An Analysis of AMS Meeting Costs

Chuck Doswell

As usual, this essay is my opinion only and does not necessarily reflect the opinions of anyone in the office where I work and certainly does not represent the position of the AMS.

A Message from AMS President George Frederick:

I encourage everyone reading these pages to also read the complete analysis provided by AMS Headquarters staff in the following two Web documents:

"Financial Analysis of AMS Meetings" (which provides the financial breakdown for the years 1995-97) and "Financial Analysis of 79th AMS Annual Meeting" (which provides a more detailed breakdown of the 1999 annual meeting, including short courses). Both are available on the AMS Ten-Year Vision pages: http://www.ametsoc.org/AMS/EXEC/TenYear/index.html

I found after reading these documents along with Chuck Doswell's detailed analysis that I reached some very different conclusions than did Chuck. I recommend that everyone judge for themselves only after reviewing all sides of the issue.

1. The data

Ever since Harold Brooks and I began to analyze the AMS budget, it became clear to us that the Meetings and Conferences had an interesting impact on the overall budget. For many of us scientists who have been involved with putting on non-AMS meetings, the AMS registration fees always seemed to be large in comparison to what we were getting. Harold and I guessed that a substantial fraction of the income from each meeting was going to pay for AMS infrastructure costs that might or might not be directly associated with that meeting.

Unfortunately, the budgets for all the meetings were lumped together in the AMS annual budget statements and it was difficult to understand clearly what was going on with regard to where the income from each meeting was going. Recently, the AMS has provided a conference-by-conference budget breakdown ... this is available only to AMS members, but it can be found in Table 1 of the AMS's own analysis found here. Hence, what I am providing are figures directly from the AMS or derived directly from them.

I have downloaded these data and have provided my own analysis of their Table:

It can be seen in Table 1 that the Annual Meeting has a significantly different financial picture than that of any of the specialty conferences, mainly as a result of the income from the exhibitors and corporate sponsors. Hence, it should be recognized that its characteristics are not broadly applicable to the other conferences. Generally speaking, conference registration fees are roughly half of the total income. "Infrastructure" generally constitutes more than half of the total expenses per conference. The registration fees always exceed the "absolutely necessary costs" of putting on the conference, whereas the registration fees generally are less than the infrastructure costs. I have calculated the "absolutely necessary costs" per paid attendee, which can be compared to the actual average registration fees per paid attendee; an interesting calculation is to see what the"absolutely necessary costs" per paid attendee would be if the exhibit income were used to "buy down" the registration costs. The "sales and pages charges" income generally exceed the "printing" costs (which is mostly for the Preprint volume if one is provided). I'll have more to say about all of this later. Now, I want to show the results for 1996 and 1997:

