Pages

Thursday, March 9, 2017

Avalanche of hate

Note: this critical article about hate and ideas about our shared global community has received >600 likes/shares and posted in Bloomberg and by top marketing firms and elsewhere.  Sequel here.  Please read and share with your friends.

Every day we witness a drip, drip, drip of negative news.  Hate crimes now occur so regularly we often fail to notice.  By the following day we are onto other matters or perhaps fresh victims: at random K-12 schools, offices, and public venues across the country.   Despite some detachment, a staggering toll speaks for itself.  Nearly a couple thousand hate incidents have happened, in just the four months since President Trump was elected.  We don’t say this in any way to place all culpability on him.  President Trump started his campaign years ago in an already worsening and intricate atmosphere.  But most of us must do better to address the broad swath of quality and emerging data on the ground: hate crimes have been swelling in the past couple years and have recently reached record sustained levels that are a risk for us all.  And history shows things will sternly run amok, if not quickly checked.  Right now, we lack one consistent and recent federal data that can help us more efficiently create a robust analysis of how hate crimes have developed over time.  And with lack of funding this will only get worse.  While our data comports to the FBI reports, those reports also we show as (without malice) biased.  However there is an ample amount of disparate data that is paired over time, and moving in the same direction.  It takes time to process it all, particularly textual intake of news, but through this process we have uncovered a sense of precisely some of the parameters of hate in the U.S., and in exactly how the situation is deteriorating over time.  Have a look at this early peek on where the country’s trends in hate have been and what we have learned from the past when nations try to correct these societal difficulties.  Here are the topics we will discuss: (a) different hate crime targets, (b) changes in Islamophobia, (c) patterns in violent versus nonviolent hate crimes, and (d) mapping where post-election hate crimes are occurring.

Discrimination and hate have been with us throughout human history, where we see ebbs and flow in the progress of good for a time and sometimes bad for a time.  We must ask ourselves if it is a richer life to live when all our neighbors are also doing well.  What we have seen in recent months is not the typical America we have experienced in our life.  At the same time, nothing seen is novel.  Thousands of years ago we saw Jewish cemeteries vandalized, and people of “different faiths” persecuted.  Women and children were not immune.  Having an uncommon sexual orientation was also restricting oneself to a hard life.  Immigrants then were treated in as hostile a way as now, whether in subtle forms or brutally.  The campaign season and the election itself have been difficult for some, but now that it’s over we should all see what more we can do to first approach the world in a more loving stance, before criticizing others.  This site too took a little time to not rush to judgement despite the pain these episodes caused on friends.  In the past decade, there have been millions of displaced Middle Easterners, casualties of a civil war that Western government was not completely independent of.  At the same time there were tens of millions of Americans who were negatively impacted by actions foreigners have had (whether through terrorism, or taken-away economic opportunities, etc.)  Very few of us are wholly immune from the common push and pull of how we think about the environment around us and the societies we want to raise our children in.  We are all legitimate in our views, we must see our common interests to excel together.  Even the best among us has had to oft times accept slightly less than 100% of their wants to move ahead.

There are high quality hate data collected by major police departments, which ultimately are grounded into the FBI’s database report (considered by most the “gold standard” in this arena).  Even then the latter data suffers greatly from a volunteer intake of adversely self-selected police departments and lack of consistency among departments on how to classify hate crimes versus other crimes.  Using the largest police department’s year over year comparisons do provide one of many inputs to assess where the crimes are headed and they and other of our data fit together in size with the FBI's most recent trending.  So for New York -see below- we can graph the year to date levels of diverse hate crime (red for 2017 and orange for 2016).  While “anti-immigrant” crimes isn’t one of their mutually exclusive categories, it sits squarely between the religious, and race categories.  Now prior to 2016, the hate crime towards Jews and immigrant Muslims was already high and worsening.  We can see the early 2016 levels were worse than the post-9/11 levels.  All the reported data should reflect to us that the first 10 months of 2017 - will all be worse than those same months in 2016!  Data and news from sources such as Propublica, Huffington Post, and other big police departments also validate the NYPD findings more generally through the nation.  Interestingly, Jews and Muslims make up a total of 3% of the U.S. population.  Immigrants make up 14%, and transgenders make up ~0.5%.  This all combines to a sub-population of 1 in 6 people in America.  And this quickly conforms to the mathematical Pareto principle: where 20% of the population are bearing the brunt of 80% of the hate crime, particularly in recent months.


