Thursday, December 01, 2011

Selective use of data

Data is good in journalism. Selective use of data, not so much.

So let's look at the following story from The Nerve, a journalistic endeavor of the S.C. Policy Council. The council is a heavily conservative advocacy organization, but for those critics who would like to dismiss The Nerve's work out of hand as a result, stop.

The reporters, many of whom worked for "established" media, know their stuff and more often than not have uncovered those annoying little tidbits that send politicians into a tizzy. It also has become another media voice documenting and amplofying FOIA violations - can't be enough of those.

Yes, the stuff could use some solid editing from time to time. And it can be a bit shrill - one clearly knows the direction the writers are coming from ("Where Government Gets Exposed" is the site's tagline). But to dismiss it out of hand is ill-advised.

However, there are things like this that bother me because they show selective use of data to put a finger on the fairness scale, so to speak. And yes, I am a USC employee, and no, doesn't matter a twit to me what the subject is. We're talking basic journalism, editing and fairness in presenting data here. So consider the following. My comments and questions, had I edited this, are in brackets and italics:

As College Tuition Rises, So Does Administrators’ Pay
The body of the story bears that headline out at only one school.
By Amit Kumar
We’ve all heard it before: Tuition at South Carolina’s public universities and colleges is rising; state appropriations for higher education are falling; and it’s something that has been going on for years.

There is, however, at least one budgetary commitment that has remained constant or even increased at state institutions of higher education in recent years: total compensation paid to university presidents and vice presidents.
"Remained constant or 'even' increased" - so if it's remained constant, why is it worth mentioning?

The Nerve analyzed total compensation packages for the presidents and vice presidents at the University of South Carolina, Clemson University, and the College of Charleston, the largest undergraduate public universities in the state. The Nerve obtained compensation records by filing S.C. Freedom of Information requests with all three universities.

While tuition at all three public universities has nearly doubled in the past decade, university administrators have been receiving steady compensation packages worth hundreds of thousands of dollars, the review found.
OK, so the news is what? If administrators' salaries had been going up, the news is clear. But if their compensation stayed "steady," then doesn't that bolster the idea that the extra money's been going for other purposes (like education and more faculty)? 
 The Nerve’s review also found that, while university administrators often point to decreases in state appropriations as justification for tuition increases, the amounts of federal stimulus dollars each of these three universities received in the past two fiscal years more than offset any cuts in state funding.
More on this in a minute, but hold this question - do stimulus dollars actually replace state appropriations?

For fiscal year 2011-2012, the total compensation package for USC President Harris Pastides is valued at $535,000; for the College of Charleston President George Benson, at $398,987; and for Clemson President James Barker, at $400,000.

Each of those packages is more than 12 times the per capita income in South Carolina, valued at $33,163 in 2010 by the U.S. Department of Commerce.
OK, gives us a bit of context. But is 12 times excessive compared with similar officials in other states (which also have been cutting budgets) or with other S.C. officials? Has that ratio increased or decreased over time - that's a major weakness throughout the story; the real issue is whether there have been increases, and have those been excessive? I know it's hard to feel sympathy for someone making $300K or $400K a year, but as journalists, we're supposed to know when a snapshot can be more misleading than a time series. BTW, check the state salary database, and 17 state employees show up (most from the universities, including the Medical University of South Carolina,  as making more than $300,000, none of them the university presidents because the presidents' actual salaries are much lower).
At USC, the total compensation paid to Pastides each year since fiscal year 2008-2009, his first in the post, has been constant at $535,000. That number includes both money from state government and supplemental private funds from University of South Carolina foundations.
What's the breakdown? Has the public portion of that changed? What relevance does the "private" part of that have to the argument at hand, which seems to be tuition rises but these folks keep dipping from the trough.
A report by The Chronicle On Higher Education, though, found that the median total compensation package for university presidents – including presidents of university systems with multiple campuses, like that of USC – at 185 of the top research institutions in the nation in 2009-2010 was $440,487 – $95,000 lower than Pastides’ package.
Useful to know.
Meanwhile, both in-state and out-of-state tuition for attending USC have more than doubled in the past 10 years, with in-state tuition rising from $5,024 in 2002-2003 to $10,168 in 2011-12. After adjusting for inflation, that figure represents a 61 percent increase in tuition and fees.

A common refrain from universities regarding their yearly tuition hikes is that the increases are necessary because the amount of money the state appropriates to higher education has decreased each year. And this is true: For instance, for the 10-year period from 2001-2002 to 2010-2011, annual state appropriations to USC-Columbia have declined by 45 percent, from $183.7 million to $101 million, according to the most recent report by the S.C. Commission on Higher Education.

However, for the past two completed fiscal years, state universities have received federal stimulus dollars as a response to the recession. While those stimulus funds, which are non-recurring, have more than offset year-to-year reductions in state appropriations, universities have still elected to raise tuition in those years.
OK, back to the original question - are stimulus funds the same as state funds? I don't know, but the journalist could have helped me out as a reader. Many federal funds come with restrictions, as opposed to the state's "general fund" dollars. If the fed money is not a perfect substitute, is the argument here that the fed money used for purpose "x," should have then freed up the same amount of state money for other uses, thus negating or lessening the need for a tuition increase? Would have been nice to explain.
 For example, in fiscal year 2009-2010, state recurring appropriations to the entire USC system declined by $20.5 million; that same year USC received $29.2 million in federal stimulus dollars, more than offsetting the loss in state dollars. Still, USC increased tuition by 3.6 percent for in-state and out-of-state students in 2009-2010.

