Thanks go to Mark Newman, Department of Physics and Center for the Study of Complex Systems, University of Michigan. Mark graciously grants Creative Commons share and share alike use of his cartograms. The featured image cartogram weights each state by its number of electoral votes and colors it by its 2012 presidential election winner.
In this article, I examine the original design objectives of the Electoral College and the biases in the Electoral College that result from our country currently having a majority of large low population states. I deliberately avoid matters of party beyond the obvious urban-rural party preference.
Reference  offers some suggestions for how these faults may be corrected. I may write about Professor Sabato’s proposal in a future article.
Thanks to Five Thirty Eight and NPR for the use of the figures illustrating this post. And also thanks to Nate Silver for his innovative work in political election modeling. This article explains how to interpret the Five Thirty Eight election products. Thinking about probabilistic and statistical models is tricky stuff. What Nate is doing is essentially political weather forecasting.
This post explains how to think about mathematical models of things. The models can be dynamical system models like those used in weather forecasting, statistical models also used in weather forecasting, or statistical models as used in political prediction. Models are useful for a particular range of subject matter and most models are carefully crafted to be useful in a particular region of a subject matter domain.
While working as a nuclear reactor plant designer, we used different models to predict accident response of the reactor than the ones we used to design its control laws and tune its control systems. We used radically different models to design the contents of the fuel pellets and the zoning of pellets along a pin and build up of pins into modules and modules into a reactor core.
What is a model
A model is an abstraction of some real world system that is useful in reasoning about and answering questions about the behavior of that real world system. This is true of all of the models described above and of the https://fivethirtyeight.com/ election model described after the break. When polling organizations sample the electorate, they collect information about respondents political opinions which are distilled into a statistical description of the electorate. Based on expressed preferences and responses to confirming questions, the pollster makes a model and reports preferences grouped by reported demographics. Remember, on the phone, no one knows you’re really a dog.
Dismal Wizard is a bit concerned because Norfolk has not logged in his returned mail ballot. The carrier picked it up on 22 September and local delivery is 2nd day or better. It only takes a minute to run the lot through the OCR to scan the outer smart mail code that ties the ballot to a poll book entry.
The wheels of government turn slow but they do turn, Norfolk, Virginia Elections Office assures me. The Registrar (head honcho) has not revealed to staff the office’s internal timeline for doing the work it needs to do. Speaking with a staffer, they are still responding to absentee ballot requests and have been holding returned ballots for envelope scanning until they are caught up with the requests. They expect they should be caught up in early October. They are deliberately holding the mail ballots until they have time to properly perform the quality processes that assure that they are all recorded. Everything happens under two person control using reader-worker procedures like those used in nuclear power.
Virginia law sets the following constraints for county elections offices
Hand counting of absentee ballots may begin at 3 PM on election day.
Machine counting of absentee ballots may begin as they are received but the tally must remain in the machine until the machines are canvased at the close of polls.
Norfolk machine counts everything using mail sorters and optical scanners. In practice, Norfolk must scan the envelopes before election day. This will happen before in person voting begins. This scanning process updates the poll book to record participation in the election. Later, the signature is verified against the application and the signature on file (usually your drivers license signature). Missing and poor quality signatures are reported to the voters at this point so they can correct the issue before early voting and absentee requests close.
Experience in the vote by mail states indicates that 1/3 of ballots are returned in the first week, 1/3 are returned on election day, and 1/3 come in between the two milestones. Norfolk staff are currently full time on responding to ballot requests with 2000 0r so in hand on 28 September. Once these are sent out, they expect to scan the returns that they have in hand. Later, they will open the envelopes (2 man control) and scan the ballots. They know better than to wait until 3 PM on election day. October 31 is the close of early voting so they’ll probably record the inventory ballot votes on that day. Totals remain in the machines until the polls close. Machines used for early in person voting and mail voting are set aside under seal and not used on election day.
Virginia 2020 Election Schedule
Virginia election statutes specify election milestones in detail. The 2020 election table appears in . It is a multi-page spreadsheet having 253 entries (why the Elections Office staffer wouldn’t answer). Some of the milestones of interest to voters are below.
