New report finds lack of association between NYC arts orgs and disadvantaged neighborhoods

[Carl Grodach]

Hello all,

 

I'd like to announce a new report conducted by myself and Nicole Foster and James Murdoch at University of Texas Arlington-- ART SPACES, ART PLACES: Examining Neighborhood Preferences of New York Arts Organizations. 

The study, which was funded by a grant from the New York Community Trust, uses data from the New York Cultural Data Project database along with industry, demographic, and built environment measures to study the location patterns and characteristics of arts organizations in New York. The results of our analysis produce three key findings:

1. In contrast to literature investigating the role of the arts as community anchors, we find a negative association between arts organizations and diverse, ethnic neighborhoods as well as poorer disadvantaged neighborhoods. This is especially the case for younger, smaller, locally focused organizations in New York City. This is a surprising and troubling finding in light of research suggesting that the arts can catalyze positive social and economic outcomes in struggling urban communities.

2. There is a strong relationship between the presence of arts organizations and the creative economy and neighborhood amenities, confirming past research examining the link between the arts and economic development.

3. There is strong relationship between the presence of arts organizations and urbanized neighborhoods that are home to young adults and singles. This substantiates the widely held assumption that the arts tend to locate in the densely populated urban core and that arts organizations are closely linked with young professionals in the ‘creative class’. 

If you are interested in the full report please contact me at carl.g...@gmail.com.

Carl Grodach

Property and Planning Discipline

Queensland University of Technology

[Andrew Taylor]

Following is a response from Ann Markusen that she asked me to post (some lingering listserver issues make it impossible for her to post).

…….

Carl and all,

This will be a great addition to the literature. However, I am concerned
about the NY CDP data. If it is anything like the California data that we
used, it woefully undercounts smaller (by revenues, though if you monetized
the volunteer labor involved, they loom much larger) and ethnic, immigrant
and multidisiciplinary organizations. We dealt with this by weighting by
the IRS data, including for very small arts organizations. See the
technical appendix to our Irvine study:

Markusen, Ann, Anne Gadwa, Elisa Barbour, and Bill Beyers. 2011.  *California's
Arts and Cultural Ecology. * San Francisco, CA: The James Irvine
Foundation, September.
*http://irvine.org/news-insights/publications/arts/arts-ecology-reports
<http://irvine.org/news-insights/publications/arts/arts-ecology-reports> *

We may have traded emails about this before, and maybe you told me the
coverage of the CDP in NY is good, but I'm still worried about drawing
implications from data that are not a sample nor a census.

Look forward to reading the whole!

When are you off to the down-under?  Wishing you the best!

Thanks Ann

mark...@umn.edu

…….

..............................................................

Andrew Taylor <ata...@artsaxis.com>

608.347.1048

--
You received this message because you are subscribed to the Google Groups "Cultural Research Network" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cultural-resea...@googlegroups.com.
To post to this group, send email to cultural...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Research Center for Arts and Culture]

Anne et al,
To add to this discussion, I believe that it is still the case that CDP includes nfps and organizations INTENDING to be nfp, which makes conclusions difficult.  Do correct me if I am misinformed.  Best, Joan Jeffri

[Carlos Manjarrez]

I believe that folks from CDP are on the CRN list.  It would be helpful is someone from that organization could respond to Joan's query (using the list) and if you could also say something about the representativeness of CDP data in New York relative to other publicly available lists or arts organizations, such as the IRS 990 and 990-N filers for past X years.

-Carlos

Sent from my iPhone

[Christopher Caltagirone]

The CDP’s mission is to empower the nonprofit arts and cultural sector with high-quality data and resources in order to strengthen its vitality, performance and public impact. We applaud every effort to collect and analyze useful data on the arts and cultural sector. It is important, however, to recognize some fundamentals about the intention and scope of our work:

First, nonprofit organizations usually submit CDP profiles to meet the prerequisites of the philanthropic organizations from which they are seeking financial support. Funders use the profiles as one of their criteria as they evaluate the submitting organizations. Those two needs – the need for support and for evaluative data -- are the core drivers for participation by more than 10,000 organizations and nearly 300 grantmakers. For ten years we have had an excellent track record in meeting those needs. We constantly strive to make our Data Profile an even more useful tool for those purposes. However the CDP data is not, and has never presented itself as, an all-inclusive “survey” or census of organization-level arts and cultural activity.

Second, once we collect data, we make it available to researchers. They pursue independent studies and may choose to base conclusions partly on their analysis of our data. The data we provide is accurate and specific and can be delineated by ranges of revenue, nonprofit status (current, pending, or part of a larger organization) and many other factors. The CDP collects and categorize data – not “estimates” of arts activity. Therefore we do not “undercount” or “underestimate” any trends – we document the neutral data from our profiles. We do not endorse the agenda or conclusions drawn by anyone who uses our data as part of their broader research.

