Category Archives: demographics

Teens, Young Adults on Different Pandemic Job Recovery Tracks

U.S. teens are leading the coronavirus pandemic recession job recovery. But why?

Conor Sen runs through a number of contributing factors that have led to this remarkable outcome:

What makes teenage employment useful to study right now is that teenagers are less affected by the factors holding back labor supply than any other demographic. If they lived at home with their parents, they weren’t eligible for economic impact payments. If they were full-time students, they’d be ineligible for unemployment insurance, making enhanced benefits a nonfactor. They’re unlikely to be parents squeezed out of the labor force by closed schools or a lack of child care. They’re obviously not older workers who may have accelerated retirement plans during the pandemic. And teens were less likely to get seriously ill from Covid-19, and so perhaps less likely to avoid working for health-related reasons.

Barry Ritholtz offers a competing theory:

In 2007, before the great financial crisis, the national minimum wage level was a paltry $5.15. This was not all that long ago. For a teenager with even the most modest withholding / FICA, their take-home is so small it’s not worth it to work. You can see that in the trends over the preceding decades. By most measures — productivity, profitability, inflation, exec comp — the minimum wage has lagged badly. Teens did the math, and said WTF, why bother?

But the minimum wage began to rise during the financial crisis despite skyrocketing unemployment. It was raised in 2008, and then in 2009, and again in 2010. Post GFC, it’s been $7.25 an hour.

Not coincidentally, at exactly that time, the labor participation rate of teenagers began trending upwards. Today, it’s even higher than it was before the pandemic began. Maybe it’s boredom, perhaps some teens just want out of the house where they’ve been stuck with mom and dad and their siblings during the past year.

Or just maybe, local employers are raising wages sufficiently to make summer jobs attractive to teens.

That is an interesting hypothesis and one we can easily investigate. Starting with the Bureau of Labor Statistics' 2020 report on the characteristics of minimum wage workers, which reports that 1,112,000 Americans earned the federal minimum wage or less in 2020. Of these, 222,000 were teens from Ages 16 through 19. Teens therefore accounted for nearly one in five minimum wage workers, the second largest group by age in the U.S.

The largest age group for minimum wage workers is young adults, Age 20 through 24. In 2020, they accounted for 307,000 minimum wage workers, or nearly 28% of the total. Together, teens and young adults represent just under 48% of all those earning the U.S. federal minimum wage or less.

If the hypothesis that local employers offering higher-than-federal minimum wage is what is drawing teens into the U.S. labor force holds, it stands to reason that young adults would be likewise motivated to enter or re-enter the job market for the exact same reason, since they make up a larger share of minimum wage workers. That would be especially true during the last several months when employers have responded to a shortage of labor by boosting wages.

The following chart reveals what happened during that time for both teens (Age 16-19) and young adults (Age 20-24). For good measure, we're showing the data from January 2007 through May 2021 to capture Barry's period of interest, which confirms the data for both groups generally follow the same patterns, but we'll be focusing on more recent months in our analysis.

Percentage of U.S. Population Employed, Age 16-19 and Age 20-24, January 2007 - May 2021

The employment-to-population ratio of teens and young adults peaked in February 2020, just ahead of the arrival of the coronavirus recession in the United States. The percentage of employed for both groups plunged before bottoming in April 2020, after which both saw a steady recovery through October 2020. The onset of the second wave of coronavirus infections through the end of 2020 saw that recovery stall, with overall employment-to-population ratios holding relatively steady during this period.

The data for both groups begins to diverge after January 2020, with the employed share of young adults holding steady while the employed share of teens has risen. Since this period coincides with increased demand for labor and rising wages, the absence of an increase in the share of young adults becoming employed in this period means we can reject the hypothesis that teens only sought jobs when entry level wages rose higher than the federal minimum wage.

As for what has led to this situation, we're afraid that employment data is subject to an abundance of confounding factors, which makes determining which factors are significant difficult to untangle. We think Conor Sen's analysis pointing to teens' ineligiblity for pandemic unemployment benefits deserves greater consideration, especially since teens and young adults aren't very different from one another with respect to the other factors he mentions.

