I was reading a new blog, ribbonfarm.com by the interesting Venkatesh Rao, because one post on the true organization of corporations had been linked approvingly on LessWrong; he presents a lifecycle of corporations by way of describing how a corporation inevitably degenerates into operating. They are constantly being born and then ripped apart by hungry entrepreneurs (a nicer name for Venkatesh’s ‘sociopaths’), and this creative destruction keeps things relatively efficient and cuts down on indefinite rent-seeking & other principal-agent problems like the Iron law of oligarchy. Very few organizations indeed survive more than a century.1
But of course, some corporations last a very long time. Millennia is not unheard of, and the Catholic Church resembles modern corporations in many respects. It’s interesting looking at their areas of expertise—food, construction, and religion seem to predominate.
All areas with what one might call ‘unearned’ advantages; if a temple tells you that Buddha endorses that temple and worshiping elsewhere might condemn you to the hell or rebirth as a hungry ghost, they probably didn’t earn that endorse on the free market with really good homiletics and especially effective philanthropy. If a construction company has been building the local Shinto temples for the past couple centuries, they’ll go on getting business regardless of whether they’re doing an efficient job. The candies produced near Ise will go on selling well despite transgressions like selling expired frozen goods as fresh and forging the documentation. When King Henry VIII expelled the Catholic Church and seized monastery assets, he seized something like 1⁄4 of the entirety of England2; one has to wonder if the monasteries were that good at agriculture & business. Modern corporations have seamy underbellies even if they seem utterly unobjectionable3.
So if one sees an old corporation, one is justified in asking how it has survived so long—does it have some sort of unique stable niche it alone can fill, or an unfair advantage? It might be a simple case of luck. With millions of corporations over the years, surely a few will simply happen to survive a long time.
But it seems to me that if this were the case, one would expect to see such corporations shifting niches and surviving in competitive areas for long periods; but instead, if one looks at examples like the former oldest corporation in the world, Kongo Gumi, one sees a story which runs more like this: they lived in a protected niche for an extremely long time and then they died shortly after venturing into more competitive areas; and indeed, Kongō Gumi was founded in 578 AD as a temple construction company and went bankrupt after it tried to expand into general real estate in the middle of a bubble.4 Gumi illustrates another failure mode: a single leader can draw on the accumulated capital (financial & social) of an ancient institution. How did Gumi, a relatively small construction company in a shrinking business, have enough credit to hang itself? Perhaps because it was such an ancient institution. The folly of one generation can be fueled by the prudence of many generations.
That is all background for the main question.
Presumably we would ‘blame’ a leadership if it was failing to accomplish its publicly claimed goals and serving itself instead (the fancy economics term for this is the principal-agent problem). A charity focused on character-building and recreation is a bit harder to judge than one trying to stamp out polio; the latter can just point to falling death-rates.8
There are a number of criteria we could try to judge Girl Scouts on:
Is the administrative overhead growing or decreasing?
Computerization and general technological developments means that the necessary bureaucracy ought to be more efficient than, say, in the 1950s.
Is the enrollment of Girl Scouts as a fraction of the young female population increasing or decreasing?
If Girl Scouts is shrinking, then the good the organization can do is necessarily also shrinking. It’s hard for Girl Scouts to help a lot of young girls learn self-discipline or character if they’re all fleeing.
Are camps being more or less utilized, or worse, being closed?
Camping and the outdoors are one of the key aspects of any Scouting organization and in a sense, what makes them more than a club for arts & crafts9. If they are going unused, then it’d be like a library no longer lending books—it may still be doing something useful, but one wonders what exactly.
How can we judge Criterion 1? Easily, it turns out. Girl Scouts is a charity, charities are nonprofit corporations, corporations keep records of spending, and often make them public; as an IRS-recognized charity, Girl Scouts is obligated to make Form 990 financial reports public, and some Googling then turns up GuideStar as a source for the reports, at which point it’s easy to download their PDF and finally answer the question—what does Girl Scouts spend its money on? (That charity filings are publicly available seems to be generally unknown; people routinely ask questions about charities that could be answered by looking at the Form 990s.)