Finally, although we continue to improve our service to the organizations we focus on in the nonprofit arts, cultural, and philanthropic communities, we do not intend to be or profess to be all things to the data-seeking world. We encourage the research community to identify and explore new areas of investigation, especially those that seek to verify the ability of arts and cultural organizations to improve communities and the economy, but it is important that our clearly stated charter is not misconstrued or viewed as a shortcoming simply because it does not provide all of the answers being sought.

We stand by the accuracy of our work and the value we deliver.

[Carlos Manjarrez]

Chris,

Thanks for responding to the list.  I think we can all agree that it is important to have a thorough understanding of the origins of the data we collect and use.  Indeed, it is essential if we are going make serious, empirically-grounded statements about the sector.  

However, knowing what an organization intended for a given collection is not enough.  For users (and consumers) to have confidence in the data they have to ask the kinds of questions that have been asked here on the CRN list.  Most of us share CDPs interest in leveraging the data for full impact.  But we want to do so in a responsible manner, which requires detailed information about limits of the data. Please realize that this is not an indictment of CDP.  This is all about the data.

So, having said that, I'd like to ask again if CDP has recently done an analysis of how representative it's data is (in terms of discipline and organizational size) when compared to other publicly available data holdings on the arts and culture sector?  If so, can you share that information with the list?  If not, do you know of other researchers that have looked into this issue for any of your partner states (apart from the California study by Ann and Anne).  

Many thanks.

-Carlos 

C. Arturo Manjarrez

Director, Office of Planning, Research and Evaluation

IMLS

--

[Research Center for Arts and Culture]

Yes, Chris, many thanks for the info. We are all grateful for CDP, a mammoth undertaking, and just trying to understand the particulars better, Joan

--
Joan Jeffri, Director
Research Center for Arts & Culture
National Center for Creative Aging
4125 Albemarle Street
Washington, D.C. 20016
202-895-9456
www.creativeaging.org/rcac
rc...@creativeaging.org

[Christopher Caltagirone]

Carlos,

Thank you for your reply.  We are going to compile some information for the list and get some additional thoughts and data posted either today or tomorrow.  Obviously we are very interested in this issue as well and look forward to further engagement on this topic.

Best,

Christopher

[Christopher Caltagirone]

Carlos,

Thanks again for the opportunity to respond.  I apologize that compiling this data took longer than expected.  Like many in the field, we are very interested in the question of representation and have found it to be an intriguing and difficult issue to address.  It is my hope that the numbers provided below will generate a wider discussion about the challenges, limitations and opportunities encountered by those engaging with these datasets.  Please note that the NCCS numbers are taken from a January 2013 dataset. 

NTEE CODE	NCCS-ALL NY	NCCS-ALL NY w/TAXPER	CDP	% to NCCS w/TAXPER
A01-A19	755	727	105	14.4%
A20-A27	1,415	1,312	375	28.6%
A30-A34	638	613	136	22.2%
A40	260	235	56	23.8%
A50-A57	548	528	147	27.8%
A60-A6E	3,084	2,999	991	33.0%
A70	250	236	35	14.8%
A80-A84	1,375	1,322	104	7.9%
A90-A99	187	180	59	32.8%
C41	44	35	6	17.1%
D50	15	15	3	20.0%
N52	80	73	5	6.8%
TOTAL	8,651	8,275	2,022	24.4%

 

There are a several issues regarding the NCCS numbers that require our attention.  One is the number of defunct organizations represented in the dataset.  There are nearly 400 organizations in the codes listed above that did not have a tax filing date (TAXPER). Removing those organizations drops the total for the selected codes to 8,275. There were also more than 200 organizations whose tax filing date is 2006 or earlier, although I left those in the counts above.

Coding errors are another issue requiring further investigation.  A23 – Cultural and Ethnic Awareness is of particular concern, as a cursory review of the more than 650 organizations (approximately half of all the organizations in the A20-A27 group) with this code reveals a number of questionable inclusions.  For example, this particular dataset included organizations such as Brightstar German Shepherd Rescue Inc., Dominican Medical Association Inc., Flushing Chinese Business Association Inc., Greater New York Korean Nurses Association Inc., and the National Arab American Medical Association Inc. Some studies have identified coding errors of up to 20% within certain categories, and there is sufficient reason to believe the counts of organizations within the Arts, Culture and Humanities category could be inflated if these errors are not addressed.

There are other issues to investigate but for now I will leave this information here for others to review.  I look forward to continuing the discussion.

On Tuesday, June 17, 2014 9:36:21 PM UTC-4, Carlos Manjarrez wrote:

We stand by the accuracy of our work and the value we deliver. <w:LsdException Locked="false" Priority="69" SemiHidden="false" UnhideWhenUsed="false" Name="Mediu

...

[Kiley Arroyo]

Dear Christopher and Carlos,

Thanks again for this important discussion and subsequent breakdown. I do not see Chris's most recent post on the Google group, do either of you? Either way, I have a few follow up questions for you.