We would also suggest investigating to what extent employers desperate to fill jobs may have lowered their standards for new hires. If we're talking about standards that favored previous experience, training, or education in hiring, then we may have a good candidate for explaining why teens and not young adults are leading the job recovery in 2021.


U.S. Bureau of Labor Statistics. Labor Force Statistics from the Current Population Survey. Employed persons and employment-population ratios by age. [Online Database]. Accessed 14 June 2021.

U.S. Teens Lead Coronavirus Recession Job Recovery by Age Group

U.S. teens have become the first demographic group to fully recover their pre-coronavirus pandemic recession job levels as measured by percentage of the population employed.

That surprising result is visible in the following chart, we've tracked the non-seasonally adjusted employment-to-population ratio the various 5-year age group cohorts whose employment status is tracked through the Current Population Survey. The chart covers the period from January 2017 through May 2021, where February 2020 represents the last month before existing trends were broken by the negative employment impact of the coronavirus pandemic arriving in the United States.

Percentage of U.S. Population Employed by Age Group, January 2017 - May 2021

In the next chart, we're showing three separate snapshots in time, for February 2020 (before), April 2020 (the bottom), and May 2021 (the latest data at this writing). The chart makes it very easy to see that the employed share of the teen population has surpassed its pre-coronavirus level, unlike every other age demographic group.

Post-Coronavirus Recession Employment to Population Ratio Job Recovery by Age Group, Snapshots on February 2020, April 2020, and May 2021

With respect to all other age groups, teens are the least educated, least skilled, and least experienced segment of the U.S. labor force. And yet, this demographic group has the first to recover to its pre-coronavirus recession level of employment. We'll explore the possible reasons for that in an upcoming post.


U.S. Bureau of Labor Statistics. Labor Force Statistics from the Current Population Survey. Employed persons and employment-population ratios by age. [Online Database]. Accessed 14 June 2021.

COVID and the Magnitude of the Border Migration Crisis in Arizona

How many migrants have unlawfully crossed the U.S. border into Arizona during the Biden-Harris border migration crisis?

That's a question we realized we might be able to answer using the state's detailed COVID-19 data after seeing the surge in detentions along the southwestern border since January 2021. The trick to doing that lies in converting the number of excess COVID-19 cases in the state arising from the incoming migrants into a back-of-the-envelope estimate of their total number.

We've previously estimated that from 19 February 2021 through 30 April 2021, Arizona has recorded at least 8,150 excess COVID-19 cases because of the border migration crisis, as shown in the following chart, based on the difference between Arizona's first and second troughs following its recorded peaks in COVID cases:

Comparing Arizona's First and Second Wave COVID Troughs

That number includes the incoming migrants and also Arizona residents who might have been infected with the SARS-CoV-2 coronavirus from their encounters with any infected migrants who entered Arizona during these months, where the real number would be a percentage of that figure. That's something we can reasonably estimate. Once we have that number, we can divide it by the percentage of migrants in the U.S. government's detention facilities with COVID-19, which the Federal Emergency Management Agency indicated was "less than 6% positive" in mid-March 2021. Doing so will give us an estimate of the number of "surplus" migrants who have entered Arizona during the border migration crisis.

The rest is as easy as building a tool to do that math. Which we did, so you can supply the one missing piece of information we don't have: the percentage of Arizona's excess COVID-19 cases represented by the incoming migrants. We've set a conservative estimate default value of 50% in the following tool, which you can change as appropriate. Then click the "Calculate" button to generate your estimate (if you're accessing this article that republishes our RSS news feed, please click here to access a working version on our site).

Arizona's Excess COVID Data
Input DataValues
Excess Number of COVID Cases in Arizona During Border Migration Crisis
Percentage of Excess Cases Represented by Incoming Migrants
Percentage of Migrants Testing Positive for COVID in Detention Centers

Estimated Surplus Migrants Entering Arizona
Calculated ResultsEstimates
Surplus Migrants Entering Arizona During Border Migration Crisis

Using the 50% default value, we estimate some 67,916 surplus migrants entered Arizona since the end of January 2021. That's a little shy of 1% of Arizona's estimated 2020 population of 7.4 million. If the percentage is higher, it's no wonder the Biden-Harris administration is taking steps to contain the negative impacts from its policies.

