Wednesday, September 18, 2013

U.S. News' College Rankings Are Terrible -- But It's Not (Completely) Their Fault

This post originally appeared in the Huffington Post:

U.S. News and World Report publishes its annual rankings of the best colleges and universities in America this month, and if past is prologue, the rankings will receive plenty of attention and criticism. College affordability advocates argue that the ranking incentivizes schools to build lavish new buildings that drive up tuition. Others point to the subjective nature of the reputation survey. Even President Obama recently criticized the rankings for "encourag[ing] colleges to game the numbers and rewards them in some cases for costs."

Despite all the criticism, the rankings' shortcomings aren't entirely U.S. News' fault. A school's reputation and money account for almost half (44.5 percent) of the rankings are proxies for information that just isn't available. Not only is it not available, but collecting and keeping it all in one place for students to shop around is actually illegal.

The 2008 reauthorization of the Higher Education Act specifically banned the government from tracking individual student education statistics and matching them to workforce outcomes. That it made it illegal for the federal government to collect crucial variables like the job placement rates, loan repayment rates, and average incomes from a particular major at a given school.

Imagine you're a high school senior (or a parent of one), and you're about to make the second biggest purchase of your life, deciding where to attend college. You're choosing between two schools: an applied engineering major at your state's flagship public university, or an elite private research and technology college halfway across the country. How many students like you graduated on time at either school? Which school produces more graduates that actually work in the field they studied? Which graduates are having an easier time paying off their loans? That's crucial information that would inform your decision.

The crazy thing is that the data is out there. Schools track how different types of students perform and the IRS and Social Security Administration track incomes. Those datasets just aren't talking to each other. Privacy advocates warn of the government abusing or leaking the information, but there has to be a way to protect students' privacy while simultaneously giving them the information they need to make an informed decision on where to attend college.

That's part of what our consortium of policy analysts, student and business advocates, and higher education representatives is trying to do. With support from the Bill and Melinda Gates Foundation, the Reimagining Aid Design and Delivery (RADD) project is working on ways we can improve college choice. Students need comparable information about institutions that offer them a quality education at an affordable price.

And students agree: In a 2012 survey conducted by Young Invincibles, student leaders ranked graduation and job placement rates as the most important factors in picking a school. Factors like academic prestige and the US News and World Report ranking were rated far lower. Researchers at the College Board and Harvard Business School found that students facing a flood of information in the Internet age reach for simplified metrics, like college rankings, to make decisions. So this new data should be presented simply too.

Our group will be designing policy recommendations, briefs, and events through March, getting the word out about the lack of helpful data in higher ed, and crafting what solutions could look like. Stay tuned.

Friday, July 26, 2013

Senate Deal Treats Students Like Cash Cows

This entry was originally posted on the Huffington Post

The Senate is voting on a deal to change interest rates on federal student loans, pegging the rates to the 10-year Treasury note. With market interest rates so low this year, students taking out loans will pay a lower rate than current law of 6.8 percent -- for now. But without strong enough protections from high rates in the future, the infographic below demonstrates how the deal will actually milk students in the coming years.

Thursday, July 11, 2013

Where are America’s Uninsured Young People?

This post was originally published at Young Invincibles at featured on Washington Post's Wonkblog:

In just three months, the first state health insurance marketplaces created by the Affordable Care Act will open, and millions of uninsured Americans will have the opportunity to purchase more comprehensive, more affordable coverage, often with the help of new tax credits.

With the deadline for enrollment looming, Young Invincibles is dedicated to making sure young people know their options and understand how to get covered.

Nationally, over 19 million 18 to 34 year olds are uninsured. Young people lack coverage more than any other age group group, making up 27% of the estimated 30 million uninsured nationwide.

That’s why we’re launching the Healthy Young America Campaign. We’ve done extensive research on insurance rates in every state of the nation and put together fact sheets for advocates to get the word out and make sure young people sign up for health insurance. We have a growing list of frequently asked questions and developed an app to help young people find coverage and health services.

