Network News

X My Profile
View More Activity

Must-read new report on high school dropouts

I have long considered high school drop-outs not only the least soluble of our education problems but the least clear. School districts have traditionally fudged the numbers, reporting their drop-out rates as only 5 or 6 percent, a grossly deceptive one-year rate.

The National Governors Association and other policymakers, ashamed of this charade, have put an end to it. Everyone is switching to a four-year drop-out rate, the percentage of ninth-graders (about 31 percent nationally) who do not receive diplomas four years later. The improved data has not only raised the level of the debate but also made possible a new report with some unnerving revelations about graduation rates.

My wife made the mistake of letting me go with her to her office last Sunday to catch up on work. While there I read the new Education Week report, “Graduation by the Numbers: Putting Data to Work for Student Success,” and kept squealing at one statistical surprise after another. I insisted on reading each one to her, delaying her efforts to get back outside on a nice weekend day.

The report’s prime author is Christopher B. Swanson, vice president for Research & Development at Editorial Projects in Education Inc., the nonprofit that owns Education Week. (Bias alert: I am on their board.) Swanson is a national expert on dropouts. This is some of his best work. He has mined the latest data in remarkable ways.

He discovered, for instance, that just 25 of the 11,000 U.S. school districts with high schools accounted for one out of every five students who failed to graduate in 2007, the most recent year with relevant data. Those 25 districts at the top of the dropout scale had a quarter million non-graduates, as many as were counted in the lowest ranked 8,400 districts.

Here they are, followed by their number of non-graduates in 2007:

1. New York City, 43,643
2. Los Angeles, 42,174
3. Clark County, Nev. (Las Vegas) 17,479
4. Chicago, 16,731
5. Miami-Dade County, 13,261
6. Philadelphia, 9,324
7. Houston, 9,266
8. Broward County, Fla. (Fort Lauderdale), 9,093
9. Detroit (class of 2009), 8,754
10. Dallas, 8,054
11. Charlotte-Mecklenburg, N.C., 6,386
12. Hillsborough County, Fla. (Tampa), 5,773
13. Hawaii (statewide district), 5,731
14. Orange County, Fla. (Orlando), 5,656
15. Palm Beach County, Fla., 5,507
16. Prince George’s County, Md., 5,426
17. Gwinnett County, Ga., 5,115
18. DeKalb County, Ga., 5,073
19. San Bernardino City, Calif., 5,051
20. Baltimore, 5,047
21. Duval County, Fla. (Jacksonville), 5,002
22. San Diego, 4,836
23. Milwaukee, 4,680
24. Albuquerque, N.M., 4,637
25. Pinellas County, Fla., 4,280

The problem is so concentrated that a concerted effort to raise graduation rates in just those districts could bring significant improvement nationally. New York being number one is no surprise. It is the biggest district, with 1.1 million students. The more students you have, the more dropouts you are likely to generate. But notice, Swanson wrote, that “despite its smaller size, the 678,000-student Los Angeles Unified generates a comparable number of dropouts, owing to a graduation rate 14 points lower than in New York City.”

I think Los Angeles’s higher dropout rate could be related to its higher portion of recent and often illegal immigrants, a factor that connects with another of Swanson’s revelations -- a 140-year history of four-year graduation rates in America.

The long graph stretches over two pages below Swanson’s article, “Progress Postponed.” It shows the percentage of high school graduates among those old enough to receive diplomas from 1870 (2 percent) through 1900 (6.4 percent), through 1940 (50.8 percent), through 1969 (the peak year, 77.1 percent), to 68.8 percent in 2007. I had never before seen this laid out with such clarity.

Why have graduation rates been declining since 1969? The timing suggests to me that the surge in immigration has something to do with it. Immigrants account for a significant number of poor people in the country. Poverty correlates with leaving high school before graduation. When I called Swanson about this he suggested an additional factor. Graduation rates were high in both high school and college in the late 1960s, perhaps a temporary phenomenon caused by even the most restless male students of my generation not wanting to expose themselves to the draft and the Vietnam War.