The following 15-minute video report describes how ICE's mishandling of COVID-19 fueled outbreaks around the country:

Many of the factors described in the video report have played out in Arizona since early February 2021 with the Biden-Harris administration's border migration crisis.

The Consumer Spending of American Households Before the Pandemic

How much, and on what, was the average American household spending in a year before the coronavirus pandemic?

The Consumer Expenditure Survey (CEX) for 2019 provides the most definitive answer to that question, which we can compare with the previous 35 years worth of annual data to see how consumer spending has changed over time. Produced as a joint product of the U.S. Bureau of Labor Statistics and the U.S. Census Bureau, the CEX compiles the information collected through tens of thousands of surveys, diaries and interviews submitted by U.S. households, or "consumer units" as the BLS' data jocks affectionately call them, which provides a tremendous amount of insight into how Americans allocate their limited resources.

In 2019, the average amount spent per American 'consumer unit' household was $63,036, which was up 3% (or $1,812) from 2018's level.

We've broken down that spending by major category and visualized the data from 1984 through 2019.

Major Categories of Average Annual Expenditures per Consumer Unit, 1984-2019

Illustrating the same major expenditure categories as a percent of the average annual expenditures of a U.S. household/consumer unit makes it easy to see the trends for each of the major categories.

Percent Share of Major Categories of Average Annual Expenditures per U.S. Household Consumer Unit, 1984-2019

We have one last chart to show how consumer spending has evolved over the 36 years from 1984 through 2019. The expenditures shown on the bottom of the chart, in shades of purple, show those expenditures whose share among the total has increased over time, while the expenditures shown toward the top of the chart, in shades of green, show the household expenses whose share of total spending has fallen.

Major Categories of Consumer Spending as Share of Average Annual Total Expenditures, 1984-2019

Here's the list of major categories of consumer expenditures whose share has risen from 1984 through 2019:

  • Housing
  • Financial Products (Life Insurance, Pension Savings & Social Security)
  • Health Care (Health Insurance and Medical Expenses)
  • Charitable Contributions
  • Education

And here's the list of major categories of consumer expenditures whose share has declined over the 36 years for which the data is available:

  • Apparel and Other Products
  • Food and Alcoholic Beverages
  • Transportation
  • Entertainment

Next year's data will be fascinating since almost all of 2020's data will have been affected by the impact of the coronavirus pandemic. We'll have to wait until sometime in September 2021 to get complete data for 2020, but we may get a preliminary glimpse from interim Consumer Expenditure Survey results earlier in 2021.


U.S. Bureau of Labor Statistics. Consumer Expenditure Survey. Multiyear Tables. [PDF Documents: 1984-1991, 1992-1999, 2000-2005, 2006-2012, 2013-2019]. Reference URL: 9 September 2020. 

The Deadly Intersection of Anti-Police Protests and COVID-19

On 25 May 2020, video of the tragic death of George Floyd in police custody in Minneapolis, Minnesota exploded across social media, quickly leading to mass public protests in many cities across the United States. Coming as many states were lifting their initial lockdowns aimed at limiting the spread of COVID-19, many public health officials worried the protests would cause a resurgence of new coronavirus infections.

In the months since, little evidence has emerged to indicate that the anti-police protests contributed to a surge of new SARS-CoV-2 coronavirus infections. A NBER working paper considered data collected in 315 U.S. cities and counties that experienced protests in the five weeks following George Floyd's death, demonstrating "the protests had little effect on the spread of COVID-19 for the entire population of the counties with protests" for all but one:

However, with the exception for Maricopa County, Arizona, we find essentially no evidence that protests contributed to significant or substantial increases in COVID-19 during the period following protest onset, consistent with our main difference-in-differences findings....