Now let’s take a deeper dive on anti-Muslim hate crimes, given the similarities many suggest this era may have towards another similar period of relatively high, Muslim hate (post-9/11), connected together with our unending “War on terror”.  Despite the happy-hour shooting death at a Kansas sports bar, of someone mistaken to be an Iranian Muslim, some might still think the current climate must be more quiet for Muslims versus then.  And this is where inanely waiting for FBI reported data (which is anyway annual summaries) only to match up to more recent and similar data would be incomplete for now.  We need information related to the time about the specific impulse (e.g., 9/11, Election Day, etc.)  And so here many datasets were judiciously and more aptly formed -through arithmetic- to carve out a continuously segmented time series that we can investigate.  Along with the FBI data, we have synthesized data from police departments, justice department, major civic organizations crossing the political divide such as the Think Progress and Southern Poverty Law Center (SPLC), non-partisan think tanks such as Pew Research, and The Guardian news reports.  The data in this article now is seamlessly complementary with one another, and provide a consistent picture across time.


What we’ve done is to segment the data into 4 month periods (also it’s now 4 months since the election).  This sub-aggregation of the data allows for a more solid comparison of major and similar time blocks.  And there is no doubt from the chart above (5 evenings worth of data scraping and scrubbing and modeling), that while there was an acute spike in anti-Muslim hate crimes instantly after 9/11, what we saw after President Trump’s election was a continuous rise in Islamophobia.  There was of course the intense reaction after the election, where for days hate crime was more than a handful of times the normal amount prior to the election.  But this has somewhat collapsed back by late-November, has only returned to a still rising and ember-like growth level.  It’s only a matter of time, as we see in the trended 5-8 month future hate crime estimate, before the 2017 number of hate crimes against Muslims exceeds that from 2001!

Now another applicable topic that arises is whether some of these hate crimes are simply overblown and simply mean-spirited teasing, with no real victims.  In other words, no one is really getting killed, right?  Sort of like what some said of the facebook live kidnappers and torturers in Chicago? 

It is fair for a statistician to separate hate crimes into the severity of the incidence, and this greatly enhances data quality as well when we focus on more binary violence statistics.  In looking at the same news sources, and reading each one in recent months and tabulating them between those related to violence (deaths, threat of death, injuries, etc.) versus non-violence (property debasement, privacy invasion, verbal harassment, etc.) we see a striking pattern.  The large rise in hate crimes is nearly all related to violent hate crime.  Certainly, the ugly murders only sometimes become national news.  But for perspective, since the election, three times as many violent hate crimes were committed versus non-violent hate crimes.  And prior to the election the ratio was reversed!  The ramp up in religious hate is also among our youngest Americans, for who we should be role-modeling better. 

And thus, there is also a large statistics problem stemming from unreported crime, or even those yet to be discovered.  Say a debasement of a wall that has not yet been noticed, or the first of a succession of punctured tires without explanation.

The last topic is on looking at mapping the hate crimes about the U.S. to see what we notice about where such crime occurs (nearly in every state!)  And peculiarly we see it is proportionately (to one's own minority population) much higher in the North.  We map below the nearly thousand hate crimes in only the month since the election.    

 
One might simply assume we are looking at a random map, where more hate crime is simply a function of larger overall populations.  Yet a more sophisticated insight derives from scrutinizing this information and other demographic data about this, is to regress the information, naturally by population weights! What we see is that the most populated states (e.g., California, New York, Florida, Texas, Illinois, etc.), not necessarily the most Hillary Clinton leaning or sanctuary-base states, have had the lowest hate crime rate among their immense immigrant population!  Not economic status or some other explanation. 

So the curious findings here is that as hate crimes have surged in recent months, and maintains at an unnervingly brisk pace, those who probabilistically should be the more concerned for themselves are those minorities simply living in isolation from other minorities (and in less populated states).  Put differently, the economics from the hate crime supply and demand function breaks down, and we see a copious amount of blunt data here -over time- that an immigrant in a red-state such as Texas, may be less worried versus an immigrant in one of the smaller and blue, New England states.
 
And regardless of where specifically such hate crimes are occurring, we should all agree that the aggregate crime level across our neighbors is unsustainably high and we must do what we can to repress it.  We are all impacted as we all have friends or loved ones who are directly effected, and the situation is complex, yet drearily unending.

No comments:

Post a Comment