Similarly, at the College of Charleston, President George Benson’s compensation has not increased since he first took the top position in 2007-2008. However, the compensation packages for five of the college’s six vice presidents have increased since 2006-2007, including three packages that have increased by more than 14 percent each after adjusting for inflation.
OK, so the story vis a vis Benson is? The five VPs is interesting. But why does the story report only VP increases for College of Charleston? What about Clemson, USC, etc.? (Note: Clemson's data comes at end - would be useful if these were grouped if the point was that VPs overall were getting raises.)
For instance, the compensation package for George Watt, the college’s executive vice president of institutional advancement, has increased by $62,687, or 20.1 percent after adjusting for inflation, since 2008-2009 – right at the height of the recession.
Sounds not so good. But what does a VP of institutional advancement do? As a reader, helps me decide whether it's reasonable.
 In the past 10 years, in-state tuition at the College of Charleston has increased from $4,858 in 2002-2003 to $9,616 in 2011-2012, or by 57 percent after adjusting for inflation. Out-of-state tuition has increased even more, by 76 percent after adjusting for inflation, or $13,356 more per year.

College of Charleston, like USC, repeatedly justifies tuition increases at least partially because of lowered state appropriations. But although the amount of general funds appropriated to public universities has decreased significantly in recent years, general funds make up only a small portion of a university’s overall budget.

For fiscal 2011-12, general funds made up only 9 percent of the College of Charleston’s overall budget. For USC, that number was 11 percent; for Clemson University, general funds were only 8 percent of its overall budget.
Perhaps it is too obvious to note that one of the reasons general funds make up "only" (a loaded word) a small part of the budgets is because the state appropriations have been cut so much? Again, trend data useful for me as a reader to determine context.
 The bulk of these universities’ budgets actually comes from other funds, which are made up of tuition and fees. Of the College of Charleston’s $220 million budget, $183.5 million, or 83 percent of the overall budget, comes from other funds. At USC, $641.8 million out of its $907.2 million budget, or 71 percent, comes from other funds; at Clemson, other funds make up $650.6 million out its $805.4 million budget, or 81 percent.
And, again, that large proportion would be a direct result of state funding being cut, right?
Universities are crying out that they need to increase tuition because their state appropriations are dwindling; but those appropriations make up only about 10 percent of their overall budgets, and cuts to those funds have been offset by federal stimulus dollars in recent years.
So synthesizing my comments: This is a collection of random facts that when put in the same graf sounds sinister but fails to answer whether stimulus dollars can be directly substitutable for state appropriations and how much state money as a proportion of the schools' budget has changed. The implied argument seems to hinge on "only," but the "only" might well be the result of state cuts, especially if the federal money is not a one-for-one replacement  - seems a bit tautological, eh?

All these same financial trends are visible at Clemson, where in-state tuition has increased by 67 percent and out-of-state tuition by 75 percent in the past 10 years after adjusting for inflation.

In that same time, Clemson President James Barker has seen his compensation package increase by $120,986, from $279,014 to $400,000 – a raise of 14 percent after adjusting for inflation.
OK, that's useful.
Some of Clemson’s vice presidents have received large compensation increases in the past decade as well, even when adjusting for inflation.
In the past 10 years, the compensation package for Doris Helms, Clemson’s vice president for academic affairs and provost, has increased by 28 percent, to $270,389 today. The package for John Kelly, vice president for agriculture, public service, and economic development, has increased by 19 percent after, to $242,732 today; and for Neill Cameron, vice president for advancement, by 15.5 percent, to $211,185 today.
OK, that's useful. So Clemson raises some flags. But when did those raises happen? Isn't it possible the bulk were before the 2008 recession? Help me out as a reader to evaluate the information. It's far more significant if they've continued getting raises even after the economy tanked.

And why don't we hear about the VPs at the other institutions - USC? So at two out of three schools, there weren't significant VP raises? But not at the largest? And how do they relate? That's cherry-picking data.

Bottom line: Might be a story here. Take your pick based on your policy/political/fiscal orientation:
  • Tuition's gone up even though the schools' financial picture shouldn't have changed because federal stimulus dollars were exactly substitutable for the state money that was cut. (Nothing in the story says that about the fed dollars; we are left to artfully conclude it.) Ergo, any tuition increase somehow is tied to the pay, although how is that possible when the pay remained steady for the presidents of two schools and we have no data in the story showing significant increases for underlings at one of those the three schools (besides Clemson)?
  •  University officials' pay has continued to rise even while the economy has gone to crap. (Unfortunately, the data we have right now shows that's the case at "only" one of three, and even then it's not clear when the increases took place.)
  • State appropriations were cut and the stimulus dollars that came in could not directly replace them. So tuition was raised, in part to cover the salaries plus any increases in those salaries over the past decade. That was wrong, since the stimulus money could have freed up other funds that could have been transferred back to other lines to cover/offset the salaries. The implication is that those top officials should somehow have refunded parts of their salaries to cover that (unspecified) part of the tuition increase.
One of these might have been true. But this story, as a "random walk" through data, doesn't support any of them. If the stimulus was not a dollar-for-dollar replacement, you could also debate things like whether the money was effectively spent, whether the universities should have added positions at such a time, did the universities properly transfer money among lines to account for the stimulus, etc.? But that's a different story than what's here.

Bottom line - numbers don't make a story credible by themselves. Careful, clear numbers with accuracy, completeness and context - and the assumptions clearly spelled out - do.

Labels: , , , ,


At 12/5/11, 9:53 PM, Anonymous Anonymous said...

I think it's fair to point out that the author of this story is not a former reporter from "established media." He is actually a "reporter intern" at The Nerve. Ironically enough, he is also a broadcast journalism student at USC.

At 12/6/11, 8:13 AM, Blogger Doug said...

Thus the "many of whom."

My larger point, reinforced by this, however, is that there is a need for good editing.



Post a Comment

<< Home