We've been following Arizona's experience during the coronavirus pandemic since we identified the state had become a national hotspot for infections in the early summer of 2020. We've continued following Arizona's experience because the state's Department of Health Services makes high quality, detailed time series data available. This data makes it possible to use back calculation techniques to identify the timing of events that changed the rate of incidence for coronavirus exposures, which in turn, changed the trend for COVID-19 infections within the state.

With that being the case, we are uniquely situated to quantify the impact of 2020's anti-police protests on the spread of COVID-19 in the one state where these two factors intersected with deadly effect. In the following sections, we'll estimate the number of excess COVID-19 cases that resulted from the mass protests that took place within Arizona, along with the number of excess hospital admissions and the number of excess deaths that were attributed to COVID-19 following the protests.

Excess COVID-19 Cases Resulting from Anti-Police Protests

Arizona tracks its confirmed COVID-19 cases by their sample collection dates. Since the onset of COVID-19 symptoms is a significant driver of testing activity, occurring a median of five days after initial exposure with 95% of tests occurring within ten days following this initial viral exposure, we can use the date for positive test results to predict when a turning point will be seen following an event that significantly changed the incidence of viral exposures within a population. These figures apply when sufficient test kits and testing capacity are available to process COVID-19 tests without delays.

In Arizona, 28 May 2020 marked the first day of mass anti-police protests, which were concentrated in Maricopa County and specifically within the city of Phoenix. Adding 10 days to this date suggests that any turning point related to anti-police protests would start after 7 June 2020.

That's exactly what we observe in the following chart, showing the number of confirmed COVID-19 cases in Arizona by their sample collection date.

Confirmed COVID-19 Cases in Arizona by Sample Collection Date, 3 March 2020 - 4 December 2020 (based on data available through 7 December 2020)

The protests had high participation through 7 June 2020, before the crowds dwindled and finally petered out on 15 June 2020.

That noted, we can use the counterfactual of the trend that existed prior to the protests to estimate the number of excess COVID-19 cases that were recorded after 7 June 2020. We estimate Arizona experienced at least 29,061 more COVID-19 cases as a consequence of the anti-police protests, which is indicated by the shaded red area on the chart.

Excess Hospital Admissions Due to Anti-Police Protests

Arizona's Department of Health Services provides its data for the daily number of new COVID-19 hospital admissions. Here, we note that COVID-19 hospitalizations typically follow some 11 to 13 days following the initial viral exposure, where we would expect to see a surge of new admissions with respect to a counterfactual based on the pre-existing trend on or shortly after 8 June 2020.

Once again, that's what we observe in the data. The next chart shows the number of daily COVID-19 new hospital admissions in Arizona during the nine months from 3 March 2020 through 4 December 2020.

Daily COVID-19 New Hospital Admissions in Arizona, 3 March 2020 - 4 December 2020 (based on data available through 7 December 2020)

Here, we estimate Arizona experienced at least 2,608 more COVID-19 hospitalizations than it would otherwise have because of the anti-police protests, which is indicated by the shaded red area on the chart.

Excess Deaths Due to Anti-Police Protests

Arizona reports data for the number of actual deaths attributed to COVID-19 per day. Typically, deaths from COVID-19 follow some 17 to 21 days after the date of initial exposure to the SARS-CoV-2 coronavirus. With 28 May 2020 as the starting point, this lag would put the expecting timing of an increase in coronavirus-related deaths in Arizona resulting from the social mixing that occurred as part of the anti-police protests on or shortly after 14 June 2020.

Or would, if not for the demographics associated with the anti-police protests. By and large, individuals below the age of 45 were the main participants in the protests, where fewer than six percent of all deaths attributed to COVID-19 in Arizona have occurred within this age group.

However, that doesn't mean the anti-police protests didn't result in a surge of COVID-19 deaths in the state. We think that after large scale protests in Arizona ended on 7 June 2020, participants carrying coronavirus infections went home, so to speak, where their subsequent social interactions would result in the infection of older Arizonans. Using 7 June 2020 as the starting point for those post-protest interactions, that would put the expected timing of a change in trend for COVID-19 deaths between 25 June 2020 and 28 June 2020.