To bring together all of our research, we created this interactive map of insurance rates among 18-34 year olds. How do the states stack up? Massachusetts has the lowest rate (not surprising considering they passed a health reform law years ago).

But for those looking to enroll as many people as possible, showing the raw numbers of uninsured youth is useful as well. California leads the nation with over 2.8 million uninsured young people.

Whether you’re a local non-profit looking to enroll young people your state, a state looking for outreach methods to reach their youth population, or a young person looking for information on your options, we want Healthy Young America to be a resource for you.

Tuesday, May 21, 2013

A Map of Deals for Developers

Developers have received $1.7 billion in subsidies and tax breaks  in the District in the last decade. Some of these projects are cash machines for the developers, with corporate tenants like Target at the DC USA shopping center in Columbia Heights, or homes for professional sports teams like the Verizon Center. That's a stunning pot of money for this little city, and to an industry that may not need that sort of public assistance.

I made a map of the projects that received the largest packages of subsidies:

You can't help but notice that only one of the top 20 subsidized projects is across the river, and that the majority are along the heart of a Northwest corridor that you wouldn't think would need any assistance to turn a profit.

Here are 20 projects that received the most public assistance, according to WAMU and the Office of Chief Financial Officer. "Deals for Developers" also looked at the campaign contributions the developers gave to city officials, but with strict regulations on campaign finance, I don't think that data set is particularly persuasive.

Subsidy Recipient
Subsidy Value
Total Donations
The Wharf
PN Hoffman, Paramount, E. R. Bacon, CityPartners, Madison Marquette, Gotham
Convention Center Hotel
Marriott, Quadrangle, Capstone, RLJ, Gould
US DOT (Anacostia Waterfront Projects)
JBG , Clark Enterprises
The Yards
Forest City
Gallery Place
Western Development, Akridge
Capper Carrollsburg
Forest City , WC Smith, Urban Atlantic, EYA
Verizon Center Renovation
DC Arena, Monumental Sports, Lerner Enterprises
NoMa residential development
Archstone, Cohen Companies, Stonebridge, Palmetto
Adams Morgan Hotel
Friedman Capital, Marriott, Beztak
Mandarin Oriental Hotel
Republic Properties, CityInterests
Grid Properties
National Public Radio, Inc.
Boston Properties, NPR
Skyland Shopping Center
WC Smith, Rappaport
City Market at O Street
Roadside Development
Living Social
Living Social
Capital Fire Station
CityPartners, Potomac Investment Properties, Paramount, DC Strategy Group, Adams Investment,
Hine Jr. H.S.
EastBanc, Stanton Development, L. S. Caldwell & Associates, Dantes Partners, Wilmot, The Jarvis Company
Kelsey Gardens Redevelopment Project
Metropolitan Development, CEMI, Akridge
Center Leg Freeway
Property Group Partners,

This was a very impressive piece of reporting from WAMU. They also provided a lot of data for other journalists to analyze and visualize. You should donate to them:

Wednesday, May 15, 2013

Wall St. is Growing Twice as Fast as Jobs

The Dow is on a tear!

Here's the Dow Jones Industrial Average from January 2008 to the present, including yesterday's new record of 15,215:

The market has recovered from its pre-crash high of 12,000, surpassing it by 25%. 

And here's BLS' unemployment rate, from the same time period:

The job recovery is there: last month's 7.5% is the lowest its been in Obama's presidency. But we're still 50% below where we were pre-crash. 