Table 3. 1997 Calculations

1997 Meetings/Locations	77th Annual/Long Beach	So Hem/So. Africa	Hurricanes/Fort Collins	Broadcast/Salt Lake City	Fluid-Middle/Tacoma	BLT/Vancouver	Bio-Hydro/Raleigh	Radar/Austin	App Clim/Reno	TOTAL 1997
Paid Attendees	1,945	198	385	169	255	273	75	329	164	3,793
Pages Published	4,041	430	736	0	408	631	0	642	156	7,044
Papers Published	1,004	0	339	0	120	298	0	295	128	2,184
Exhibits	283	0	0	29	0	0	0	8	0	320
Income
Sales & Page Charges	$268,422	$0	$52,185	$5,308	$29,220	$43,058	$0	$45,296	$14,487	$457,976
Registration Fees	$476,658	$19,038	$76,179	$49,045	$48,845	$54,033	$7,840	$68,635	$36,896	$837,169
Exhibits/Corp Sponsors	$533,869	$0	$0	$21,160	$0	$0	$0	$5,490	$0	$560,519
Total Income	$1,278,949	$19,038	$128,364	$75,513	$78,065	$97,091	$7,840	$119,421	$51,383	$1,855,664
Registration as a % Total Income	37.3	100.0	59.3	64.9	62.6	55.7	100.0	57.5	71.8	45.1
Registration per paid attendee	$245.07	$96.15	$197.87	$290.21	$191.55	$197.92	$104.53	$208.62	$224.98	$220.71
Expenses
Staff	$396,375	$8,435	$54,407	$37,540	$33,092	$41,648	$3,473	$56,333	$20,725	$652,028
Overhead & G&A	$188,598	$3,832	$25,166	$20,232	$15,224	$19,026	$1,516	$25,309	$9,588	$308,491
Printing	$112,732	$0	$20,815	$1,555	$11,922	$17,320	$690	$18,094	$5,170	$188,298
Onsite	$482,275	$8,206	$38,905	$31,819	$20,051	$26,010	$1,017	$40,337	$14,895	$663,515
Audio/Visual	$88,457	$0	$5,250	$1,250	$5,350	$0	$0	$7,413	$4,726	$112,446
Total Expenses	$1,268,437	$20,473	$144,543	$92,396	$85,639	$104,004	$6,696	$147,486	$55,104	$1,924,778
Net	$10,512	($1,435)	($16,179)	($16,883)	($7,574)	($6,913)	$1,144	($28,065)	($3,721)	($69,114)
Staff as a % of Total Expenses	31.2	41.2	37.6	40.6	38.6	40.0	51.9	38.2	37.6	33.9
Overhead & G&A as a % of Total Expenses	14.9	18.7	17.4	21.9	17.8	18.3	22.6	17.2	17.4	16.0
Infrastructureas a % of Total Expenses	46.1	59.9	55.1	62.5	56.4	58.3	74.5	55.4	55.0	49.9
Sales&Page Charges-Printing	$155,690	$0	$31,370	$3,753	$17,298	$25,738	($690)	$27,202	$9,317	$269,678
(Onsite+AV-exhibit) per paid attendee	$18.95	$41.44	$114.69	$70.47	$99.61	$95.27	$13.56	$128.45	$119.64	$56.80
(Onsite+AV) per paid attendee	$293.44	$41.44	$114.69	$195.67	$99.61	$95.27	$13.56	$145.14	$119.64	$204.58
Registration-(Onsite+AV)	($94,074)	$10,832	$32,024	$15,976	$23,444	$28,023	$6,823	$20,885	$17,275	$61,208
Registration-infrastructure	($108,315)	$6,771	($3,394)	($8,727)	$529	($6,641)	$2,851	($13,007)	$6,583	($123,350)

In general, the results from 1995 are validated by those from 1996 and 1997, although variances appear from one conference to another. I am told that this period should be representative of times in the past and the future ... that is, the AMS staff tells me that this 1995-1997 period is not unrepresentative.

2. Inferences

What can one infer from all of this? I am willing to let you (the reader of this page) infer what you wish and you are quite free to manipulate the numbers in ways I've not thought of, perhaps to satisfy your own curiosities about the situation. However, I am not going to let this opportunity go by without saying that I am gratified that our original conjectures about the AMS meetings and conferences have been validated by this exercise: a significant amount of the cost of these conferences, and the reason for the relatively high registration fees is the "infrastructure" component ... that is, we are paying for the privilege of having the AMS staff handle many of the arrangements of our conferences.

Let me hasten to say that I do not believe it is inappropriate for the AMS to charge us (more properly, for many of us, our supporting organizations pay for our participation in conferences) for services rendered. I also agree that it is proper for the AMS to add on an extra fee for "overhead" ... in much the same way that universities charge funding agencies an "overhead" cost. I am told by the AMS staff that the exact costs for AMS staff participation in a conference is generally unknown ... they are not keeping track (in a detailed way) of the hours worked by AMS staffers solely in support of meetings and conferences. I can understand the reasons for this ... it could involve a lot of complexity to keep track of each staffer's time spent in support of meetings and conferences. From this lack of detailed record-keeping, it is my inference that the amounts offered as "charges" for AMS infrastructure at each meeting are more or less rough estimates. When the dust settles from the accounting, it appears that numbers work out such that the conferences as a whole come close to "breaking even." Not every dollar that is taken in during the conference goes specifically to pay for that conference's costs, and those costs only. This is clear from more detailed accounting that I have seen as a result of being a member of the AMS Council. For instance, some part of the cost for the Washington, D.C. office of the AMS has been charged to the income derived from conferences.

At the risk of being repetitious, I am not saying that it is wrong for the AMS to be doing this. If we want the AMS to provide the level of support to which we have become accustomed, we should not expect to receive this for nothing!

What is at issue is the extent to which the individual conference program committees (and their supporting constituencies) wish to continue with this amount of AMS infrastructure support for their particular conference. Some program committees might choose to reduce the costs of participation by opting not to take advantage of any of the services rendered by the AMS staff. As an example, the Cyclone Workshop is organized entirely outside the aegis of the AMS ... since the Cyclone Workshop incurs no costs for "AMS infrastructure" support, it has been possible to keep the registration costs very low. In such situations, the participants in the conference have to do more work and provide on their own whatever services are needed to put on the conference. This approach can be inexpensive but it means more work for the participants. Another option for program committees might be to maintain the status quo with regard to AMS infrastructure support. Clearly, some program committees might want to avail themselves of some fraction of the AMS infrastructure support, but not all of it.