Criterion #2 is harder, but total enrollment is public information, as is US census data about how many young girl there are.
Criterion #3 is difficult to research; it seems unlikely that camp attendance or closing data is easily found online, and it is possible that not even Girl Scouts National (GSUSA) knows the answer. We’ll punt on #3, and see how far we can get on #1 and #2 without looking at the tax filings.
Here are some facts to consider. Girl Scouts was founded in 1912 and is a century old. The GSUSA leadership is apparently unelected. And the current era is well-known as one of stratospheric executive compensation, which would be as easy to justify in charity as anywhere else (“shouldn’t our managers get a fair living wage? You don’t want the Girl Scouts crippled by headhunters poaching it.”) Particularly troubling are the cookie sales themselves. At first glance, they seem almost designed to concentrate control & money, rather than decentralize it:
- The council probably control the cookie sales through standard methods like copyright or trademark or trade secrets
- prices and availability reportedly are now set nationally for each area. This is a very recent development.
- Cookies have been official fundraisers and sold since the 1920s; they were commercialized in 1934, and commercial bakers used nationally in 1936
- Competing would be hard because so much of the value is in the brand (and brands are so prized by manufacturers because they are a license to print money).
- Economies of scale discourage alternatives; there are 2 official bakers in 2011, compared to 14 bakers in 1961.
- There are few fundraising alternatives: Girl Scouts sometimes sell calendars or magazine subscriptions or nuts or popcorn, but the sales of those are peanuts and most only sell cookies.
- Returns of cookies are forbidden
So, before we peek at the tax filings, what predictions would we make? It’s important to predict before we look at more data; there are too many cognitive biases which let you decide that the results confirm your theory or beliefs (even if you were just lied to). If your theory can account for any result, then it’s a pretty worthless theory. (If your predictions don’t eliminate outcomes as possibilities, then you wind up holding weird beliefs like WWII sabotage confirms the existence of Japanese saboteurs and the absence of sabotage also confirms the existence of said saboteurs, or that confessing to witchcraft confirmed you as a witch and not confessing also confirmed you. Which is daft.)
Well, one would expect a high salary for the GS CEO (>$300k seems like a sum I would start to wonder about), one would expect substantial benefits15, a large number of employees in general (empire building); a major red flag would be any employees shifting employment between Girl Scouts and the 2 licensed bakers (and any shared directors basically end our inquiry right there).
This essay for me has been something of an intellectual exercise. Without knowing the facts of the matter, going on general knowledge and calculations, I would try to make a case for corruption in GSUSA. The case so far may or may not be convincing depending on how much trust one puts in the 2% figure. The IRS tax filings would reveal important data like employee salaries, which trump anything I’ve said so far. Will my suspicions be vindicated? Or will I be revealed as a cynic or paranoiac, and the Seattle Times and others right in pointing to the councils as hogging all the money?
When I first wrote this, the 2009 was the latest filing available, so we’ll start with it.
What interesting things do I note on the first read?
the headquarters are on Fifth Avenue in NYC—what is a frugal charity doing on a street whose Wikipedia entry describes how it is the most expensive street in the world?
the ‘governing body’ has just 30 voting members—so it actually is elective (in a very narrow sense)
there are 539 employees
it received $6m in donations & grants ($1.5m government) and paid out $4.1m in similar (all to various local councils, it seems)
it received $37.7m in ‘program service revenue’
The later section (page 10) is very confusing to me. I’m not sure how to read it. It may imply that $32m came from membership dues and the other $5.1m from ‘meeting and learning events’.
investments are an odd item; in 2008 they earned $2.15m, but are recorded are -$3.9m in 2009! Did they get burned badly in the housing bubble?
More interestingly, the detailed breakout section includes an item for ‘royalties’ which must be the cookie royalties! The royalties are $8,122,461, much less than the $16m figure. They also earned $2.8m from ‘dividends, interest, other similar amounts’.