Last day to register to vote is 10/13
Last day to request absentee ballot by mail is 10/23
Last day to request absentee ballot in person is 10/31
Last day of early voting absentee ballots is 10/31
Last day to request a Mulligan absentee ballot is 10/31
Absentee ballots polls close on 11/06. Absentee ballots received after the election day poll close are counted on 11/07.
Election counting processes complete on 11/10
Election certificates sent to elected candidates on 11/11
City officials have 4 days in which to count absentee ballots and seven days in which to check, total, and certify the official results. The statutes don’t specify how the jurisdiction must schedule its internal work flow, just how local mechanics must match up to larger election mechanics. Other statutes establish verification and validation procedures. Part of these is the testing of machines with standard ballots prior to their use in an election.
Note that many expatriates, military personnel, and diplomatic personnel vote by mail from overseas. Mail ballots must be postmarked election day or before. The mail ballot close is set to allow these important groups to vote. National elections have about one million OCONUS participants.
Election Night Results
The election night results are unofficial and incomplete because they don’t include the mail ballots requested late in the election process. The city also has to verify the results to ensure that all machines have been read out and included in the total. The totals themselves have to be checked. Some elections are close and may swing back and forth as the last results are located and reported. In one case, the election was an actual tie broken by game of chance.
Election night results usually tell the tale because most elections are not so close that the outstanding mail ballots will change the result. It is reasonable to expect last minute mail ballots to have a vote distribution similar to the vote already cast as these last minute requests are driven by last minute schedule changes. This is not true with the early ballots which have historically had a Republican Party bias in many states.
Virginia Absentee Voting Law Changes
With the liberalization of Virginia election law, it is likely that those who have to work on election day will dominate the absentee and early voting. Lower wage occupations and child care logistics will drive this change. Many lower income voters were staying on the sidelines because they could not afford to give up a half-day of wages to vote. Similarly, many lower wage households had multiple jobs constraining the day job worker to take evening childcare while parter worked. The new Virginia mail and early voting changes will change Virginia election demographics. These new voters may well skew Democratic.
Our Beloved President
Don’t Vote By Mail
Our beloved President has been disparaging use of the absentee ballot process thinking that Democratic Party voters used it more heavily than Republican Party voters. In prior years, it was the other way around. Republican voters have been more likely to vote using absentee procedures. This is typically because older mobility limited voters were voting absentee and this demographic skewed Republican. This will be the case in Florida. Is he suppressing his own vote in a key small margin state?
Vote early and again on election day
At one time, our beloved President was advising Republicans to vote absentee and attempt to vote again on election day. This can be interpreted as a felonious act in several ways. In many jurisdictions soliciting multiple votes from an individual is a felony. In many jurisdictions, queuing of absentee voters in the election day queue can be taken as a voter suppression tactic because each person has to be looked up in the poll book and told to vote or leave. In many states (yes, you, Georgia), districts that are majority party out of power are deliberately under provisioned with election equipment and poll workers. This is a voter suppression tactic. The President’s double voting scheme would aggravate the problem. Simply make less voting resources available to those people.
Voter suppression schemes are illegal
The various schemes to challenge absentee balloting can also be taken as voter suppression schemes. The Republicans are doing this in plain sight. Let’s gum up the works by challenging all of the signatures! Clever. But we’ll only do it in minority Republican districts. Same thing with challenging all of the machine rejects were boxes were not completely filled.
The featured image shows the status of my Virginia 2020 fall election ballot request. The elections office has received and approved my request and it is queued for the mail. As always, there is more. After the break, I’ll show you how to track your absentee ballot through the process using the Virginia Elections Office Citizen Portal.
The 2020 legislature changed Virginia voting laws to permit general voting by mail and to remove the ballot witnessing requirement. These changes allow anyone to vote by mail in any Virginia election without a note from their mother.
Exercising your franchise is important to the vitality of US, Virginia, and Norfolk City polity. I have attempted to summarize Virginia policy and procedures for voter registration and absentee balloting. Please read the linked pages and use the links for all voting-related activities. The display text and link text are identical. Please confirm this by looking at the link target display in your browser, typically appearing in the window footer.
Reference , the citizen portal, is where you register and verify that your ballot was received and accepted. The other links give background information.
Reference  shows a typical locality Elections Office website. There, your general registrar will give information unique to the locale, especially local ballot questions and office citizen procedures and hours.