Tuesday, May 14, 2013

Company Town: See How Many D.C. Area Residents Are in Government Jobs

Originally posted in DCist on May 14, 2013:

Photo by Brian Mosley
Seems that nobody ever tires of pointing out that Washington is a "company town," dominated by a single industry and affording little room to any other interest. Of course, any D.C. resident knows that the city and its surrounding communities are more diverse than that, but the appellation fits. Los Angeles has entertainment, San Francisco has technology, Detroit had cars, and we have government.
Whether federal, state, or city, no era of sequestration will dislodge the public sector as the Washington region's most dominant source of employment, but not everybody's neighbors are working for the man. Analyzing figures released by the U.S. Census Bureau, we can see which neighborhoods are most dependent on government, and where other industries play a larger role.
The maps below were made for DCist by Tom Allison, who previously offered a crowdsourced Google mapplotting rooftop bars around town. For this set, Allison analyzed employment data in the Census Bureau's American Community Survey.
The American Community Survey supplements the decennial census that all Americans complete. Each month, roughly 250,000 residents are sent a lengthy questionnaire asking information that was previously only solicited from people lucky enough to receive the old "long-form" version of the census. The bureau analyzes the resulting data to extrapolate trends about life in the United States.
Map by Tom Allison
Using the survey's five-year estimates for responses pertaining to employment, Allison applied that data to a map of the greater D.C. area, including Montgomery and Prince George's counties in Maryland, and Alexandria, Arlington, and Fairfax County in Virginia. It's hardly a surprise that the census tracts—roughly analogous to established neighborhood boundaries—around local military installations are the deepest shades of orange and red. Navy Yard, in Southeast D.C., is as red as it gets, with more than 50 percent of residents on the government payroll. Tracts around the Pentagon are consistently yellow and orange, while Fort Belvoir, Va. is very red, too.
The northeastern section of Prince George's County also features a large red splotch. No surprise there, as it surrounds the University of Maryland. Over 13,000 people work full- and part-time at the university, making it one of the state of Maryland's biggest employers.
Higher concentrations of government workers can also be found in D.C. wards 5, 7, and 8, and much of Prince George's County is solidly yellow, indicating that between 30 and 39 percent of residents in those zones are government workers. Those wards and Prince George's County have majority black populations, and there is a demographic correlation between public sector employment and ethnicity. A 2011 study at the University of California, Berkeley found that between 2008 and 2010, 21.2 percent of all black workers across the United States were government employees, compared with 16.3 percent of non-black workers.
But the racial breakdown is more easily set aside in the D.C. area. The Census Bureau's definition of what constitutes a government worker encompasses "employees of any local, state, or Federal governmental unit, regardless of the activity of the particular agency." And that includes occupations beyond direct public administration, such as health or transportation.
The greater D.C. area has no shortage of governments, intergovernmental agencies, and transportation authorities to choose from, and Allison's maps, both for the percentage and raw number of government workers, are a colorful reminder of the fact that this region is dominated by the public sector. As Mayor Vince Gray or any other city official will gladly point out, nearly 60 percent of D.C.'s workers are in the private sector, with industries like hospitality, education, construction, and technology leading the pack.
But on Allison's maps, most of the D.C. is still a sea of yellow and orange, reflecting that in those census tracts, public-sector employment is between 11 and 29 percent. That's not a majority, of course, but it's still far more emphatic than an analysis of anywhere else. A map of the entire United States would show an ocean of green. Nationally, 14.9 percent of workers are in government jobs, according to the same five-year estimate.

Wednesday, May 8, 2013

Two Maps That Explain VA's Changing Electorate

Cartograms are an effective way to project data, especially when the values of a given variable are extremely disproportionate. You've probably seen cartograms before: the area of polygons are distorted to a ratio of some other variable. They've been around since the 19th century, and a classic one is the U.S. states distorted to reflect their populations.

Dr. Stephen Farnsworth, a former professor of mine at the University of Mary Washington, sent me some very interesting cartograms that visualize his analysis of population change in Virginia, a favorite topic here at Map Attacks. Dr. Farnsworth and his colleague across the hall in the geography department, Dr. Stephen Hanna, published the report on the changing Virginia lanscape electorate in the Virginia News Letter last week.

Here's a typical, static map of Obama's performance by county in Virginia in 2012.