There’s one set of items I have no idea to interpret. ‘Gross amount from sales of assets other than inventory’ lists ‘(i) Securities: 53,116,009’ and then ‘Less cost or other basis and sales expenses’ lists ‘(i) Securities: 59,853,735’ for a ‘Gain or (loss)’ of -6,737,726.
Another curious entry is ‘Gross sales of inventory, less returns and allowances’ ($37.9m), ‘Less cost of goods sold’ ($16.6m) for a ‘Net income or (loss) from sales of inventory’ of $21.2m. It’s possible that this is some sort of cookie reselling scheme, but more likely it’s just GSUSA selling Girl Scout merchandise like uniforms at a healthy 57% markup. A full set of Scout equipment can be very expensive.
the ‘Salaries, other compensation, employee benefits’ expense item constitutes $37.3m; and GSUSA spent $34.6m the previous year
Here is another area where the details (page 11) begin to trip me up. ‘Compensation of current officers, directors, trustees, and key employees’ amounts to only $2.4m! ‘Other salaries and wages’ is a total of $24m. The remaining $9m seems to go to $2.7m in pension contributions, $5.7m in ‘Other employee benefits’, and $2.2m in payroll taxes.
All ‘Other expenses’ make up $44.5m
Some of them make one wonder. $489k in investment management fees seems somewhat reasonable if $14m is being held in ‘savings and temporary cash investments’, $93m in ‘publicly traded securities’ & $22m in ‘other securities’, as does $913k in legal fees; but what did they bury in the $11.8m laconically listed as ‘Other’? Did they really have to spend $6.1m on ‘occupancy’ and $3.7m on ‘travel’. It is surprising how little they pay in insurance ($305k) but then again, the insurance often cited as one of the necessary overhead expenditures probably is more a problem for the councils which actually own the camps & facilities.
the form, in case you missed it, spells out the implication: GSUSA was $17m in the red in 2009, and $7m in the red in 2008. How on earth does that work? Where is the money coming from? Am I misunderstanding something? Did they really choose to spend an extra $3m on salary & benefits even as they fell off a financial cliff?
the summary is particularly stark: Total liabilities 2008, $19.35m; 2009, $40.8m. Its net assets in 2009 amount to $118.9m.
If it continues to deficit spend at $20m a year, then it’d be down to $100m in 2010, and $80m in 2011, and bankrupt around 2015 or so. I had expected to find a large salary/
benefits item like $38m, but I had not expected to find GSUSA in such parlous state.
the checklist does have one interesting item, #28, which asks about former employees etc. having a business relationship; this was all marked no, so if there’s a revolving door, it’s using some loophole or dodge, which makes it less likely. Phew! One bit of good news.
part 7, section A covers the juicy salary information—but omits the details! All it reports is that 70 individuals are included, who collectively earned $3,793,979 in salary & benefits from GSUSA ‘and related organizations’, and at least one earned >$150,000.
Fortunately, on page 17 (part VII, section Aaa), we find that the CEO gets $435,352 and $63,058, for a total of $498,410. The VP makes $431,758, the chief of staff $310,630, the executive vice president makes $311,618, the senior vice president makes $248,979, the ‘VP & general manager’ makes $297,407, and a passel of 7 other vice presidents of various sorts make similar sums in the 300-190k range.
On page 42, we read of still more payments. The CEO has 82k in ‘other compensation’, 48k in ‘deferred compensation’, and 16k in ‘nontaxable benefits’ (healthcare?). These may or may not subsume her $63k from ‘other organizations’ previously mentioned, but if we add it all together (), she got $561,468. Nor is she unique—the other executives are all getting their other/
deferred/ nontaxables as well.
the ‘Independent Contractors’ section is more detailed. 2 ‘Technical Services’ firms earned $2.3m; 1 legal firm earned $0.9m; and 2 marketing firms earned $1.2m.
This doesn’t seem so bad, but a final item mentions that there are 45 other independent contractors each paid >$100,000! (If they averaged $250k each, then the 45 were paid $11.25m, for total payments to independent contractors of ~$15.6m.)