As of this writing (8 August 2020) I am unable to verify if the witness signature is still required on the outer envelope. It is the last bit of Jim Crow remaining in Virginia election procedures. I believe the legislature was to remove it in the emergency budget session scheduled for 18 August 2020.
This started out as a draft of a note for Senator Mark Warner, one of the Democratic Party brain trust working on the 2020 campaign strategy. Thinking before the pandemic was that the campaign would focus on a return to normalcy. Post-pandemic, it is clear that many elements of national resiliency need a major overhaul.
Our government has suffered from two things, potentially related. First, one of our national political parties has been using a strategy of divide and conquer to keep power to itself for oligarchical purposes. This includes campaigns based on fear of them.
Second, our social programs were developed individually in this hostile environment to treat symptoms rather than in support of essential governmental and civic resiliency objectives.
At every point on our journey to our current resiliency system, one party waged a two pronged fear campaign — fear of communism or now socialism, whatever that is, and fear that they, people not like us, might get something that we didn’t or cause us harm more directly.
The resulting difficulty of passing coherent legislation has led to a patchwork of well-intentioned but buggy programs. This patchwork is full of gaps and special cases making it difficult for the patchwork to support maintenance of national resiliency and unable to support many of us whose occupations do not permit a secure standard of living. For these purposes, a secure standard of living includes safe housing, essential health care for chronic and acute diseases and trauma, retirement security, and economic during earned income interruptions.
This post makes the argument that we need to take a step back, change our policy from one of aiding individuals ad hoc to one of achieving a resilient government and society that can survive foreseeable challenges by aiding individuals in a holistic fashion. It is largely about keeping domino chains from topping.
Our Congress’s response to the 2020 COVID-19 outbreak has been largely ad hoc. In spite of Congress’s “best” efforts, it has largely missed the mark by slow delivery of ad hoc relief measures that treat individual economic effects of the pandemic. The economy is far too complex for such a strategy to be successful.
Rather relief needs to enter the economy at the same points from which economic activity originates and flow through the economic networks that service normal transactions. Where possible, relief should permit the individual transactions that would occur if isolation measures were not in place. Normal emergent flows would support most of the economy.
This article considers the characteristics that promote societal resilience. How we go about achieving these characteristics is a complex undertaking that will emerge a bit at a time. Much of what we need already exists but needs refinement and protection from schemers. But essential elements of individual and familial security require an overhaul and processes that now require legislative action need to become automatic.
This post grew out of research I did to prepare a discussion about income in the United States for my church’s discussion group. As I prepared the presentation materials for the opening of discussion, I learned quite a bit about how fortunate I was and how things fit together. This post is based on the following references.
Income, Poverty, and Health Insurance Coverage in the United States: 2012, United States Censure Bureau Report P60-245, 2013.
An understanding of this information is important to making both personal and public policy decisions.
Standard of living
How much income does it take to support an individual or household in the United States? I was surprised to learn these figures. Unfortunately, I’ve forgotten the source so take these as 2013 thumb rules
$30,000 individual self-sufficiency possible
$40,000 individual savings for retirement and set backs
$10,000 per additional member of the household.
The $30,000 figure represents the income needed to live independently and provides basic housing, transportation for work, food security, and basic health care in the absence of chronic diseases or injury.
The $40,000 figure represents savings for retirement and for maintenance of reserves to meet minor health emergencies, out of work contingencies, save to replace a vehicle, etc.
The $10,000 figure represents the incremental cost of adding an additional non-working resident to the household. Thus, a single head of household with 2 children requires $50,000 for a basic standard of living and $60,000 for a secure standard of living. For a two parent household, raise these figures to $60,000 and $70,000.
Individual Income Distribution
This figure shows the Census Bureau’s 2012 estimate of individual income density in the US. The bar height is proportional to the number of individuals in a $2500 band, for example from $40,000 to $42,999. Normalizing by the total number of people surveyed gives an estimate of the probability density function of income levels in the US.
There are some inconvenient truths here.
The distribution is not Gaussian
The distribution is bottom weighted
The distribution is noisy
There are high-low income band pairs, cause unknown
The important thing to take away is how income is distributed. A large swath of young (< 15) are counted as zero income. The median individual income of about $40,000 is well below the middle of the range considered in the survey ($50,000). Household income has a similar distribution with the median income being $51,000.