There’s more. I tend not to blame schools for large dropout rates. The higher percentage of students from impoverished families, the higher the school's dropout rate is going to be. The educators have little or no way to change the circumstances of their students’ family lives. But Swanson and his staff analyzed the demographic factors associated with dropouts and published a chart that makes me wonder if I should be so quick to excuse the schools.

Swanson identified a pool of urban districts that matched the profile of the nation’s largest urban school systems and then singled out those whose graduation rates were significantly higher than what would be predicted from their size, poverty level and other characteristics. Above, I identified the 25 districts that produce the most dropouts. Here is the flip side of that, districts that seem to have had success reducing the dropout rates despite my assumption that they couldn’t. The district names and locations are followed by the number of percentage points their graduation rate in 2007 exceeded what was predicted:

1. Newport-Mesa, Newport Beach, Calif. (29)
2. David Douglas, Portland, Ore. (20)
3. Texarkana, Texarkana, Tex. (19)
4. Memphis, Memphis, Tenn. (18)
5. Visalia, Visalia, Calif. (18)
6. Jonesboro, Jonesboro, Ark. (18)
7. Mesa, Mesa, Ariz. (16)
8. Hamilton County, Chattanooga, Tenn. (16)
9. Madera, Madera, Calif. (15)
10. Phenix, Phenix City, Ala. (14)
11. United, Laredo, Tex. (14)
12. Fort Smith, Fort Smith, Ark. (13)
13. Evansville-Vanderburgh, Evansville, Ind. (13)
14. Hemet, Hemet, Calif. (12)
15. Riverside, Riverside, Calif. (12)
16. Ferguson-Florissant, Florissant, Mo. (12)
17. Cumberland, Fayetteville, N.C. (12)
18. Long Beach, Long Beach, Calif. (11)
19. Little Rock, Little Rock, Ark. (11)
20. Muscogee County, Columbus, Ga. (10)
21. Warren Township, Indianapolis, Ind. (10)

The most eye-catching charts and revelations are in the back of the report, but “Graduation by the Numbers” has good articles in front, too. The Stockton, Calif., schools, for instance, reassigned several office staffers to check the whereabouts of students listed as dropouts. They interviewed neighbors. They pursued leads. They tracked down large numbers of students who had not left school, but had moved to another area. That, and improved programs to keep students in school, reduced the district’s dropout rate from 52.5 to 17.7 percent in just one year.

I am reading, and rereading, the whole report. It provides a useful context for just about everything I have ever read about dropouts. If this is the hardest problem to solve in education, we are going to need the best data we can find, and Swanson has provided a lot of it.

Read Jay's blog every day at

Follow all the Post's Education coverage on Twitter, Facebook and our Education web page,

By Jay Mathews  | June 10, 2010; 1:00 AM ET
Categories:  Trends  | Tags:  140 year history of drop-outs, 21 districts that have reduced drop-outs, 25 districts that produce one of every five drop-outs, drop-outs, high school graduation rates  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: Look for fun, not facts, on your campus visits
Next: The principal who created a wellspring of innovation


The wonder is that more kids don't drop out.

They're compelled under law to attend, overseen by people who place wildly varying importance on education from "who cares?" to "very", treated with a degree of arbitrariness that argues against the importance of the, supposed, reason for which they are compelled to attend and any measurement of attainment is strictly a function of their own skills and circumstances.

Posted by: allenm1 | June 10, 2010 7:51 AM | Report abuse

Heckman and LaFontaine have done incredibly important work on dropouts. Here are a few links.

video lecture:

Posted by: redhouse18 | June 10, 2010 10:51 AM | Report abuse

Dude, we here in Florida are not surprised at the results: former Gov. Jeb Bush can take partial credit for this amazing outcome. We in education also partner extensively with the Department of Corrections to ensure a smooth transition for our non-graduating school leavers.

Posted by: pittypatt | June 10, 2010 11:49 AM | Report abuse

Mr. Matthews, am I missing something? Where did Education Week obtain their data?