Fine. Makes sense. Democrats in the D.C. suburbs and southeast, Republican strongholds in southwest and pretty strong in the central areas except for Richmond. But Virginia has two Democratic Senators and Obama won twice. The solution to this cognitive dissonance of course is that not all counties are created equal. So redraw the area of the county polygons to reflect population, and you get a much better idea of what's going on:

An even stronger visual of what's going on in Virginia politics is seen here, in a cartogram of the increase in percent of votes cast for each candidate:

Stay tuned for some new cartograms about Washington: Crime, population size, and median income will be the usual suspects, but if there's anything you'd like to see, especially something that might tell a different story than just "Northwest is rich and safe and Southwest is poor and dangerous," let me know and I'll do my best to accommodate.

I also recommend everyone check out Hanna and Farnsworth's full report here:

Monday, April 29, 2013

Mapping Z Scores

Z Scores can be an easy and effective way to organize and analyze a dataset with problematic measures of central tendency. Z scores, also known as standard deviation units, demonstrate how far away (positively or negatively) any given data point is away from the mean of that data set. You could just show the difference between the data point and mean, in the original units, but that might not give you as clear a picture of how just how much that point differs from the central tendency.

Take population growth by Virginia county for instance: According to the American Community Survey's 2012 data, most counties are adding people, but some are losing residents. In fact, there's a huge range in population change. Henry County, for instance, has lost 1,185 residents in between 2009 and 2011. Fairfax, in northern Virginia, gained over 36,000 people.

We have also have a significant positive skew here. Look at how much larger the mean is than the median in the below SPSS output:
In fact, only a handful of counties greatly affect the central tendency of the dataset (Surprise! They're in Northern Virginia). The graph below shows how three counties spike far above the generally stable cluster near the x axis.

So the mean and median isn't helping us very much figuring out population change in Virginia. The "average" county isn't adding 1,379 people annually, as the mean would suggest.

That's where measures of dispersion help a lot. SPSS already gave us the standard deviation, so we just take subtract the mean from each value and divide by the standard deviation, and we have Z scores for each county.

Join those Z scores to a shapefile of Virginia, classify under natural Jenks breaks, and voila:

I've been doing a lot of thinking about the Virginia Gubernatorial Race and the changing demographics of the state. And looking at a map like this demonstrates that Democrats really don't need to compete throughout the state in statewide races. They can win by continuing to shore up Northern Virginia, and will continue to with these population changes. Not a bad plan, unless you want to win back the state senate.

Map of D.C.'s Rooftop Bars

Some folks over at the D.C. Reddit group have been trying to compile a complete list of the city's rooftop bars, and someone asked for a map. So here you go:

Friday, April 12, 2013


If you got a job last month, congratulations. You were one of a very lucky 0.75% of people who wanted one too. Not as bad as the lottery, but not a recipe for prosperity. 

A Lost Economic Generation?

Young people have it pretty rough in this country right now. Most age groups and demographics do too, but the statistics are worth taking a closer look at:

  • Median net worth of people under 35 fell 37% between 2005 and 2010; those over 65 took only a 13 percent hit.
  • The national unemployment rate is 7.6%; For people between 20-25, it's 13.3%. For 18-19 year olds, it's 22.1% (remember these percentages aren't out of the total population, but rather the portion of the age group that are actively seeking work but can't find it.  And as this post from Young Invincibles points out, that $4.25 an hour internship or 20 hour a week bartending gig means, that in the eyes of the Bureau of Labor Statistics, you are very much employed.)

Why should we care? Aren't the young less likely to have obligations like mortgages, children, and aging parents? Take a look at this chart that shows how rising costs in education and health care, combined with sluggish earnings trend, is putting extra financial pressure on the nation's young people. 

As the gap between the top green line and the other three shrinks, young people will delay moving out on their own, making major purchases, and also be more willing to work those low paying jobs that America's hirers are more than happy to keep them in. 

For each year lost to unemployment, not only is that a year of forgone cash earnings, that are very likely to be spend to spur the economy as a whole, but it acts a drag on future earnings as well. Even with a more significant economic recovery, the lack of gained experience and skills could amount to a $20 BILLION LOSS  in earnings over the next decade. Of course that's lost taxable income as well, so advocates protesting the President's proposed cuts to safety net programs should probably consider advocating investing in the generation that's responsible for the tab of those benefits. 