Another point I want to emphasize is that I have been using the AMS-cited expenses for "Onsite" and "AV" costs. It might be possible for program committees to negotiate less expensive arrangements than are typical when the AMS does the negotiation. For example, it has been suggested that universities can offer inexpensive venues for specialty conferences. It has been the position of the AMS staff that (1) universities often are inadequate facilities for AMS conferences, and (2) that they may not be as inexpensive as they seem to be. If some program committee wishes to alter the venue from the traditional choices for AMS conferences (large hotels in major metropolitan areas), they should be free to try this out. They may be able to negotiate lower "Onsite" and "AV" costs on their own. Thus, a program committee might be able to make extensive use of a local arrangements committee to reduce the overall costs of a conference (including hotel registration costs and other factors outside of the accounting presented here) by finding inexpensive alternatives to the "standard" arrangements provided by the AMS staff.

3. Do most conferences "lose money"?

It is the position of the AMS staff, repeated so frequently it has become an accepted "truth," that conferences typically "lose money" ... when pressed, the staff shifts this position slightly, saying that on the whole, averaged over time, the conferences come close to breaking even overall. Given the numbers they (and I) have revealed in Section 1, I'd like to clarify this position. If we consider only the "absolutely required costs" associated with the conferences, then it's quite obviously a myth that the conferences lose money.

This situation was clear to me when I considered those non-AMS conferences I had attended ... when the conferences were over, after having charged substantially lower registration fees than is typical for AMS-sponsored conferences, we always had money left over. The big problem we had was what to do with this "profit" so as not to create a scandal! So why were AMS conferences, charging substantially higher registration fees, such "losers"? The answer now is quantitatively clear, validating the original notions I had when Harold and I began to look into this budget question in the first place. We are paying for a lot of AMS infrastructure with our registration fees, some unknown (but not zero!) fraction of which is not associated directly with the conference at all. The AMS conferences only "lose money" when the cost of the AMS infrastructure is included. This situation is not denied by the AMS staff in my private conversations. The facts are now available for all to see, and the facts were provided by the AMS staff, so there can be no question of their validity. Perhaps the AMS staff will not like all of my interpretations, but the facts cannot be at issue.

Sometimes the issue of the Preprints comes up when asking about the costs of AMS conferences ... from the data provided in Section 1, it's clear that the Preprints typically make a profit (there was only one modest exception to this in the three years for which data were provided by the AMS), which can be substantial. Hence, this is primarily a non-issue regarding the conferences; if anything, this is yet another money-maker for the AMS ... it certainly provided a "profit" over the course of each of the three years for which data were provided by the AMS ( $353,192 in 1995, $330,758 in 1996, and $269,678 in 1997 for a three-year total of $953,628!). Preprints are typically derived from camera-ready manuscripts provided by the authors, so it is almost certainly possible to reduce the costs of printing the Preprints considerably by exploring other printing options than those provided by the AMS. This might reduce the page charges correspondingly ... in some non-AMS-sponsored conferences I've attended, the Preprints were included in the Registration Fees but no page charges were required! Admittedly, the quality of the cover and bindings might have been less than those of the typical AMS Preprint volume, but the participants might well be willing to accept this modest decrease in quality in exchange for saving themselves (and their supporting organizations) hundreds of dollars in page charges for each paper.

So I believe the answer to the question of this section must be: No, AMS conferences "lose money" only in a substantially-qualified way. I have indicated repeatedly that the AMS has a right to recoup the expenses incurred during their activities in support of conferences, and I even concur with their right to charge an "overhead" cost in association with that support, above and beyond the costs associated with the conference itself. My main goal here has been to clarify the magnitude of this cost to the members and to make sure that the members understand what they are getting and why their AMS-sponsored conferences "lose money," even as their non-AMS-sponsored conferences "make money" in spite of charging the participants substantially lower registration fees. I now have quantitative confirmation of what Harold and I guessed had to be the case right from the beginning. It's now up to the members to ponder how they might wish to react to the information provided in Section 1.