Below are links/data (more current then 2007) that I found from the state of MD school system website. It shows very high trends of graduation rates for two of the larger school systems:

2009 Graduation Rate MoCo Public Schools: 87.38%
2009 Dropout Rate: 2.72%

2009 Graduation Rate PGCPS Public Schools: 84.56%
2009 Dropout Rate: 1.34%

Link to Graduation Trend MoCo PS from 2002-2009:

2009: 87.38
2008: 89.08
2007: 90.37
2006: 91.58
2005: 91.43
2004: 91.95
2003: 92.45
2002: 91.81

Graduation Trend PGCPS from 1996-2009

2009: 84.56
2008: 83.09
2007: 84.88
2006: 86.56
2005: 86.82
2004: 86.72
2003: 89.54
2002: 89.03

Posted by: TwoSons | June 10, 2010 11:57 AM | Report abuse

For TwoSons. This is quite weird. It looks like neither district has switched to the 4 year system in their reporting yet. It must take some time to get the data collection system adjusted. Their grad rates, particularly in PG, wont look this good once they do.

Posted by: Jay Mathews | June 10, 2010 12:50 PM | Report abuse

Mr. Mathews,

This comes straight from the State of Maryland School System Website.

There is a big difference between what MSDE is showing vs what you've reported:

MoCo PS Dropout Rate: 9-12
2007: 2.71%
9-12 Student Total: 49,562 (Dropouts: 1,342)

PGCPS Dropout Rate: 9-12
2007: 3.82%
9-12 Student Total: 48,093 (Dropout: 1,838)

Posted by: TwoSons | June 10, 2010 1:01 PM | Report abuse to PGCPS Dropout Rate (it seems that I included MOCO links twice above):

PGCPS Dropout Rate: 9-12
2007: 3.82%
9-12 Student Total: 48,093 (Dropout: 1,838)

Posted by: TwoSons | June 10, 2010 1:07 PM | Report abuse

Maryland State Department of Education's Definition of Dropout Rate:

Dropout Rate

The percentage of students dropping out of school in grades 9 through 12 in a single year.

The number and percentage of students who leave school for any reason, except death, before graduation or completion of a Maryland approved educational program and who are not known to enroll in another school or state-approved program during the current school year. The year is defined as July through June and includes students dropping out over the summer and students dropping out of evening high school and other alternative programs.

The dropout rate is computed by dividing the number of dropouts by the total number of students in grades 9 - 12 served by the school.

Note: Students who re-enter school during the same year in which they dropped out of school are not counted as dropouts.

Reported since November 1990: System and State levels.

Reported since November 1991: School.

Additional information may be found on the School Improvement in Maryland Web site at and the Maryland
State Department of Education Web site at

Posted by: TwoSons | June 10, 2010 1:22 PM | Report abuse

Sorry Mathews,

But the only "wierd" thing that seems to be observed in your 1-25 are school districts/systems with the HIGHER PERCENTAGE OF MINORITIES are look pretty bad in drop out numbers, when, in fact, it MAY NOT be the case.

I hope all who are viewing your blog and data within, does their individual research relating to their respective school districts.

This information seems extremely flawed and against majarity minority school systems.

Posted by: TwoSons | June 10, 2010 1:48 PM | Report abuse

But the only "wierd" thing that seems to be observed in your 1-25 are school districts/systems with the HIGHER PERCENTAGE OF MINORITIES are look pretty bad in drop out numbers, when, in fact, it MAY NOT be the case.
way too many typos in prior post..sorry...

But the only "wierd" thing that seems to be observed in your 1-25 are school districts/systems is that they are HIGHER PERCENTAGE IN MINORITIES which look pretty bad in drop out numbers, when, in fact, it MAY NOT be the case.

I hope all who are viewing your blog and data within, does their individual research relating to their respective school districts.

This information seems extremely flawed and against majority minority school systems.

Posted by: TwoSons | June 10, 2010 1:50 PM | Report abuse

for TwoSons--thanks for all of the data. As you said, they are looking at a one year dropout rate, not a four year rate, which is the new national standard. Research has shown for years a very high correlation between high poverty levels and high dropout rates. Nationally the average 4 year high school graduation rate is about 70 percent, but it drops much lower for urban districts with lots of poor kids.