This scenario fits what we observe in the following chart showing the number of deaths attributed to COVID-19 each day in Arizona from 3 March 2020 through 4 December 2020.

Daily COVID-19 Deaths in Arizona, 3 March 2020 - 4 December 2020 (based on data available through 7 December 2020)

We estimate Arizona experienced at least 594 more coronavirus-related deaths than it would otherwise have due to the anti-police protests, which is indicated by the shaded red area on the chart.

Personal Injuries and Wrongful Deaths

What if the organizers of Arizona's anti-police protests were held accountable for the excess COVID-19 hospitalizations and deaths that occurred as a consequence of their actions?

That's a real question because there were activists in Arizona who recognized the risk of spreading COVID-19 infections, who chose to not engage in mass protests. BLM activist Lola Rainey explains her thinking for why she didn't organize mass street protests in Tucson, Arizona:

"When you talk about taking to the streets an organized movement like that, you have to be aware that you're putting other people's lives at risk. One, because we are in a pandemic, and there are a lot of things you cannot control when people are out in the streets like that,” said Rainey.

Alas, that kind of responsible thinking wasn't anywhere to be found in Phoenix, Arizona, where disputes among protest organizers all but ensured an uncontrolled situation that would promote the spread of coronavirus infections, with participants engaging in behaviors that ran counter to the guidance of public health officials.

In that regard, the various organizers and promoters of the anti-police protests could be held legally liable for the excess COVID-19 hospitalizations and deaths that resulted from their collective actions and negligence, if any enterprising trial attorneys were looking to score some easy personal injury and wrongful death settlements.

Using data from California, the median personal injury settlement is $114,305, so Arizona's protest-related 2,608 excess COVID-19 hospitalizations could represent a $332 million payday for the injured and their lawyers. Meanwhile, with a median wrongful death award of $2.2 million, the corresponding payout for wrongful death claims could total over $1.3 billion. With numbers like that, how long do you suppose it might be before class action lawsuits start being filed by hungry attorneys representing damaged clients?

Why Is the Deadly Intersection of Anti-Police Protests and COVID-19 in Arizona?

The previous section explains "why Phoenix?", but a larger question is "why Arizona?"

We suspect the answer lies in a unique confluence of events. For one, Arizona is unique in its climate. Unlike nearly all other places where mass anti-police protests occurred at this time, the city of Phoenix was already registering daily high temperatures between 100 and 110 degrees Fahrenheit at the end of May 2020 and in early June 2020. Those kinds of temperatures mean that groups of protesters would periodically break away from protesting to cool off in air-conditioned environments. Such environments have been found to be conducive for spreading respiratory viruses like SARS-CoV-2, providing an additional factor for increasing the incidence of infections beyond that from direct participation in the protests.

We think the timing of Arizona's initial lockdown period also played a significant role. Arizona lifted its statewide lockdown on 15 May 2020, which led to a steady but managable growth in the number of infections in the following weeks. At the time of the protests, that increase had provided a critical mass of infected individuals that would fuel a surge of new infections under the conditions of a large, uncontrolled event like mass political protests where social distancing would not be effectively practiced.

Though Arizona's leaders could not have possibly forecast such an event would occur when it did, we can compare the state's experience with that of Nevada, where Las Vegas is the only other major city in the U.S. with a climate similar to Phoenix to test this factor. Here, we find that Nevada delayed lifting its lockdown for nearly two weeks after Arizona, which meant it had a much smaller pool of infected individuals at the start of its protests. Those lower numbers would have greatly reduced its risk of experiencing a surge in cases at the time of the protests, which potentially explains "why Arizona and why not Nevada?", even though both share similar hot, arid climates.

If it wasn't already clear, the role of protest organizers and those supporting or promoting them in contributing to the conditions that promoted the spread of COVID-19 in Arizona cannot be overstated.

There may be additional, lesser factors to consider, but these are the big three that could very well explain most of why the intersection of anti-police protests and COVID-19 proved to be so deadly in Arizona and not elsewhere in the United States.

