Category Archives: Data

UConn-Notre Dame, Minute-by-Minute

by Categorized: Data, Sports, UConn Date:

A nail-biter ’til the end – and then a heart-breaker for Huskies fans.

UConn-South Carolina, Minute-by-Minute

by Categorized: Data, Sports, UConn Date:

It was a close game – for 19 seconds. But after that, it was all UConn, as the women steamrolled over South Carolina to advance to the Final Four for the 11th consecutive season.

Surprise! People are Moving Out of Connecticut

by Categorized: Data Date:

United Van Lines, which moved about 110,000 people from one state to another last year, keeps tabs on where folks are going and where they are leaving. And it will come as a surprise to no one that more people are packing in Connecticut than unpacking in the state.

Connecticut ranked fourth among states with the highest percentage of outbound moves, at just over 57 percent, meaning roughly four people moved out for every three who moved in. More specifically, United had 1,229 customers move into the state, and 1,637 move out.

Only New York, New Jersey and Illinois had more lopsided numbers. Illinois, at the top of the list, had 63.4 percent of relocations occuring with the state in the rearview mirror, meaning there were nearly two outbound moves for every inbound move.

At the other end of the scale, Vermont was the most popular of the 48 contiguous states, with more than two inbound moves for every outbound move. Other popular states: Oregon, Idaho, Nevada and South Dakota – each of which had more than three inbound moves for every two outbound moves.

In Connecticut, 55 percent of inbound moves were made for job reasons, compared with 40 percent of outbound moves. About 9 percent of new residents cited retirement as the reason for the move, compared with 24 percent of movers leaving the state.

The data, of course, are not a complete picture of interstate relocation, capturing only those moves involving one professional mover. That could skew the data by over-representing wealthier movers – overstating the number of moves out of wealthier states.

But the numbers should still set off alarms in those states with significant imbalances – including Connecticut, which has been on the high end of the moving company’s list every year since 2011.

Click the map below to see data on each state’s moves. States shaded green had a net inflow of movers, with bright green states having at least 55 percent of their moves inbound. Red states had a net outflow, with bright red states having at least 55 percent of their moves outbound.

UConn-Mississippi State, Minute by Minute

by Categorized: Data, Sports, UConn, Uncategorized Date:

It took the UConn women more than half the semi-final game to find their mojo and produce 20-plus minutes of nail-biting competition. But Mississippi State met them basket-for-basket, and Morgan William’s buzzer beater in overtime ended the Huskies’ legendary 111-game winning streak – and created a new legend of her own. Read Paul Doyle’s on-the-ground account here.

Hover over the chart below to see the minute-by-minute score. And use the scroll bar or arrows in the bottom chart to see how individual players performed during the 45 minutes of play.

How Much Did Your Vote Count in the Presidential Election?

by Categorized: Data, Government, Politics Date:

With this year’s divisive election leading to a divided result – with Hillary Clinton winning the popular vote and Donald Trump winning the presidency – calls are rising once again to scrap the electoral college system.

Some of this is Monday-morning griping; had the results been reversed, it would surely be Trump’s supporters suddenly aghast at the way we’ve elected presidents since the ratification of the 12th Amendment in 1804. But it is undeniable that the electoral college system is a departure from the traditional one-person-one-vote philosophy that typically guides democracies.

How big a departure? To quantify that, we measured the relative value of each state’s voters in this election by calculating how many votes for the winning candidate were required to procure each electoral college delegate.

Three factors affect the weight of a particular state’s voters:

Voters in small states have an advantage – because delegates are assigned based on the number of each state’s representatives and senators in Congress. Representatives are apportioned based on population, but every state has two senators regardless of  population. So, for a tiny state with a single representative, adding two more delegates triples the number of electoral college votes for that state – thereby tripling the potential power of its voters. But in a huge state such as California, with 53 representatives, the addition of two more delegates increases its electoral college power by only a few percent.

Voters in states with low turnout relative to population have an advantage – because the fewer votes cast, the more weight each individual vote carries. The electoral college delegates are assigned based on a state’s total population. So in states with low voter turnout – either because of below-average voter eligibility or below-average voter registration or simply because residents didn’t go to the polls – those who do cast ballots carry more weight than voters in high-turnout states.

Voters in states where the margin of victory was narrow have an advantage – because the victor in each state (with a couple exceptions), wins 100 percent of that state’s delegates whether that candidate swept 80 percent of vote or won with a tiny plurality. So the closer the race, the smaller the number of votes for the winner. And the smaller the number of votes, the more weight each vote carries.