Friday, April 5, 2013

What D.C. Neighborhoods Could Use a Metro Station?

The D.C. Metro system is perilously close to capacity, with supply of trains and stations already not meeting demand. Total trips are down slightly this year, but population growth projections as well as growing trends in area residents preferring mass transit to owning and driving their own cars, has spurred the Washington Metro Area Transit Authority to plan expanding the Metro system. A new 49 page strategic plan calls for building new tunnels through the city, a new tunnel under the Potomac River, and a pedestrian walkways between Metro Center and Gallery Place, and Farragut West and Farragut North, and a new station for the Pentagon.

Combine this expansion with the Silver Line and street cars, it's an exciting time for D.C. mass transit. So H Street and Tyson's Corner get their new transit assets, but what about other areas of D.C.? Particularly, what high density residential areas aren't currently served by Metro?

The D.C. Data Catalogue has existing landuse data on every block in the city. By projecting medium and high density residential areas, we can use ArcMap's measuring tool to see what neighborhoods are situated a mile or more from existing Metro stations. 

How about a Gold Line starting in Mclean (a future silver line station), continue northeast to Glover Park and connect to Van Ness on the Red Line, then serving Brightwood, before turning South and connecting to Georgia Ave, a new Crestwood Station, and continue on down to the New York Avenue station.

On its way Southeast, this Gold Line could also service the burgeoning H Street corridor with its connections to the proposed streetcar, before crossing the river and serving multiple neighborhoods in Southeast. The city is more connected, congestion declines, Mayor Gray's "One City" vision realized...

One can dream...

Monday, April 1, 2013

What a 50 Mile Evacuation Zone Means For Virginia and D.C.

It's been two years since the 2011 tsunami and 9.0 earthquake that devastated Japan and wrecked the Fukushima nuclear power plant. Fortunately, the World Health Organization released a report this month concluding that negative health effects from the released radiation would be minimal or not observable.

Louisa County, Virginia, near where I grew up and the family farm still operates, was the epicenter of a smaller 5.8 magnitude earthquake in August 2011. Near by are two nuclear reactors at Lake Anna. After the quake the plant's safety measures worked, and no radiation was emitted. But according to the Nuclear Regulatory Commission, the Anna reactors face an annual 1 in 22,727 chance of the core being damaged by an earthquake and exposing the public to radiation. 

Dominion Power, the company that operates the plant, said that Lake Anna was built to withstand a 6.2 earthquake. The Richter scale is logarithmic, meaning that increases at the top of the scale are more dramatic than at the bottom. Even so, it would seem that the 2011 quake was cutting it close to Lake Anna's limits. 

Two other nuclear plants in the area could also pose a risk in the event of a strong earthquake, Surry Nuclear Power Plant in the Tidewater region and Calvert Cliffs in Maryland, southeast of Washington D.C. 

Drawing a 50 mile radius around each plant demonstrates the cultural, national security, and of course population assets that would be caught in the evacuation zones.

Why 50 miles? Transcripts released last year revealed that the Nuclear Regulatory Commission (NRC) thinks that a 50 mile evacuation zone would be appropriate for a disaster along the lines of Fukushima. Then NRC Chairman Gregory Jaczko also urged Americans living within 50 miles of the plant to evacuate. It turned out that the cooling pools had not completely evaporated as previously thought, but the comments raised alarm in energy and environmental community.

Below is a national look of all the nuclear reactors in the country, with a 50 mile buffer shaded around them.

Key Takeaways:
  • Earthquakes pose a threat to our nuclear power plants, most of which were built when Jimmy Carter was president
  • Many environmentalists say it will be impossible to cut emissions from fossil fuel sources to combat climate change without expanding our nuclear profile
  • After Fukushima the body in charge of regulating domestic nuclear energy production re-evaluated their evacuation zones to fifty miles. Highly valuable national assets are within these fifty mile zones. 
More questions than answers. But laying out the problems ahead is the only way to make well-informed decisions.