion, this number is misleading. 

I always had a hunch that A.J. was a better pitcher at home than away.  I decided to compile data over the course of his Yankee’s career in an effort in trying to justify my feeling.  

Here is what I found.  Over A.J.’s 27 appearances this year (all starts), he has pitched at home 15 times and away 12 times.  However, his WHIP’s are significantly different.  At home, his WHIP is 1.25 and away it is 1.73.  If you look at that over the course of an entire ball game, that is more than four less base runners per game!  He gives up about 2.6 less earned runs per game at home, and he is averaging one less walk per nine innings.  These are all big numbers, especially when you are comparing them to the options the Yankees have.  Colon has a higher WHIP (1.35) than Burnett at home, and has pitched fewer innings. However, Colon’s WHIP away from Yankee Stadium is an impressive 1.11, ideal for a game three away start for the Yanks.  Hughes has been out all year and we are not really sure what to expect from him.  He can be lights out or get hit for 10 runs.  Nova, who is a fantastic pitcher, is a rookie and doesn’t have that big game experience, plus has an at home WHIP of 1.52.  Garcia, who just came off the disabled list, has an at home WHIP of 1.35.

The reason why I am looking at WHIP is that it is an extremely telling measure of a pitcher’s success.  The more hitters that get on base, the greater likelihood that the batter scores.  With more men on base, an error or routine pop-up or groundball can mean a run.  The less runners on base means the less chance of giving up runs, and a greater margin of error for the team to get out of an inning unscathed.  Also, WHIP measures the pitcher’s ability to get outs.  The pitcher is in control of walks and hits.  When he pitches the ball, no other player has control of the location or type of pitch.  There are times when teammates make exceptional plays that take away hits, in essence, lowering the pitcher’s WHIP.  However, routine balls that should be outs turning into hits even this out and judgment calls by umpires that go against the pitcher.    

The point I am trying to show is that A.J. is a good pitcher at home.  His big flaw is the number of homeruns he gives up at Yankee Stadium (the short right field does not help).  If the Yankees are at home in a big game, (is CC has already thrown that series) throw A.J. out there.  I would love to see him bring a win to the Yanks.   

Real Effects of the Stimulus Package for the Second Quarter of 2012 and the Future

We are almost two full years removed from the passage of the stimulus package.  As an homage to the passage of the bill, I decided to take a look at the CBO’s report on what exactly the act has accomplished, in particular, the months of April through June.  There is a lot of great information released in this report.  Perhaps some of the most telling data is which fiscal tools was most successful and what Congress should use to attack high unemployment.  

Looking at the data, the act funded 550,000 jobs in the three-month span.  In the history of the bill, it has accomplished:

• Raised real (inflation-adjusted) gross domestic product (GDP) by between 0.8 percent and 2.5 percent,
• Lowered the unemployment rate by between 0.5 percentage points and 1.6 percentage points,
• Increased the number of people employed by between 1.0 million and 2.9 million, and
• Increased the number of full-time-equivalent jobs by 1.4 million to 4.0 million

These numbers provide some insight on the past performance of fiscal policy.  However, the CBO provides insight on which fiscal tools added the most to the gains in employment. 

Below is data for the multiplying effect of the contents of the stimulus package.  Any number over 1.0 is a positive gain:

 

This nifty table provided in the report shows which programs contributed most to economic growth resulting from the stimulus package.  This shows us what programs we can cut spending from, and where we can add investment.  As you can see from the right hand column, government investment in energy related research has a huge multiplier effect.  The same applies to Federal government money transfers to states for infrastructure.  What does not contribute to economic growth are tax benefits for high income earners.  The increase in the individual Alternate Minimum Tax exemption amount, and the first-time homebuyer credits, both which are designed for use among higher income individuals, has little effect on creating jobs and increasing economic growth.  The true growth comes from investment in infrastructure, and the middle and lower class.  Housing assistance, highway construction, education for the disabled, unemployment insurance, and clean water and drinking funds are some of the programs (which many are public goods which all people benefit from) that have helped grow the economy during the past few years. 

My thoughts: Lawmakers will be coming back after the summer recess and immediately address the issue of jobs (2 years late is better than never, right?...) .  There is empirical data that shows what can be used, and what can be done to improve the economy.  One of two two things will happen.  The first is Congress will attack the unemployment problem head on, invest in public goods that will benefit the people in many positive ways.  While at the same time, look at the tax code and other areas where there is useless spending, and make the country more efficient.  The second and more realistic scenario, Congress will have a bitter fight about what not to do, continue to fund tax credits that have no positive economic impact, and wait for full employment to return sometime after the next financial crisis.  The Republicans will have successful defeated Obama’s “jobs plan” and declare victory, while the Democrats call “compromise” victory, and we all lose.

HOORAY!

AAA real shake in confidence for S & P

A little over a week ago S & P downgraded US bonds to AA.  There is no doubt that you probably heard of all the imminent doom that is associated with downgrades. But I don't completely trust S & P's move, and I am not alone.  10 year treasury bills have dropped about 30 basis points since the downgrade, so it seems that investors do not worry too much about it much either.

This raises the question if we can trust S & P anymore.  Look at it like this.  The first half of the twenty-first century saw an enormous spike in bonds backed with mortgages. What investment firms did was combine the worst mortgages they could find, polish them off, bundle them together and sold them to investors as gold.  You would imagine that an agency that was giving them a rating would assume they are junk, and assign them the worst rating you can imagine.  However, these junk bonds were given the most secure rating, so everyone assumed they were a safe bet.  One of the major players in giving them that great rating was S & P.  When the market collapsed and it turned out all those bonds had inaccurate ratings, S & P looked bad.  I mean really bad.  One of the most trusted rating agencies, which gets a profit from rating bonds of the companies who they happen to be rating, could whiff so hard is a really bad look.  Only thing they can do is try to toughen up.  Stamp American bonds as AA, rather than AAA to make themselves look serious.

This doesn't take away the fact that when banks would send bonds over for ratings, and they did not understand them, they just shrugged their shoulders and trusted the bank.  For a better part of the decade, that is how the system worked.  Hundreds of billions of dollars flowed through these bad bonds and nothing was thought of it.  But now since S & P realized they messed up so bad, it is time for them to get serious.  Unfortunately for them, no one is really buying it. 

It's been such a long time...

Decided I am going to try and get this blog going again after a year hiatus.....but it is Friday so that may not happen till Monday.