Those factors can move in different directions and occasionally cancel each other out. In Texas, for example, the state’s large population weakened the value of each vote. But the state also had low voter turnout, increasing the value of each vote. The result: one vote for Trump in Texas carried almost exactly the same weight as the average of all winning votes across the country. And California, despite being the most-populous state and producing a larger than average margin of victory for Clinton – factors that pushed down the weight of each vote – had such low voter turnout that each individual vote for Clinton ended up carrying 22 percent more weight than the national average.

But while there is variability among the factors, a state’s small size is clearly the most significant in boosting a state’s relative power. The three states whose voters carried the most weight in the electoral college all have a single representative, but three delegates: Alaska, Wyoming and Vermont. Alaska also had a smaller than average margin of victory for Trump and a smaller than average voter turnout, making its voters the most heavily weighted in the election, with each vote for Trump counting for 2.8 times the electoral-college power of the average vote nationwide.

At the other end of the scale, Clinton voters in Massachusetts – with high voter turnout, a higher-than-average vote spread and a slightly above-average population – carried the least electoral-college weight, with each vote counting for about 70 percent of the average power of votes nationwide.

Put another way, in Alaska, one electoral college delegate was assigned for roughly every 43,500 votes for the winner, while in Massachusetts, it took about 178,600 winning votes to produce one delegate.

The map above shows the relative weight of votes for the winning candidate in the 50 states and the District of Columbia. States shaded progressively greener had higher weights; states shaded progressively redder had lower weights. Click on a state to see its winner and the relative value of its votes for the winner. There’s a clear geographic split, with most Western states being over-weighted and most Eastern states under-weighted, with the exception of small New England states. That is primarily a reflection of most of the Great Plains states.

In Reliably Blue Connecticut, a Shift Toward Red

by Categorized: Data, Politics Date:

It was no surprise that Hillary Clinton won Connecticut’s seven electoral votes Tuesday, but looking under the hood of that easy victory shows a clear shift to the right for most – but not all – of the state.

Here’s a town-by-town breakdown of Tuesday’s vote for president, with greater support for Clinton shown in progressively bluer shades, and greater support for Donald Trump shaded progressively in red. Click on a town to see its vote totals. (Towns in yellow were late reporting complete results.)

As expected, the major cities were solidly blue, as were the ring towns around Hartford, most shoreline towns and the northwest corner. And there was no shock in Trump’s edge in a swath of towns from Trumbull in the south to Hartland in the north.

But there were surprises Tuesday. Trump’s dominance in much of Eastern Connecticut was a dramatic improvement over Mitt Romney’s showing four years ago. And on the flip side, Clinton fared much better than Barack Obama in Fairfield County, and also had stronger support in the Farmington Valley. It may be that while both Romney and Trump are extremely wealthy, Trump’s populism resonated in rural towns far better than Romney’s silver-spoon persona. And Clinton’s friendliness toward Wall Street could explain her strong showing in the Gold Coast.

To gauge how broad Trump’s support was in small towns this time around, here’s a clickable map showing the Obama-Romney race in 2012, where nearly the entire eastern half of the state either favored Obama or gave Romney only a slim margin.

To fully appreciate the depth of the swing, this map shows the vote shift between the two elections, with redder towns showing a stronger shift to the Republican candidate and blue towns showing a strong shift Democratic. In many eastern towns, Trump had a double-digit percentage advance over Romney’s 2012 support. But in Fairfield County, Clinton enjoyed an even bigger swing in votes, helping her to win a majority in all the southern Fairfield County towns that have reported.

Since 1997, More the 1,000 U.S. Officers Have Been Murdered

by Categorized: Data, Law Enforcement, Public Safety, Uncategorized Date:

The policemen gunned down in Dallas Thursday night join a long and sad list of officers intentionally slain year-after-year in the line of duty. From 1997 to 2015, according to FBI statistics, 1,005 members of U.S. law enforcement members were “feloniously killed” – and a surprisingly large number died in ambushes.

Over the last two decades, officer murders peaked in 2011, when 72 were slain. Preliminary numbers for 2015 are far lower: 41.

Those numbers are a fraction of all law-enforcement deaths. Intentional killings account for about a third of the officers who die in the line of duty, with many more dying in car accidents and from job-related illnesses. The overall numbers have dropped dramatically from spikes in violence against police in the 1920s and 30s, and the 1970s.

