INEQUALITY CONTROL Economist Joseph E. Stiglitz’s article (“A Rigged Economy” [The Science of Inequality]) on how we got to today’s lamentable economic state in the U.S. is spot on. Yet let me give my own, more simple explanation: When I was a young man, during the three decades after FDR and the New Deal, the maximum federal tax rate (applied at the time to income exceeding an amount that has ranged between $100,000 and $500,000) was between 70 and 91 percent. As such, federal taxation was highly progressive; the rich were few and were not so obscenely wealthy, and most important, the middle class was dominant. In 1981, shortly after taking office, Ronald Reagan slashed the top bracket’s rate to 50 percent and then, in 1986, to 28 percent—a tremendous windfall for the rich that continues unabated (today’s top rate is 37 percent on income exceeding $500,000). One has only to look at Stiglitz’s graphs, in which everything takes a turn for the worse after 1980, to see how our current tax code lines the pockets of the rich and steadily erodes the middle class. We either return to a progressive tax policy or continue the descent into plutocracy. R. C. GIBSON Irvine, Calif. I agree with the points that Stiglitz (who is my former professor) makes about the causes of inequality, as well as those that James K. Boyce makes about the links between environmental degradation and inequality in “The Environmental Cost of Inequality” [The Science of Inequality]. But Stiglitz’s list of needed policy changes falls short, as does Boyce’s reliance on environmental activists to save flora, fauna and natural resources. We know that the problems with slavery would not have been fixed simply by calling for more regulation and stiffer penalties. Our laws today ensure that a few can claim excessive wealth and power. By what right do those owning firms have the power to decide how the income and wealth generated by the talent and labor of many are used? Stakeholders—employees, customers, the communities affected by a company’s decisions—have rights that require greater recognition. Stakeholders’ interests should be represented on the boards of big firms. Those with revenues exceeding $1 billion should be required to have a national charter that would lay out obligations and penalties. Where would the power to institute such changes originate? My fellow economists are very reluctant to talk about political parties, yet we want to influence political platforms. We can at least begin to identify not only where the public interest lies but also what kind of political group is most likely to represent those interests. MARIANNE HILL South Portland, Me. REDISTRICT JUDGE In “Geometry v. Gerrymandering,” Moon Duchin describes mathematicians’ efforts to create statistical methods to detect and replace biased voting district maps. Last November’s election in Missouri had an amendment on the ballot, approved by about 62 percent of the vote, to change “the process and criteria for redrawing state legislative districts during reapportionment.” (While many argued that the ballot wording was deceptive, one needs examine the details. The full statement can be found here: www.sos.mo.gov/elections/petitions/2018BallotMeasures) I wonder if any of the ideas or analyses Duchin presents could be used toward validating the method outlined in the constitutional amendment before actual redistricting maps are constructed. MORITZ FARBSTEIN St. Louis, Mo. The Markov chain Monte Carlo (MCMC) process for redistricting that Duchin describes requires the public to trust both the mathematics and the mathematicians. The possible configurations are so enormous that it reminds me of all the possible outcomes in a game of chess. And yet even beginners play chess without being overwhelmed by the vast number of moves. Perhaps the entire process could be treated more like chess, with the two sides taking turns choosing a district to maximize its number of voters instead of letting one side make all the moves for both. If one side outnumbers the other, that side may be given proportionally more choices. The final result would be approved by a judge or a redistricting committee. There is no need to resort to the massive computations in MCMC as long as the process of choosing the districts is fair. BENJAMIN JONES via e-mail DUCHIN REPLIES: The Missouri amendment that Farbstein refers to belongs to a crop of state-level reform measures approved by voters in 2018 (joining Colorado, Utah, Ohio and Michigan). Missouri’s was especially detailed: specific criteria were laid out, including a formula to define “partisan fairness” and a precise way to measure “competitiveness.” A