4. AMS Short Courses

I want to consider the separate accounting as given by the AMS for the Short Courses at the 79th Annual Meeting (in Dallas - 1998):

I've highlighted the specific items relevant to the totals for the six short courses in green and my additional calculations are shown in blue (as with the previous tables). From the accounting provided above, it seems that the short courses have netted a modest surplus, which probably has been dispersed to offset other expenses. One short course's luncheon was sponsored in 1999, which is where the $1500 comes from according to the explanations provided by the AMS. It is my hope that this separate accounting for the short courses will be continued.

5. Thoughts, conclusions, and recommendations

In my informal discussions with the AMS staff, I find them generally to be projecting a spirit of seeking to comply with the wishes and needs of the membership. I believe that the AMS staff wants to convey more widely within the membership that spirit of responsiveness to the wishes of the organization's participants. Therefore, it's my belief that if program committees want to depart from the traditional relationship between themselves and the AMS staff, they should expect to find, not resistance but rather a willingness to try new things. If that responsive spirit is not forthcoming, in the experience of the program committees (and their associated STAC committees, when applicable), then the members should feel responsible for communicating their reactions to the AMS staff ... for example via the AMS Web Page, or by whatever means is possible. Silence will be interpreted as acquiescence and apathy!

One barrier to an effective transition to a more flexible relationship between program committees and the AMS staff is the lack of quantitative knowledge of the costs of various activities of the AMS staff (including the "overhead rate"). The information provided by the AMS staff (and me) in Section 1 (above) gives some crude guidelines about the expected costs associated with the current suite of AMS staff-provided services. Thus, the choice at the moment seems to be either black or white; the "infrastructure" costs are either (a) the full suite, or (b) none. Should a program committee wish to consider a more flexible range of options, involving a lesser level of AMS staff support, but not zero, it is not possible (to my knowledge, anyway) to know what those costs might be. If a program committee is going to set its registration fees appropriately for those situations involving some intermediate level of support from the AMS, they need a "price list" that would tell them precisely what specific services are being offered and what each service costs. This would permit the degree of flexibility that I believe many program committees would want, and that flexibility is what I believe the AMS staff is interested in providing.

This "price list" also would be helpful in allowing the program committees to have full understanding of, and control over, the budget for each conference. Thus, I believe that the program committee should be intimately involved with their conference budget, from early in the initial planning, right down to the final accounting. It seems to me that this level of "local" control by the program committees would shift the burden of responsibility from the AMS staff ... if the constituents associated with a particular conference expressed unhappiness over the choices of the program committee, that unhappiness should be directed first at the members of the program committee. I think the obvious response by the program committee should be to offer any unhappy constituents an opportunity to make the decisions the way they want at the next conference .. i.e., to serve on the next program committee. I am generally in favor of clarifying the relationship between the activities of the Society and the membership. If the members are unhappy with the way things are going, it should be clear that (a) they can do something about it, and (b) if they aren't willing to take responsibility to change things, then there is no justification for their griping. This shift of responsibility should act to encourage a greater level of participation by the membership in the preparation for meetings and conferences.

It's my perception that the AMS staff has moved significantly toward developing a responsiveness regarding meetings and conferences that has not always been present in the past. The membership needs to understand that the budget of meetings and conferences can be much more in their hands, if they wish to take on that responsibility. If everyone is happy with the status quo, then the AMS "infrastructure" costs are pretty clearly defined in what they (and I) have presented in Section 1. I have tried to make clear what those costs are and their justification under the status quo. What you, the membership, wish to do with this information is up to you.

If the new flexibility regarding meetings and conferences becomes a reality, it seems obvious to me that the participants in the experiments (made possible within the boundaries of this new flexibility) need to share their perceptions about the outcome. It's still my belief that many program committees fall back on the traditional approaches of the status quo because it's what they know. Most program committee members are serving in this capacity for the first time and have little experience upon which to draw in departing from the traditional conference structures. When new concepts regarding meetings and conferences are tried, there needs to be some sort of "institutional memory" of the outcomes of the experiments done at the conferences and meetings. An example of how this could be done can be found here. This should be made available to incoming program committee members as part of the information they can use to take advantage of what I hope will be a substantially widened range of options in putting a conference together.

With any luck at all, this could increase the degree of member satisfaction with their conferences ... because, after all, the conferences belong to the membership, not to the AMS! Or at least that is the way I believe we should all be thinking about this. That is, the AMS staff is paid to help the members get what they want from their scientific/professional society ... if that is truly the case, not only does it say the AMS staff should be responsive to the wishes of the members; it also says the members should take responsibility for the activities of the AMS, as a whole!