Graduation Rates and White Social Pathology

by James A. Bacon

Over the weekend, I heard a story that a teacher from Wilkes County, N.C. in the foothills of the Blue Ridge Mountains, told about discipline problems he had encountered at school. My friend kept us spellbound as he described how a trouble maker, whose mother had taught him “not to take nothin’ off nobody,” engaged in a series of physical confrontations with students and teachers. Our initial reaction was, wow, what a jerk. But as my friend delved deeper into the tale, we gained a little sympathy. The kid’s parents were separated and his mother turned tricks in the back seat of the car. One day, not long after the incidents at school, the father returned to the house. He shot the boy in the stomach, shot the mother in the head, killing her, and then put the gun under his chin, and killed himself.

The story brought to mind Charles Murray’s book, “Coming Apart,” about the social and moral disintegration of the white working class in America. There has always been a violent streak to Southern white working-class culture, of course, so a single tragic anecdote means little in itself. But having blogged recently about Virginia’s dropout rate (which is actually improving) and the fact that Virginia’s immigrant population has achieved greater success (whether measured by education, unemployment or income) than native Virginians, I wondered if Murray’s thesis accurately described what is happening here.

I found myself perusing Virginia Department of Education statistics on public high school graduation rates for the class of 2012. The broad patterns are familiar to everyone. Women graduate on time at a higher rate than men: 90.6% compared to 85.5%. Whites graduate at a higher rate (90.8%) than blacks (82.7%) and Hispanics (80.9%) but at a lower rate than Asians (94.7%).

What was interesting and counter-intuitive, however, was what emerged from taking a jurisdiction-by-jurisdiction look at graduation rates. While the statewide pattern of women graduating at higher rates than men held true for the vast majority of localities, the pattern of of whites graduating at a higher rate than blacks was far less likely to do so.

Statewide, there is a 5.1 percentage point differential between the graduation rates of women and men. Women graduated at higher rates in every Virginia city and county save 15, and in nearly every one of those the difference was minimal. By contrast, there was an 8.1 percentage point differential in the graduation rate between whites and blacks statewide — yet blacks showed a higher graduation rate than whites in 31 jurisdictions.

In some jurisdictions with higher black graduation rates, the differences were miniscule. In others, blacks represented a tiny minority in a predominantly white student body, comprising only 11 individuals, for instance, in Grayson County, where all 11 (100%) graduated compared to only 84.6% of their white classmates. Perhaps those can be dismissed as meaningless statistical anomalies.

But what do we make of Caroline County with black and white student bodies of roughly equal size, where the black-white graduation gap was 9.0 percentage points — in favor of blacks? Or Cumberland County where the graduation gap was 10.3 percentage points in favor of blacks? Or Nottoway County where the gap was 14 percentage points? Or Lunenburg County where the gap was 19.6 percentage points? All statistical flukes? I doubt it.

Across Virginia’s suburban and inner-city jurisdictions the pattern of higher white graduation rates prevail almost uniformly. The exceptions to the rule occur primarily in Virginia’s rural counties where, the stereotype holds, racial prejudice and discrimination are the strongest. These numbers would seem to demolish that widely held view. As a fallback position, one might argue that rural counties are equal in their poverty. How, then, does one explain a Nottoway or Lunenburg County? How does one explain such superior outcomes for blacks?

To answer that question brings us full circle. I would love to think that it means rural blacks are prospering. Given the general state of the economy, however, I find that difficult to believe. Alternatively, one must ask, do these numbers reflect a breakdown of white, working class families and value structure? It would be dangerous to draw hard-and-fast conclusions from one year’s worth of graduation data. Any serious inquiry would require examining standardized test scores, unemployment rates, incomes, teenage pregnancy rates, rates of substance abuse and rates of incarceration and tracking trends over time. But such an effort might yield fascinating conclusions.

Share this article


(comments below)


(comments below)


  1. Peter Galuszka Avatar
    Peter Galuszka

    What an unspeakably tasteless photo given an otherwise interesting post.