A Resurgence of Cases

Finally, we should recognize the surge in coronavirus cases that Arizona is currently experiencing. We've previously traced the origin of the current surge to 2020 political campaign events occurring two and a half weeks before the 3 November 2020 election, which stand out because they involved many more people than the anti-police protests did. On 3 December 2020, Arizona Governor Doug Ducey initiated new restrictions on large events and businesses in the state. We should first start seeing any effects from the new restrictions in the state's data for cases and hospital admissions reported after 14 December 2020.

Celebrating Political Calculations' Anniversary

Our anniversary posts typically represent the biggest ideas and celebration of the original work we develop here each year. Here are our landmark posts from previous years:

  • A Year's Worth of Tools (2005) - we celebrated our first anniversary by listing all the tools we created in our first year. There were just 48 back then. Today, there are over 300....
  • The S&P 500 At Your Fingertips (2006) - the most popular tool we've ever created, allowing users to calculate the rate of return for investments in the S&P 500, both with and without the effects of inflation, and with and without the reinvestment of dividends, between any two months since January 1871.
  • The Sun, In the Center (2007) - we identify the primary driver of stock prices and describe a whole new way to visualize where they're going (especially in periods of order!)
  • Acceleration, Amplification and Shifting Time (2008) - we apply elements of chaos theory to describe and predict how stock prices will change, even in periods of disorder.
  • The Trigger Point for Taxes (2009) - we work out both when, and by how much, U.S. politicians are likely to change the top U.S. income tax rate. Sadly, events in recent years have proven us right.
  • The Zero Deficit Line (2010) - a whole new way to find out how much federal government spending Americans can really afford and how much Americans cannot really afford!
  • Can Increasing the Minimum Wage Boost GDP? (2011) - using data for teens and young adults spanning 1994 and 2010, not only do we demonstrate that increasing the minimum wage fails to increase GDP, we demonstrate that it reduces employment and increases income inequality as well!
  • The Discovery of the Unseen (2012) - we go where so-called experts on income inequality fear to tread and reveal that U.S. household income inequality has increased over time mostly because more Americans live alone!

We marked our 2013 anniversary in three parts, since we were telling a story too big to be told in a single blog post! Here they are:

  • The Major Trends in U.S. Income Inequality Since 1947 (2013, Part 1) - we revisit the U.S. Census Bureau's income inequality data for American individuals, families and households to see what it really tells us.
  • The Widows Peak (2013, Part 2) - we identify when the dramatic increase in the number of Americans living alone really occurred and identify which Americans found themselves in that situation.
  • The Men Who Weren't There (2013, Part 3) - our final anniversary post installment explores the lasting impact of the men who died in the service of their country in World War 2 and the hole in society that they left behind, which was felt decades later as the dramatic increase in income inequality for U.S. families and households.

Resuming our list of anniversary posts....

Previously on Political Calculations

Here's our previous Arizona coronavirus coverage presented in reverse chronological order, with a sampling of some of our other COVID analysis!


Arizona Department of Health Services. COVID-19 Data Dashboard. [Online Application/Database].

Maricopa County Coronavirus Disease (COVID-19). COVID-19 Data Archive. Maricopa County Daily Data Reports. [PDF Document Directory, Daily Dashboard].

Stephen A. Lauer, Kyra H. Grantz, Qifang Bi, Forrest K. Jones, Qulu Zheng, Hannah R. Meredith, Andrew S. Azman, Nicholas G. Reich, Justin Lessler. The Incubation Period of Coronavirus Disease 2019 (COVID-19) From Publicly Reported Confirmed Cases: Estimation and Application. Annals of Internal Medicine, 5 May 2020.

U.S. Centers for Disease Control and Prevention. COVID-19 Pandemic Planning Scenarios. [PDF Document]. Updated 10 September 2020.

COVID Tracking Project. Most Recent Data. [Online Database]. Accessed 10 November 2020.

More or Less: Behind the Stats. Ethnic minority deaths, climate change and lockdown. Interview with Kit Yates discussing back calculation. BBC Radio 4. [Podcast: 8:18 to 14:07]. 29 April 2020.