legitimate worry for such reforms is that trade-offs in redistricting priorities are so complicated that well-meaning rules might actually conflict or have unintended consequences. This raises scientific questions, and they are approachable! Sampling from the universe of plans can illustrate the cost to one priority as another is introduced and can give a state-specific baseline or normal range for metrics used to evaluate a commission’s proposals. Regarding Jones’s letter: One of the problems caused by rampant redistricting abuse is precisely the erosion of public trust, and restoring it will require transparency. MCMC for redistricting needs to be open-source, peer-reviewed and fully auditable. In terms of the problem’s complexity, strategy games like chess—or Go—are great examples. The rules are simple, but mastery is elusive. Note that advocates of outlier analysis propose to use MCMC only to evaluate plans and not to select them. There must still be a role for local knowledge, community input and all things human. But a bird’s-eye view of the possibilities can help ensure that no group’s interests are trampled, no matter what process is used. CORRELATION TO MURDER Maia Szalavitz’s story “Income Inequality and Homicide” [Forum] refers to psychologist Martin Daly’s assertion that income inequality predicts murder rates better than other variables do. I am perplexed by Daly’s use of murder rates alone to nail down his conclusions rather than looking at a more relevant tally of violent assaults overall. Death rates alone are often used in discussions of gun violence and highway speed limits. And yet death is just one of several possible outcomes of a violent assault. The end-of-year murder rate is more dependent on the access and actions of responding EMTs and hospital trauma teams. The impact of trauma center success stories in high-crime areas may be ignored in statistical studies and yet might be the primary reason for drops in regional murder rates. JOHN ANDREWS Milford, N.J. ERRATA “Back in Time,” by Dan Coe, should have referred to galaxies likely at a redshift of around two as being three billion years old, or nearly a quarter of the universe’s age, rather than 10 billion years old. “A Rigged Economy,” by Joseph E. Stiglitz [The Science of Inequality], should have referred to the “North American Free Trade Agreement,” not the “North Atlantic Free Trade Agreement.”
Economist Joseph E. Stiglitz’s article (“A Rigged Economy” [The Science of Inequality]) on how we got to today’s lamentable economic state in the U.S. is spot on. Yet let me give my own, more simple explanation: When I was a young man, during the three decades after FDR and the New Deal, the maximum federal tax rate (applied at the time to income exceeding an amount that has ranged between $100,000 and $500,000) was between 70 and 91 percent. As such, federal taxation was highly progressive; the rich were few and were not so obscenely wealthy, and most important, the middle class was dominant. In 1981, shortly after taking office, Ronald Reagan slashed the top bracket’s rate to 50 percent and then, in 1986, to 28 percent—a tremendous windfall for the rich that continues unabated (today’s top rate is 37 percent on income exceeding $500,000).
One has only to look at Stiglitz’s graphs, in which everything takes a turn for the worse after 1980, to see how our current tax code lines the pockets of the rich and steadily erodes the middle class. We either return to a progressive tax policy or continue the descent into plutocracy.
R. C. GIBSON Irvine, Calif.
I agree with the points that Stiglitz (who is my former professor) makes about the causes of inequality, as well as those that James K. Boyce makes about the links between environmental degradation and inequality in “The Environmental Cost of Inequality” [The Science of Inequality]. But Stiglitz’s list of needed policy changes falls short, as does Boyce’s reliance on environmental activists to save flora, fauna and natural resources.
We know that the problems with slavery would not have been fixed simply by calling for more regulation and stiffer penalties. Our laws today ensure that a few can claim excessive wealth and power. By what right do those owning firms have the power to decide how the income and wealth generated by the talent and labor of many are used? Stakeholders—employees, customers, the communities affected by a company’s decisions—have rights that require greater recognition. Stakeholders’ interests should be represented on the boards of big firms. Those with revenues exceeding $1 billion should be required to have a national charter that would lay out obligations and penalties.