    Jim, you really do need an editor.

  2. As far as photos are concerned, Peter wins hands-down all of the time.

  3. I like the fact that Jim clearly deeply cares about education in his overall perspective of the world we live in.

    the data is provocative and begs further consideration although I find myself a bit skeptical…

    I also think back to a agrarian country where sod-busters were largely uneducated and decided that education of their kids was important enough to pay taxes for and thus began the seeds of modern world public education.

    children in K-3 are largely innocent of their circumstances and their fate. They all learn if there is a decent teacher who is also perceptive enough to discern the child’s family circumstances and perhaps put a little extra on task for that child.

    K-3 teachers will tell you that at some point, they can see the family forces starting to take effect on a child who initially loved learning but then, inevitably, gets drawn more into family traditions – which are not all wonderful things in the world of traditions.

  4. oh.. and the kids in K-3 who steal other kids lunch money – usually do not end up the way we’d all hope.

    the bigger question is – why is a kid stealing other kids lunch money to start with – in the 2nd or 3rd grade.

    You say you don’t want teachers teaching “character”?

    but I bet you want them to stop stealing and bullying, eh?

  5. DJRippert Avatar

    The difference in graduation rates will come down to religion. Churches have remained racially segregated across America. I’ll bet you that the places where one race substantially out performs the other in rural America is a function of religion. In many rural areas a strong church or churches set the moral tone for the congregation. This, in turn, affects the rate of divorce, substance abuse, etc. And divorce (especially in families of modest means) affects the kids.

    In some places, the white churches have fallen apart. Divorce rates among rural whites approach those in urban and suburban areas. In those same places, many black churches continue to thrive. The African Americans in those areas are more likely to have strong nuclear families.

    This is just a hunch but I’d be willing to bet a six pack on that hunch.

    1. Interesting theory. I wish I could find the data to test it.

      1. DJRippert Avatar

        Go to Lunenburg County on a Sunday morning and visit the churches? About 12,000 people live in that county and about 40% are black. That’s just under 5,000 African Americans. If the black churches are full and the white churches are mostly empty – you’ll have a start.

        1. reed fawell Avatar
          reed fawell

          Rippert is on the nose yet again.

  6. used to be Catholic schools did well for both races… hmmmph

    oh by the way, I did to agree with TMT about Peter’s photos…

  7. DJRippert Avatar


    You may have to spend a couple of Sundays in Lunenburg County. There are 59 churches serving the 5,000 households in the county.

    That’s an average of 84 households per church.

    There are 140 churches listed in Henrico County serving the 108,121 households in Henrico County.

    That’s an average of 772 households per church.

    Isn’t that interesting?

    1. Interesting data, but ambiguous. The larger number of households per church simply could mean that Henrico County has more mega-churches (there are several 1,000+ congregation churches near where I live) and that Lunenberg has a lot more small-congregation country churches. Still, I agree with you, the numbers suggest that there *could* be greater religiosity in Lunenberg.

      Next step: See if there is a correlation between church attendance (or non-attendance) and indicators of social dysfunction like teen pregnancies, divorces, dead-beat dads, etc.

  8. re: ” If the black churches are full and the white churches are mostly empty – you’ll have a start.”

    I think we have veered sharply off course here in a very weird way.

    If you drive to a rural county in Va and you see a church and it’s maintained – then it has a congregation.. that shows up on Sunday’s and puts enough money in the collection plate to keep the church maintained.

    I’ve seen no downhill degradation of white rural churches in Va.

    What I HAVE noticed is that churches remain largely white or largely black.

    that evangelical, “primitive” baptist, snake-handling, etc tend to be the white churches.

    this whole line of thought here: ” Graduation Rates and White Social Pathology” is wickedly weird given the Massive Resistance history of Va where blacks were continued (from the Civil War on ) to be denied access to public education and whites continued their access to public education.