One random item I was struck by was $192k spent on direct lobbying (of a legislative body), and a total of $866,500 in direct lobbying 2005-2009. One doesn’t expect the Girl Scouts to have much lobbying to do.
One suggested cost for cookie money is scholarships, but the endowment section ($112m) mentions spending $1.2m on ‘grants or scholarships’, so I doubt any cookie money is going directly there. It’s par for the indebted course that the GSUSA endowment declined to $107m.
The 2010 Form 990 is likewise available from Guidestar.
The first page is as striking as before:
voting members is down 2 to 28
employees is down 21 to 518
- but compensation is still up, from $37.3m to $40.1m
Total revenue: $80.7m. Total expenses: $85.5
- $4.8m in the red is much better than 2009’s $17m
Revenue seems mostly similar (roughly same amount of grants & dues)
- but what on earth is GSUSA doing to earn $85k for ‘software maintenance’?
Expenses is pretty similar
- ‘compensation of current officers, directors, trustees etc.’ has fallen from $2.4m to $1.9m, despite the previously noted increase in employee compensation
- ‘other salaries and wages’ is $23.5m—less than 2009?
- the 2009 report listed just $1.8m for ‘Other’, but the 2010 lists $12m! What on earth?
- Travel fell $0.8m, so perhaps they have been cutting back
The “total net assets or fund balances” by the end of the year totaled $112m, down from $118.9m in 2009; this is more than I would have guessed from a $4m deficit.
2010 lobbying was substantially higher—$307k vs $192k
pg 21 covers the GSUSA endowment with convenient current/
prior year comparisons.
- No prizes for guessing that the balance, contributions, & grants all fell
- however, investment income soared to a gain of $9.5m from a loss of $1.4m
- “other expenditures for facilities and programs” increased by a solid $1.1m
Salary information is very interesting (page 39):
- the CEO compensation for Cloninger now totals $621k, up from $500k—a remarkable $121k salary increase
- next is Corsello (CFO/senior VP), who now makes $500k—a jump of ~$70k
The 2011 return was finally posted in late 2012. To redo the 2010 analysis:
voting members up 1 to 29
employees down to 490, a loss of 28
- but compensation is still up, from $40.1m to $42.98m
Total revenue: $90m. Total expenses: $89.59m
So (barely) in the black! With the 2011 fiscal year, the bleeding seems to have stopped.
The revenue improvements are mostly due to large increases in “Investment income” and “Other revenue”; these seem to be related to “Securities” ($5m; this is the same puzzling “Gross amount from sales of assets other than inventory”) and merchandising sales (“Net income or (loss) from sales of inventory”; but not the “royalties” I speculated were cookie money since that was $8.1m in 2009 but has fallen to $7.9m in 2011).
the “Salaries, other compensation, employee benefits” expense item is $42.98 vs $40.1m, as already mentioned
pg10: “Compensation of current officers, directors, trustees, and key employees” now amounts to $2.1m, “Other salaries and wages” $27.938m; with the remainder being $4.7m for the pension (doubled), $6.2m for “Other employee benefits” (+$0.5m), and $2m in payroll taxes (-$0.2m).
Schedule O (pg59) explains part of the pension change as
NET UNREALIZED GAIN/(LOSS) -8,329,587 PENSION EXPENSE 5,410,836 CHANGE OF DEFERRED GIFTS -142,795
compensation is not much different from 2010. The CEO has another $21k in total compensation, the CFO/VP has jumped $12k, the chief of staff fallen $19k, the “VP & General Manager” held steady, etc.
Overall, a number of existing trends held, but the picture improved. As suggested in previous years, the pension may be an issue. (Perhaps the pension is related to the steady shedding of employees.)