Where would the power to institute such changes originate? My fellow economists are very reluctant to talk about political parties, yet we want to influence political platforms. We can at least begin to identify not only where the public interest lies but also what kind of political group is most likely to represent those interests.
MARIANNE HILL South Portland, Me.
REDISTRICT JUDGE
In “Geometry v. Gerrymandering,” Moon Duchin describes mathematicians’ efforts to create statistical methods to detect and replace biased voting district maps.
Last November’s election in Missouri had an amendment on the ballot, approved by about 62 percent of the vote, to change “the process and criteria for redrawing state legislative districts during reapportionment.” (While many argued that the ballot wording was deceptive, one needs examine the details. The full statement can be found here: www.sos.mo.gov/elections/petitions/2018BallotMeasures)
I wonder if any of the ideas or analyses Duchin presents could be used toward validating the method outlined in the constitutional amendment before actual redistricting maps are constructed.
MORITZ FARBSTEIN St. Louis, Mo.
The Markov chain Monte Carlo (MCMC) process for redistricting that Duchin describes requires the public to trust both the mathematics and the mathematicians. The possible configurations are so enormous that it reminds me of all the possible outcomes in a game of chess. And yet even beginners play chess without being overwhelmed by the vast number of moves.
Perhaps the entire process could be treated more like chess, with the two sides taking turns choosing a district to maximize its number of voters instead of letting one side make all the moves for both. If one side outnumbers the other, that side may be given proportionally more choices. The final result would be approved by a judge or a redistricting committee.
There is no need to resort to the massive computations in MCMC as long as the process of choosing the districts is fair.
BENJAMIN JONES via e-mail
DUCHIN REPLIES: The Missouri amendment that Farbstein refers to belongs to a crop of state-level reform measures approved by voters in 2018 (joining Colorado, Utah, Ohio and Michigan). Missouri’s was especially detailed: specific criteria were laid out, including a formula to define “partisan fairness” and a precise way to measure “competitiveness.” A legitimate worry for such reforms is that trade-offs in redistricting priorities are so complicated that well-meaning rules might actually conflict or have unintended consequences. This raises scientific questions, and they are approachable! Sampling from the universe of plans can illustrate the cost to one priority as another is introduced and can give a state-specific baseline or normal range for metrics used to evaluate a commission’s proposals.
Regarding Jones’s letter: One of the problems caused by rampant redistricting abuse is precisely the erosion of public trust, and restoring it will require transparency. MCMC for redistricting needs to be open-source, peer-reviewed and fully auditable. In terms of the problem’s complexity, strategy games like chess—or Go—are great examples. The rules are simple, but mastery is elusive.
Note that advocates of outlier analysis propose to use MCMC only to evaluate plans and not to select them. There must still be a role for local knowledge, community input and all things human. But a bird’s-eye view of the possibilities can help ensure that no group’s interests are trampled, no matter what process is used.
CORRELATION TO MURDER
Maia Szalavitz’s story “Income Inequality and Homicide” [Forum] refers to psychologist Martin Daly’s assertion that income inequality predicts murder rates better than other variables do.
I am perplexed by Daly’s use of murder rates alone to nail down his conclusions rather than looking at a more relevant tally of violent assaults overall. Death rates alone are often used in discussions of gun violence and highway speed limits. And yet death is just one of several possible outcomes of a violent assault. The end-of-year murder rate is more dependent on the access and actions of responding EMTs and hospital trauma teams. The impact of trauma center success stories in high-crime areas may be ignored in statistical studies and yet might be the primary reason for drops in regional murder rates.
JOHN ANDREWS Milford, N.J.
ERRATA
“Back in Time,” by Dan Coe, should have referred to galaxies likely at a redshift of around two as being three billion years old, or nearly a quarter of the universe’s age, rather than 10 billion years old.
“A Rigged Economy,” by Joseph E. Stiglitz [The Science of Inequality], should have referred to the “North American Free Trade Agreement,” not the “North Atlantic Free Trade Agreement.”