Posted by: Jay Mathews | June 10, 2010 4:17 PM | Report abuse

Mr. Mathews, one would appreciate how the numbers were calculated (or created) over a four year period. Do you have a link to share?

Btw...I don't recall stating "they are looking at a one year dropout rate" I just shared the data thats available via MSDE.

The graduation trend that I viewed, especially for MoCo and PG doesn't seem to show this level of students not graduating. 9-12 nongraduates and definately not a difference between THOUSANDS of students not graduating between these two school systems of similar size.

Both MoCo & PG has had a downward trend since 2002 (although year-to-year) based on information above. How is it possible that MoCo lead the trend of actual graduates if dropout numbers have increased since 2002?

Again, I must be truly missing something.

Posted by: TwoSons | June 10, 2010 4:46 PM | Report abuse

Jay has himself long been responsible for confusion by promoting an indicator of quality which is taken to be a rate :AP classes divided by number of HS seniors. Of course, he's only a journo, not a statistician. And the "Challenge Index" is called just that, an index.
But, these problems of increasing and measuring school completion rates have been with us and taken seriously by statisticians at the Census Bureau and other agencies for well over half a century.

There is some irony that a high fraction of the population has surrendered privacy and confidentiality in commercial and social networking affairs while we are still hamstrung in getting good estimates of school completion by old fears of governmental invasion of privacy and laws that forbid population registrars. If every agency entitled to confer a HS or HS equivalency diploma were required to record and report the confidential birth-certificate data on the awardee, excellent geographically- specific birth-cohort-specific estimates would soon be available and maintained. Several OECD nations have had that capacity for two or more generations.

Posted by: incredulous | June 10, 2010 5:06 PM | Report abuse

School system numbers/percentages change from year to year.

One high school system could have
50,000 students @ 10% dropout rate for one year is still 5,000

Another high school system could have
10,000 students @ 10% dropout rate that equals 1,000 students

The impact of nongraduation numbers are based on school system size. Is it fair to compare Howard County dropout rate (much smaller school system) with Montgomery County? No. How about a school system that has a higher population of ESOL or Special Need students and comparing another school system of similar size that has a lower levels of diversity and more private school studeny enrollment?

Student population will differ from year to year to include student enrollment (increase/decrease) which are not calculated (or made available) until the next SY. Still then there are marjins of errors in enrollment numbers due to mid year withdrawal (because of family relocation), school boundary and zoning changes.

Some HS grades are 10-12, 9-12 and 6-12. So, how is it possible to calculate an entire school system if, within the 4 year period, students are in a middle school that houses 7-9th graders or a high school that houses 6-12th graders within that same school system?

I truly would be interested where the figures within the lists above came from and how they were actually calculated to create these whole number of nongraduates over a 4 year period.

Posted by: TwoSons | June 10, 2010 6:13 PM | Report abuse

Mr. Mathews, I tried to leave a post early this morning - got the awaiting approval note. Did it land in the Potomac?

Posted by: shadwell1 | June 10, 2010 7:53 PM | Report abuse

I just think that districts should start early to keep kids involved in school. We can tell who at risk kids are. The kids need mentors and activities to keep them involved.

Posted by: celestun100 | June 10, 2010 8:24 PM | Report abuse

I would have to agree that the number of immigrants has added to the drop out problem. I would also have to say that my personal experience with public school (K-12) in is compulsory babysitting, little to no meaningful teaching or learning is going on. The attempt to mainsteam, the rising number of non-English speakers and the fact that parents simply don't get involved. I see this in the local public schools where I live in San Diego county. I'm not holding my breath that it will improve anytime soon.

Posted by: kodonivan | June 10, 2010 10:11 PM | Report abuse

Their grad rates, particularly in PG, wont look this good once they do.

Posted by: Jay Mathews | June 10, 2010 12:50 PM | Report abuse

whatever Mathews.

The fact remains that MoCo graduation numbers are dropping and PGCPS is closing the gap between the two school systems when looking at current graduate percentages.

It's obvious by your post that your bias toward PGCPS is continuous and also unfortunate.