Because the data is gnarly, it is helpful to think of it by percentiles as shown by the figure below.
The figure shows some of the more important income bands. A percentile boundary represents the fraction of the population making less than that income level. For example, the tenth percentile tells us that 10 percent of the population earns less than $10,500. One quarter of us earn less than $22,500. One third of us earn between $30,000 and $62,500. Similarly, one quarter of us earn more than $77,000 and one fifth of us earn more than $92,000. Our doctors and dentists earn more than do 98 percent of us. The top 1.5% of income earners make more than $167,000 and to leave the 99% requires an income greater than $350,000.
Income and Standard of Living
Let’s interpret the income figures in terms of standard of living. Most importantly, the bottom 1/3 of us do not have the income to live independently. The middle 1/3 of us range from struggling to get by to independent with some savings. A 90th percentile income, although statistically wealthy, is not practically wealthy and requires careful choices of housing, automobiles, children’s education, etc.
Occupation and Income
The table above shows median income for commonly encountered occupations. Median income is that income level dividing the occupation into two equal sized groups. Half make less than median and half make more. In choosing occupations from the reference, I was careful to choose occupations we commonly interact with. So barbers, auto mechanics, plumbers, waiters, cooks, janitors, dentists, surgeons, etc are all present as are some glamorous occupations like airline pilot.
Race and Income
The table above shows median household income by racial group. Just the facts, no opinions and no rewarming of racial stereotypes. But half of black and Hispanic households are struggling as are maybe half of all households.
Age and Income
This figure shows median income by age group taking the 10 year slices commonly used.
Implications for Markets
Other than the racial disparities, skill, experience, and the emergent nature of economic system behavior go a long way toward explaining these data. Broadly consumed goods and services must be either inexpensive or subsidized. For example, we all need our hair cut. To be affordable places an upper bound on the fees for this service and the earnings in the profession. The providers of this service don’t have a lot of pricing power because half of their market earns less than $40,000.
The data show that the keys to upward mobility are educational attainment and experience in our profession, trade or occupation. But demand for services sets limits on upward mobility. First, we can’t all be rock stars or brain surgeons. The demand is not there. As the data shows, the most important thing we can do is to finish high school. After that, we can attend a trade school, apprentice in a trade, or attend college to acquire professional knowledge and gain experience in our profession being attentive to changes in demand for our services.
Implications for policy
The data suggest a few implications for public policy
Tax where there is money to be had, that is the top quartile.
Services in broad demand must be inexpensive or subsidised
Goods in broad demand must be inexpensive or subsidised
Jobs are demand driven. Subsidizing the bottom 1/3 of us produces demand for goods and services. Subsidizing the top 1% produces speculation or savings.
And a few implications for personal choices.
Education affords access to an occupation, trade, or profession that is in demand and pays well.
Choose occupations, trades, or professions for which demand is growing or under-served.
Choose occupations, trades, or professions that pay well
Be attentive to changes in demand for your occupation, trade, or profession and follow demand.
April 1’st Unitarian Church of Norfolk service was a lay service conducted by the youth group. Their theme was making the world a better place one small act at a time. No sixties grandiosity, just ten’s pragmatism — do the doable, it will matter to this one. The youth carried this theme through the service beautifully in all of its elements. One, Eric Vick, is a budding singer songwriter who killed the following original work.
why make a change?
Why do anything?
That’s all i hear coming out of this world
why make a change?
it won’t do anything.
the worlds as crummy as it’ll ever get
that’s all i hear when i walk down the streets,
that’s all i hear when i turn on the news.
this is what we say, to the world.
you can’t save them all, well what about just one.
which one is good enough, which one deserves it.
i can’t save them all, but I’ll go ahead with it.
I’ll choose this one, and know its worth it.
Negativity, shouldn’t slow us down
we are the molders and builders of earth
our thoughts provoke, and sway their minds
what would you say if i told you one thing?
one small change could do, great things.
One Thing copyright by Eric Vick, used by permission. Eric Vick Hampton Roads teen, not the Eric Vick at myspace.com.
This service closed to a standing ovation and was easily the best since my return to UCN in June 2010. Bravo Zulu teens.