Tag Archives: data source

Is there a relationship between education level and marriage?

The U.S. Bureau of Labor Statistics post 60 percent of college graduates born from 1980 to 1984 were married at age 33 (6/30/2020) provides the chart copied here.

At age 33, people with higher levels of education were more likely to be married and less likely to be cohabiting than those with lower levels of education. At the time of their 33rd birthday, 32 percent of high school dropouts, 42 percent of high school graduates with no college, 49 percent of people with some college or an associate degree, and 60 percent of college graduates were married. Twenty-eight percent of those with less than a high school diploma were cohabiting, compared with only 13 percent of those with a bachelor’s degree and higher.

For related data, see the Census Bureau Educational Attainment in the United States: 2019 Table 2.  For people 25 years and over no more than 53% of those that didn’t complete at least an Associates degree here married. For those that have a Bachelor’s degree, Master’s degree, Professional degree, and Doctoral degree the percentage are 63%, 68%, 71%, and 73% respectively.

How has Black educational attainment changed?

The Census Bureau post Black High School Attainment Nearly on Par with National Average  by Jennifer Cheeseman Day (6/10/2020) notes:

In 1940, when the U.S. Census Bureau started asking about educational attainment, only 7% of Blacks had a high school education, compared with 24% for the nation as a whole.

In recent years, Black educational attainment has been much closer to the national average and today, 88% of Blacks or African Americans have a high school diploma, just shy of the national average, according to census data released last month from the Current Population Survey.

Related to the graph copied here:

The national average dropout rate declined from 19% in 1968 to about 6% in 2018. The Black dropout rate fell more steeply from 33% to 5%, bringing it in line with the national average.

Average enrollment for young adults increased from 26% to 41%. At the same time, the proportion of Black young adults in college more than doubled, rising from 15% to 38%.

The article contains five other graphs and links to the Census Bureau data sources.



How hot was May 2020?

From the NOAA Global Climate Report – May 2020:

The global land and ocean surface temperature for May 2020 tied with 2016 as the highest in the 141-year record at 0.95°C (1.71°F) above the 20th century average of 14.8°C (58.6°F). The 10 warmest Mays have all occurred since 1998; however, the 2014–2020 Mays are the seven warmest in the 141-year record. May 2020 also marked the 44th consecutive May and the 425th consecutive month with temperatures, at least nominally, above the 20th century average.

The global land-only surface temperature for May 2020 was also the highest on record at 1.39°C (2.50°F) above the 20th century average of 11.1°C (52.0°F). This was 0.04°C (0.07°F) above the previous record set in 2012. The 10 highest May global land-only surface temperature departures have occurred since 2010.

The May 2020 global ocean-only surface temperature was near-record warm at 0.79°C (1.42°F) above average. This value was only 0.01°C (0.02°F) shy of tying the record warm May of 2016.

May time series data here. Climate.gov provides a summary of May 2020 in their post Was May 2020 warm and dry or cool and wet across the U.S.? It depends… by Rebecca Lindsey (6/9/2020)

What is the connection between crime and lead?

Kevin Drum asks a good question in his post How Many Cops Does New York City Need?  First note that violent crime has been dropping since around 1990 (see his graph copied here for examples). In particular for NYC:

The per capita number of police officers increased by about 10 percent through 2000 and then declined by about 20 percent through 2018. That’s nearly flat over the entire period. Violent crime, by contrast, plummeted 60 percent from its peak in 1990 through 2000 and then declined another 40 percent through 2018. That’s a total decrease of nearly 80 percent between 1990 and 2018.

Violent  crime has decreased even though the per capita number of cops has been nearly flat. So, why did crime decrease? There is overwhelming evidence that removing lead emissions from cars is the main driver of crime decline. I strongly encourage you to read Drum’s 2018 summary of the evidence.

So, why doesn’t anyone talk about lead and crime?

The second problem is among activists on both left and right who have their own pet theories. On the left, we tend to blame poverty, institutional racism, poor schooling, lousy housing, and so forth. On the right, the favorite targets are the breakdown of the family, too few cops, too few prisons, drugs, the decline of religion, and so forth. There is very little convincing evidence for any of this, while lead poisoning explains everything. But if lead poisoning is the answer, then everyone has to give up their pet theories about what happened between 1960 and 2010. That’s a tough ask.

We forget that there was a lot of violent crime in the 1980s. It had an impact on society in many ways. But, we are past that and removing lead from the environment is a permanent fix to the violent crime wave of the past. This should allow us to think differently about societal needs for policing.

The Drum post has two other graphics. The Statistics Projects page has the relevant lead and crime data.


How has Covid-19 impacted unemployment by race?

The chart here comes from using FRED (Federal Reserve Economic Data). Since at least the 1970s Hispanic or Latino (using FRED terms) unemployment was consistently between Black or African American and White and more recently slightly closer to White unemployment. For possibly the first time since the 1970s Hispanic or Latino unemployment (18.9%) exceeded Black or African American (16.7%) in April 2020. The Feb 2020 to April 2020 increase in unemployment for the four groups in the chart are (in order from smallest to largest) Black or African American (10.9%), White (11.1%), Asian* (11.8%), Hispanic or Latino (14.5%). It would seem that by both the total increase and the magnitude of unemployment that the Hispanic or Latino population was hit hardest by Covid-19. Moving to May, while Hispanic or Latino unemployment has decreased, along with White unemployment while Black or African American is stable and Asian increasing, they still exceed the other three groups.