Posted by: PGCResident1 | June 10, 2010 10:36 PM | Report abuse

and would like to add, less $$ per pupil cost toward PGCPS vs MoCo while the gap is closing between the two systems.

PGCPS students, parents, teachers and administrators work hard to obtain our gains and will continue to do so.

Doing more with less (for decades) is not an easy task but the work is getting done and positive progress WILL continue.

Posted by: PGCResident1 | June 10, 2010 10:46 PM | Report abuse

Jay --

Your enthusiasm for Chris' work is evident! Fun to see you so excited about the report...

I agree that the report -- which I have read cover-to-cover -- has some interesting information. But the report has a major problem: Chris used attrition rates and called them dropout rates.

Attrition rates are *not* dropout rates.

I understand the desire to do state-by-state comparisons (everybody loves 'em), yet these comparisons cannot be performed until all states upgrade their data systems. In the meantime, the desire to perform the comparisons drives researchers like Chris to use attrition…which is akin to using the lowest common denominator.

Posted by: EdDude | June 11, 2010 2:17 AM | Report abuse

A complicated subject. One more angle: the startling rise in the number of kids diagnosed with ADHD should not be overlooked when discussing the drop out rate. Scientists are not clear on whether early TV viewing has played a role, whether chemicals, such as pesticides, have led to this learning disability, or whether it is simply more widely diagnosed now.

But one thing is clear: the drop out rate is extremely high among kids with this particular learning profile.

My son, a high schooler, struggles with ADD, and it is taking all of our resources, including plenty of money for tutors, therapists, and private school, to get him to the finish line. How many families can do this?

Posted by: trace1 | June 11, 2010 7:43 AM | Report abuse

What happens when you translate those totals for the highest dropout numbers to percents? after all you are reporting some of the largest districts in the country here. I know that my local district here is central Florida (not Orange county) has about the total number of students that are reported as drop outs in NY and yet my daughter's HS had a drop out of between 40 and 50%. Of 500 freshmen over 200 did not graduate. Is 43,00 students 40-50% of NY graduates?

Posted by: kmlisle | June 11, 2010 10:20 AM | Report abuse

Be careful what you write, Jay. Your words are inspiring crackdowns in areas that don't need to be cracked-down. As the teacher of ESOL students, I often see fantastic students marked as "drop-outs" because they aren't able to graduate within four years of entering high school. This is very often the case. Just as "two sons" mentioned, your statistics are skewed. Despite how hard my students work to graduate, local politicians would rather they graduated with inferior "standard-level" (not college-friendly) diplomas rather than stay the extra year so that they can take courses such as English 12 and graduate with an Advanced Diploma. Please don't fuel this unhealthy and discriminatory fire. I already see it every day. I see it when guidance counselors cut ESOL students from college-prep programs because they're "taking too long" and I see it when students aren't pushed into higher level courses. The discrimination has to end and you can be a part of stopping it. This article needs to be amended.

Posted by: nat15222 | June 11, 2010 6:38 PM | Report abuse

Why don't we talk about how Rhee is addressing the dropout rate in DC. She is tying the summer employment with attending summer school.

Now students who decide to skip school, walk the halls, etc can wait until summer and actually get paid to attend summer school where the amount of work given to a regular class pales in comparison.

We are rewarding students with money for acting poorly during the school year. What is the incentive for anyone to take regular classes to graduate?

Also, where is the concern about the fact that these kids are graduating barely literate with very little work ethic?

Sadly, these concerns are pushed to the side in the mad dash to increase graduation rates so Rhee can prove how she is making huge gains with DCPS. I guess this is another example of low hanging fruit?

Most people will fall for this charade and continue to sing her praises as a miracle worker.

Posted by: letsbereal2 | June 12, 2010 8:46 AM | Report abuse

Demographics of individual school districts aside, I have to wonder if the focus on standardizd testing isn't a contributor to declining graduation rates?

I also have to wonder if the continuing criticism of teachers and the public schools isn't another contributor. As I have often said in this comment section, America has done a lousy job promoting the importance of education....we are putting all the ownership of public education on teachers.