So, what’s the upshot here? I’m far from an expert on these sorts of financial matters—an interested layman, no more. To me, the picture the filing paints is of a somewhat bureaucratic & self-serving organization paying steadily more in salary, benefits, and pension to a gaggle of executives (half a million or more to its CEO) even as its revenues & endowment shrink and it sinks deeper in debt. This picture improves in the 2011 filing, suggesting I may have exaggerated the problems by ascribing the sins of the recession to the management.
But regardless, GSUSA apparently is not doing this on the back of the cookie prices, which seem to be relatively unimportant: the $8m in royalties (if that’s the cookie revenue) is less than 1⁄3 of their membership fees and less than 1⁄2 of their merchandise profit. The price increases outpacing inflation may be due to a desperate or greedy GSUSA, but they seem to be running at a loss only in the last few years, long after much of the price increase, and increases could only net them a little more revenue.
Honesty compels me to say that the culprits are almost certainly the local councils, who reap something like 60% of the revenue and can recapture the other 10% from the troops through the methods described in the previously linked articles.
On the other hand, GSUSA does seem to be in trouble. So perhaps I was right after all? If I was, it is an equivocal sort of right, and one I can’t really claim as a victory. Introspecting, I think I may have fallen prey to a sort of conjunction fallacy which one might call a ‘storytelling bias’—I had a nifty story about a reliable gusher of cookie revenue corrupting a self-selected elite, and I ignored that the elite could become sclerotic and self-serving by many mechanisms, not just the one that captivated my mind at the moment.
- Charity is not about helping -(the Folding@home distributed computing project)
In 2009, the Japanese business analysis and survey firm Tokyo Shoko Research (a combination of Dunn and Bradstreet and TRW in Japan), conducted an examination of the founding dates of the 1,975,620 enterprises in their database.They found 21,666 companies which have existed for over 100 years. The Bank of Korea conducted a similar evaluation of their database and found that there are 3,146 firms founded over 200 years ago in Japan, 837 in Germany, 222 in the Netherlands and 196 in France. There are 7 companies in Japan over 1,000 years old; 89.4% of the companies with over 100 years of history are for profit businesses. However, a closer examination of the history of these long-surviving enterprises reveals that many underwent takeovers, buyouts and essentially a complete restructuring of mission and the nature of the business the firms were engaged in—often more than once in their history. Thus, the chances of a business entity (excluding religious and academic institutions) surviving for >100 years is 1.096%.
But it turns out that cities and companies differ in a very fundamental regard: cities almost never die, while companies are extremely ephemeral. As West notes, Hurricane Katrina couldn’t wipe out New Orleans, and a nuclear bomb did not erase Hiroshima from the map. In contrast, where are Pan Am and Enron today? The modern corporation has an average life span of 40 to 50 years. This raises the obvious question: Why are corporations so fleeting? After buying data on more than 23,000 publicly traded companies, Bettencourt and West discovered that corporate productivity, unlike urban productivity, was entirely sublinear. As the number of employees grows, the amount of profit per employee shrinks. West gets giddy when he shows me the linear regression charts. “Look at this bloody plot,” he says. “It’s ridiculous how well the points line up.” The graph reflects the bleak reality of corporate growth, in which efficiencies of scale are almost always outweighed by the burdens of bureaucracy. “When a company starts out, it’s all about the new idea,” West says. “And then, if the company gets lucky, the idea takes off. Everybody is happy and rich. But then management starts worrying about the bottom line, and so all these people are hired to keep track of the paper clips. This is the beginning of the end.” The danger, West says, is that the inevitable decline in profit per employee makes large companies increasingly vulnerable to market volatility. Since the company now has to support an expensive staff—overhead costs increase with size—even a minor disturbance can lead to significant losses. As West puts it, “Companies are killed by their need to keep on getting bigger.”