The link here to FRED is only for the graph of Black or African American unemployment. Use the Edit Graph button (top right) and then Add Line (middle top tab). Search for the other groups and add them to the chart. The chart will provide data starting in 1972. The graph is interactive and the data is available.

*Asian Unemployment numbers are not seasonally adjusted while the other three are – FRED didn’t have seasonally adjusted for Asians or I couldn’t find it.

How many jobs from 18-32?

The U.S. Bureau of Labor Statistics post People born in early 1980s held an average of 8.2 jobs from ages 18 through 32 (6/3/2020) includes the graph copied here. They note

Women with higher levels of education held more jobs than women with lower levels. Women with a bachelor’s degree held 8.8 jobs from ages 18 through 32, compared with 6.5 jobs for female high school dropouts. Men held a similar number of jobs regardless of their level of education.

People held an average of 4.5 jobs from ages 18 to 22. The average number of jobs dropped to 3.3 from ages 23 to 27, and then dropped more, to 2.3 jobs, from ages 28 to 32. The pattern of people holding fewer jobs as they aged was similar among women and men and across racial and ethnic groups and levels of education.

The chart data is available and there are links to the original survey.

Do we use more coal or renewable energy?

The EIA article U.S. renewable energy consumption surpasses coal for the first time in over 130 years by Mickey Francis (5/28/2020) has the data.

In 2019, U.S. annual energy consumption from renewable sources exceeded coal consumption for the first time since before 1885, according to the U.S. Energy Information Administration’s (EIA) Monthly Energy Review. This outcome mainly reflects the continued decline in the amount of coal used for electricity generation over the past decade as well as growth in renewable energy, mostly from wind and solar. Compared with 2018, coal consumption in the United States decreased nearly 15%, and total renewable energy consumption grew by 1%.

The obvious question is are we using less electricity and if not what other source is picking up the slack? The eia Electricity explained page has an interactive time series of electricity generation.  Natural gas has picked up the slack.

The first article has two other charts and links to data.


How hot was April 2020?

For those of us living in the northeast or a good part of the U.S. we might have felt that April was cold and it was. It is easy to use that as evidence that climate change is “fake news” yet it is good to keep in mind that if it is cold where you are it is likely much warmer somewhere else. The map here is from NASA’s GISS Surface Temperature Analysis page where similar maps can be made for a variety of time periods.  Here we can see that for April 2020 parts of the U.S. and Canada where one of the  few cold spots in the world. The rest of the planet was warmer.

NOAA’s Global Climate Report – April 2020 notes:

Averaged as a whole, the global land and ocean surface temperature for April 2020 was 1.06°C (1.91°F) above the 20th century average of 13.7°C (56.7°F) and the second highest April temperature in the 141-year record. Only April 2016 was warmer at +1.13°C (+2.03°F). The eight warmest Aprils have occurred since 2010. April 2016 and 2020 were the only Aprils that had a global land and ocean surface temperature departure above 1.0°C (1.8°F).

Time series data is available on the NOAA page. Note that NASA uses 1951-1980 as their baseline while NOAA is using the 20th century. This accounts for the slight differences in their calculations on April’s anomaly from the baseline.

But it doesn’t look like sea levels are rising?

The NASA article Can’t ‘See’ Sea Level Rise? You’r looking in the Wrong Place by Alan Buis (5/13/2020) combines the quantitative facts of sea level rise with stories of places feeling the impact.

“Thanks to satellite and tide gauge data, we know that sea level is rising about 3.3 millimeters (0.13 inches) a year, a rate that grows by another 1 millimeter (0.04 inches) per year every decade or so,” Willis said. “Each year, global warming is currently adding about 750 gigatonnes of water to the ocean – enough to cover my home state of Texas about 1 meter (more than 3 feet) deep. We can’t really eyeball a few millimeters of sea level rise a year just by looking at the ocean because of waves, tides, etc. But we can definitely see the effects of it, both short- and long-term.”

Ocean Isle Beach NC:

I passed a woman walking her dog and asked her about the homes. “There used to be two streets of houses in front of these homes,” she told me. “Now they’re oceanfront.”

Norfolk, VA:

Over the past couple of decades, high tide flooding here has accelerated rapidly, and now occurs about 10 days a year, causing flooding in downtown Norfolk.

Sea Level data from NASA’s Sea Level Page.

How should we measure COVID-19 deaths?

The CDC’s new webpage Excess Deaths Associated with COVID-19 provides one method to measure pandemic related deaths:

Estimates of excess deaths presented in this webpage were calculated using Farrington surveillance algorithms (1). For each jurisdiction, a model is used to generate a set of expected counts, and the upper bound of the 95% Confidence Intervals (95% CI) of these expected counts is used as a threshold to estimate excess deaths. Observed counts are compared to these upper bound estimates to determine whether a significant increase in deaths has occurred. Provisional counts are weighted to account for potential underreporting in the most recent weeks. However, data for the most recent week(s) are still likely to be incomplete. Only about 60% of deaths are reported within 10 days of the date of death, and there is considerable variation by jurisdiction.

The interactive graphics allows the user to choose a jurisdiction and different data types. The graph here is for the U.S. and weekly excess deaths. All data can be downloaded as a csv file.