As long as we continue to trash public schools and public school teachers, why would kids respect the need to be educated an education...especially those kids who come from homes where studying and being educated is not given a high priority?

Does anyone buy a product that is, on a consistent basis, rated as inferior?

Posted by: ilcn | June 13, 2010 10:34 AM | Report abuse


As much as I admire Swanson's work, there are benefits and problems with his graduation rate index. For instance, my district had a 39.9% graduation rate during the 1991 recession. But when it opened magnet schools, the rate jumped to the mid 40% range where is grew slowly until last year. So, some students weren't dropping out but moving to the suburbs. So, the graduation rate was a hybrid between a dropout rate and a customer satisfaction rate.

As long as we can subdue the blame game and keep contradictortory thoughts in our heads at the same time, we can benefit greatly from Swanson's research. But last year, a change in the State's defintion meant that our graduation rate soared by nearly 30 points and the district refuses to attach an asterisk to it. My school's Improvement Plan is an example. In one paragraph they cited a one-year dropout rate of 33.7% from two years ago (which would proeject out to 80 something perecent over four years) and then our currect graduation rate of 97%. The author of the plan isn't stupid. That's just how you play the game in a culture of compliance.

Of course, we're just the minor leagues. For real dishonsty go to Texas and Klein and Rhee.

Posted by: johnt4853 | June 13, 2010 11:31 AM | Report abuse

Dropping out is due to many factors, some of which are an emotional turn off to school. It should not be handled as a purely academic problem. Yes, there are academic problems involved, but that is never the only issue.

I feel students who are at risk for dropping out need encouragement. I feel this is one of the most solvable issues in education. I don't think more emphasis on standardized testing is the answer.

Many times kids with serious problems outside of school can express themselves through creative writing in their own voice (which may have to be in a language other than standard English for some kids), art and music. Those programs are cut back in many places. But those are the programs that the at risk population needs to stay in school.

Sometimes the kids who are going to drop out just need to make a connection with one teacher to stay in. It is often the creative writing, ethnic studies, home economics, computer applications, art, or music classes that are deemed "unimportant" that are the reason that some kids keep coming to school.

Think about your own worst personal crisis. Then think about how you felt when somebody (maybe at work) tells you "No excuses." Maybe you threw yourself into work or school to help forget it. But what if your crisis went on for years?

What would a normal human reaction to somebody who was telling you, "Sorry, you like art, but it is just a "fun" class. You need to spend more time on English Lit (substitute here any subject you dislike), so we are giving you two periods of that class."?

I think many, many people would want to quit and if nothing were stopping them, they would.

Posted by: celestun100 | June 13, 2010 11:36 AM | Report abuse

I have to add that the arts, music, languages are always considered expensive. But they are actually a very good deal for taxpayers. Good elective programs probably save an incredible amount of taxpayer money in the long run. Less drop outs mean less crime. Them more educated the population is, the better the society.

Posted by: celestun100 | June 13, 2010 11:41 AM | Report abuse

Sorry, "them" should be "the" above.

Posted by: celestun100 | June 13, 2010 11:44 AM | Report abuse

One more thing. What is the drop out rate for religious or parochial schools? Often they are touted as having good records for academics in low income areas. But, what if the reason they are successful has to with their acceptance of the importance of emotional health? (In parochial schools this was called spirituality and prayer and while the expectations were always high, compassion was too).

Posted by: celestun100 | June 13, 2010 11:51 AM | Report abuse

Chicago? 2007? 4th highest? Isn't that where Duncan came from? Change agent? Pay-for-Performance? Data?

Posted by: ilcn | June 13, 2010 12:34 PM | Report abuse

Here is a school that is doing something about dropout rates. Learning Works charter school in Pasadena, CA employs chasers to track down kids who aren't coming to class. The school is for students who've dropped out of a traditional high school, spent time in jail, have been kicked out and are teen moms. The chasers keep them engaged in school and the reason it works is because the chasers once were in their shoes. Read the story at

Posted by: jennymancino | June 14, 2010 7:56 PM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company