While deadly encounters during traffic stop may seem to be the most common scenario in which officers are killed, they account for fewer than one in five slayings. Since 1997, more officers were killed in ambushes – including unprovoked attacks and cases in which officers were lured into danger – than during traffic stops. About one in 13 killings occurred while police were responding to a domestic dispute.

Firearms were used in 92 percent of all murders of police since 1997. Handguns were used most often, although rifles and shotguns were used far more often with cop killings than with homicides in the general population. In recent years, about 7 percent of firearm homicides were committed with rifles or shotguns. Among killings of police, long guns were used more than four times as often.

Connecticut State Budget, Fiscal 2017

by Categorized: Data, Finance, Government, Politics Date:

Legislators are preparing to vote on a nearly $20 billion budget for the next fiscal year – a revised spending plan that trims more than $800 million from what legislators approved 11 months ago. Here’s a breakdown of the proposed budget by department and by line-item description. Hovering over the bubbles will display each department’s total budget. Clicking on a department will show line-item spending for that agency.

FOIC Rules Teacher Rating Data is Public. Now Let’s Open Up Individual Evaluations.

by Categorized: Data, Education, Transparency/FOI Date:

Over the objections of teacher unions, the Freedom of Information Commission ruled Wednesday that there’s no reason to keep secret the fact that New Milford school officials think about three-quarters of their teachers are accomplished and one quarter are exemplary.

If that doesn’t sound like the sort of information that calls for breaking out the lawyers to keep under wraps, welcome to the strange world of teacher evaluations – the only public-employee evaluations in Connecticut that are kept confidential by law.

Thirty years ago, the legislature cut off public access to “records of teacher performance and evaluation,” swallowing the argument that without secrecy, parents would engage in “teacher shopping” by poring over personnel files and then harassing principals to pair their kids with the “good teachers.” I don’t know how silly that argument sounded three decades ago, but every time I mention it in casual conversation these days, the universal response is, “But everyone already knows who the good teachers are.”

Nevertheless, it’s been the law since 1984 that the public shalt not see a teacher’s performance review (and by “teacher,” the legislature decided that term should apply to librarians and reading specialists and assistant principals and every other certified school professional except the superintendent.)

But the Commission Wednesday permitted a tiny peek under the tent, ruling unanimously that the law protecting records of teacher performance was intended to apply to an individual, identifiable teacher’s evaluation, and not to anonymous, aggregated data for an entire district.

That is a victory for John Spatola, who was turned down when he asked the New Milford school system to release data related to the state’s far-from-perfect effort to uniformly assess the competency of the state’s teaching force. The state plan established a variety of criteria that placed teachers into one of four categories; exemplary, proficient, developing and below standard. (New Milford substituted the term “accomplished” for “proficient.)

So Spatola, a former member of the New Milford Board of Education, asked for data showing the number of local teachers who were placed in each of the four categories. New Milford officials were reluctant to release the data without some assurance that they could do so without breaking the law. The local and state teachers’ unions, meanwhile, intervened and urged the FOI Commission to keep a cloak on the numbers, saying their release would be illegal and would harm teachers.

The legal argument is built on the broad language of the statute, which cuts off public access to “any” record of teacher performance and evaluation. The unions argued that that language covered the aggregated data Spatola was seeking. The Commission disagreed, adopting a hearing officer’s finding that the law was meant to shield records related to individual teachers. “The requested records, which do not identify individual teachers or individual schools,” the hearing officer wrote, “cannot be used for ‘teacher shopping.’ ”

While Spatola hails the ruling as a triumph for transparency, it was somewhat academic; New Milford’s numbers, and those of many other school districts, had already come to light when they were entered as an exhibit in an ongoing lawsuit against the state Department of Education. But the data that came out of that case was heavily redacted, with some information missing for two-thirds of the state’s districts, and all information missing for nearly half.

Given the FOI Commission’s ruling, I’ve sent a request to the State Department of Ed for an unredacted copy of the data. I’ll keep you posted.

In the meantime, isn’t it time for the legislature to remember that it’s the public who ultimately employs the state’s public-school teachers? And to rethink a law that prevents the public from seeing performance evaluations for some of its most important employees?

UConn-Syracuse, Minute-by-Minute

by Categorized: Data, Sports, UConn Date:

A 16-0 Syracuse run late in the third quarter added a dash of excitement, but overall, the UConn women dominated on their march to a fourth consecutive NCAA championship.

Use the scroll bar or arrows below to see how the individual players performed during the 40 minutes of play.