    I think you’d need to further sub- divide rural Virginia as the western and south western parts of Va are culturally different from the South and south central parts of Va where there were much larger populations of black folks.

    check out where the Booker T Washington National Monument is located and draws thousands of school kids every year:

    37.1217 -79.7337

    1. DJRippert Avatar

      This is a good survey:

      You’ll notice that the number of respondents who claim “no religion” is 7% in the so-called black belt and 25% in Appalachia. Given the geographic proximity of the black belt to Appalachia, that is a very telling statistic.

      Southern Virginia is split between Appalachia and The Black Belt. In many cases, one county falls in one category while the adjacent county falls in the other.

      Jim’s post noted differences between white and black graduation rates at a county by county level.

      In areas where money is scarce, I believe that divorce and children born out of wedlock are major contributors to poverty and high school dropout rates. In turn, I believe that strong religious affiliations tend to discourage divorce and out of wedlock pregnancy.

      The final question is whether there are differences in the strength of religious affiliation by race. Given the results of the Casey Report in the “no religion” category it seems that white dominated Appalachia is almost four times more likely that the African American dominated Black Belt to have “no religion”.

      1. Hmmm, more interesting data. The Ripper’s hypothesis strengthens…

  9. HillCityJim Avatar

    Dr. Roland Fryer has been studying the B/W achievement gap for years and is someone worth following. Read his bio and view his lecture he presented this Spring at Ringling College in Florida. Jim, you will like it when he says “its not the money.”

    1. Fryer is doing interesting work. For those who don’t have time to watch his whole speech, this Forbes article — — summarizes some of his key points.

  10. re: religion, poverty, graduation rates and out-of-wedlock

    I’m flummoxed… I admit it. I would never, ever have come to this conclusion…

    does the data show this for all the counties in a region – like Appalachia or is it only some counties?

    re: “studying the achievement gap”.

    we ought to be clear what the traditional “gap” is thought to be.
    and what the “gap” in this blog post is asserted to be. right?

  11. re: the Carsey Institute:

    seems to be a stand-alone type organization as opposed to an organization contributing to a general body of knowledge:

    from Wiki entry for the Institute:

    This article relies on references to primary sources or sources affiliated with the subject. Please add citations from reliable and independent sources. (August 2010)

    The topic of this article may not meet Wikipedia’s general notability guideline. Please help to establish notability by adding reliable, secondary sources about the topic. If notability cannot be established, the article is likely to be merged, redirected, or deleted. (August 2010)

    a little investigation on GOOGLE will show that the institute is not widely referenced by other institutions doing similar research and studies – a potential red flag for me.

    For me to accept a study as credible – it has to have some level of replication by other organizations.

    Anyone can assert something and have data to support it. I want to see it looked at by other organizations and especially organizations that may not necessarily agree with their premise.

    In particular – the Pew Research Folks among others.

    I’m not writing them off, but I have questions that need answers.

  12. DJRippert Avatar


    I am officially promoting you to the ranks of “Problem Admiration Society” member.

    I guess I knew the religion angle would rankle some of the commentators. Maybe I should have stuck to the growing rate of divorce in rural America and left the religion aspect aside.

    However, that wouldn’t explain a statistical difference between white and black graduation rates.

    So, I ask myself, “What’s the least integrated aspect of rural society?”. The schools are integrated, businesses are integrated, the workplace is integrated, small towns seem (mostly) integrated. Religion is segregated. And religion has historically been a big factor in the lives of rural Americans. This is especially true with regard to attitudes toward divorce and out of wedlock pregnancy.

    So, LarryG – you know my hypothesis. If you have an alternate hypothesis, please state it.

  13. I do not have an alternate hypothesis but I’m not yet convinced by the premise which is pretty ambitious I think.

    you have to be careful about causation from data… it helps to have several different respected and credible organizations supporting it.

    If you go back and look at the other regions in the group you selected from, you’ll see even wider disparities in religiosity.

    but more than that – graduation rates for places other than the rural areas you selected – to include suburban and urban grad rates.

    I’m still a skeptic and no I do not think there is a global conspiracy.

Leave a Reply