I also found estimates which went as high as 1⁄3! Fukuyama cites an estimate of the Church as controlling 1⁄3 of Denmark as well as 1⁄2 of Norway’s land (the source apparently being Lockhart 2007, Denmark, 1513–1660: The Rise and Decline of a Renaissance Monarchy).↩︎
Why is IKEA, normally considered a paragon of virtue with its commitment to renewable energy, such a bizarre legal labyrinth? What moral high ground was forfeited when IKEA decided to help finance Communist regimes by shifting manufacturing to cheaper Poland by 1965? Much of The Omnivore’s Dilemma explores the deep compromises made by American organic farming in order to scale to large sales volumes. The images of IKEA and organic farming remain largely intact & untarnished despite the considerable coverage both have received over the years.↩︎
The real estate bubble aspect makes the tale of Kongō Gumi even sadder; a 1400-year old corporation has seen many bubbles come and go on the local and international scene, and so really ought to know better.↩︎
Since 1969, tax-exempt nonprofit institutions must spend a certain percentage of their income each year. At first the clause required either 6% or the rate of return on assets, whichever was greater. In 1981 the Internal Revenue Service revised the rule and put it at 5%…
This little-known rule is a significant part of American arts policy. On average, most foundations spend 5 to 10% of their assets, with many foundations clustered right at the 5% level. The legal constraint thus appears to bind in many cases. Indeed when the legal requirement changed from 6 to 5% in 1981, the aggregate grant rate fell from 8 to 6% over the following decade. Before 1969 many foundations were little more than methods of tax evasion and storehouses for private wealth. They were explicitly marketed to donors in those terms, and 70% of business-linked foundations had no programs for charitable giving whatsoever. The Ford Foundation, a leading arts donor, had been set up in 1937 but did not actively make grants until the 1950s…Note that from 1981 to 1999, 84.5% of the increase in foundation assets (which rose almost threefold over that period) came from new donations and the creation of new foundations, rather than from returns on existing foundation assets [Mehrling, Perry. 1999. “Spending Policies for Foundations: The Case for Increased Grants Payout.” Economics Department, Barnard College]
$20 would be a small Girl Scouts cookie order for me, and I have bought for many years.↩︎
Indeed, ‘saving lives’ is an easy enough metric that there is a ‘meta-charity’, Givewell, which ranks charities by efficiency; it is sobering to reflect that the price of a life is just $1000 or $2000. If one’s ethics say that failing to stop a murderer is as morally bad as murdering the victim yourself, then that leads to an unpalatable conclusion about every $2000 one spends on luxuries like vacations or restaurants.↩︎
My family is a Scouting family and though I was not a Girl Scout, I went on my fair share of Girl Scout camping trips and did more than my fair share of their arts & crafts. I was left with a higher opinion of the former than the latter, and a sympathy for the girls who would rather join Boy Scouts.↩︎
“Last year, Girl Scouts in the old Lake Erie Council, which served Cuyahoga, Lake and Geauga counties, sold more than 986,000. Each girl peddles an average of 115 to 130 boxes.” quoted in The Plain Dealer, “Some Girl Scout cookies change their names, but the flavor’s the same”.↩︎
“Global cocoa prices: 2010-2011 forecast”, Euromonitor International:
Prices for September-delivered cocoa peaked at £2,465 a tonne on 19 July, the highest level for a second-month contract in 32 years.
tonne is $4.27 per kilogram or $1.94 a (Imperial) pound or 12 cents an ounce.↩︎
Walmart refuses to post prices and most reviews take an almost perverse pleasure in omitting the exact price; so I’m going on this blogger’s offhand mention.↩︎
Figures can vary a lot. If you look at the following quote, it seems to imply that , leaving $1.70. Subtract $1.2 for baking, and they need 50 cents per box to buy advertising, handle the non-existent returns, and offer incentives to the girls?
Cookies cost $3.50 per box.
Of this amount, Bruney said, 40-55 cents, for an average of 47 cents, goes directly to the troop. The council receives another $1.33 per box to subsidize events and camps.
The rest of the proceeds pay the bakers, buy advertising and incentives for the girls, and take care of any debt from people not paying for their cookies, Bruney said.
$50,000 of health benefits is worth more than $50,000 to any self-serving employee, because it is taxed lightly, may keep her salary from being taxed in a higher bracket, and is harder for the public to object to.↩︎