Chris: Do the numbers in columns 1 and 2 (NCCS-ALL NY and NCCS-ALL NY w/Taxpayer) reflect only registered 501c3's with annual budgets above $50k? If so, do you have any idea how many might exist below this threshold? It seems to me we need both to accurately compare and comprehend how representative the CDP NY is, right?

Like you and several of our colleagues, I have analyzed how representative both the CDP and NCCS data is with regards to 501c3 arts and culture organizations. Last spring I analyzed NCCS data from 4 counties in SE MI, covering the years 1990 and 2000-2013. I found it necessary to process the data in two groups: Sub-set 1 (organizations with annual budgets of more than $50k) and Sub-set 2 (organizations with annual budgets of less than $50k). I have attached a document which outlines the data processing methodology used to create each group and the subsequent sample sets by year. As you'll see, the latter process was completed to help insure records included within each year, were in fact for 'live' organizations and the criteria for establishing such. This approach was deemed defensible by partners of mine at NCCS.

Despite the rigor of our method, we still lacked comprehensive data for nonprofit organizations with annual revenues below $50k (since they fall below the federal filing threshold). Any information we were able to obtain and manipulate, was drawn from a combination of the IRS BMF and Core file. As you probably know, the IRS often will not keypunch organizations that fall below this threshold (gross receipts <$50k).  However, NCCS has found that this is not a perfect rule – occasionally organizations with less than $50k in gross receipts will slip through and get keypunched.  That said, it is impossible to tell whether an organization that has 0 for income fields actually has 0’s across the board, or is merely under $50k and wasn’t keypunched.  Our partners at NCCS advised that after the creation of the 990-N postcard system, a significant majority of the organizations for which “filed_with_zero_assets_and_income_last_two_years” is “TRUE” are 990-N postcard filers, and aren’t reporting any real financial information and recommended they be analyzed separately from your other data, which is why data was processing as Sub-set 1 and 2. When creating a brief on the impact of the recession on nonprofit organizations NCCS looked only at organizations for which that final field “filed_with_zero_assets_and_income_last_two_years” was “FALSE” AND the organization reported over $50k in revenue AND “filed_within_last_two_years” was “TRUE”. We applied a similar logic.

As a result of processing NCCS data this way, we found that there were actually more records in Sub-set 2 than 1, meaning there were actually more tiny nonprofit arts organizations operating in SE MI than those with annual budgets at or above the $50k filing threshold. We would not have known this if we simply analyzed the primary NCCS data set. I appreciate the collective concern expressed about how representative the CDP data is, but would add that we should be equally cognizant of whether / how representative the NCCS data is particularly if it is to serve as this kind of baseline.

Thoughts and critiques warmly welcome.

Kiley

[Christopher Caltagirone]

Kiley, thank you for your questions.  The first column (NCCS - ALL NY) reflects the total number of organizations for the reported categories in the NCCS/IRS Business Master File (as of January 2013).  It is my understanding, based on information provided on the NCCS website, that the BMF contains "descriptive information for all active organizations that have registered for tax-exempt status with the IRS." (http://nccs.urban.org/database/overview.cfm)  Therefore this column would include any and all 501(c)(3) organizations regardless of budget size. 

The second column, NCCS-ALL NY w/TAXPER, removes only those organizations that have no end date in the TAXPER (Ending date for tax period) field. I considered excluding organizations with filing dates prior to a certain year but decided for the time being to leave in any organization that had any ending date recorded in the file. I am certain that a number of inactive organizations are included in the counts recorded in the second column, but again, I was deliberately casting a wide net with the expectation that additional refining of the parameters would likely narrow the universe further.

I originally considered removing organizations with no values for income or assets but decided against this approach so as not to exclude small organizations that might not have this information recorded in the BMF but are still considered active.  With respect to budget size, I performed a quick analysis on the CDP organizations and their most recently completed Data Profile and found that 20% reported less than $50k in revenue. I would be very curious to know how the NCCS NY counts would change were you to apply your methodology to this data.

If I have read your documentation correctly, you found 330 organizations with revenue under $50k, and 308 with revenue over $50k.  Is this correct?  Of those 330, how many had any financial data recorded in either the BMF or Core file?  Is one of the primary issues the lack of financial data available from either the BMF or Core files for those organizations with revenue under $50k?

One other thought I would like to briefly address here.  We also haven't addressed the subject of representation by geography.  We recognize the preponderance of CDP data in urban areas, although we do have suburban and rural organizations participating in CDP.  There is robust representation in many geographies where CDP is established and participation is strong.

[Carlos Manjarrez]

Hi all, 

 

I'd like to chime-in on this issue of "representativeness" and would invite others to do so as well.   While I think the comparison that Chris provided is useful, I fear that the CRN thread around the issue of CDP data has sent us down the wrong path.  I’ve seen this happen in other discussion as well.  For some reason there is an assumption that CDP data should be representative of ALL arts and culture organizations in the states where the initiative is operating.  But I think that is the wrong way to look at this data resource.

 

I would argue that CDP was never designed to be representative of all arts and culture entities.  It was designed as an administrative data collection that specifically targeted formally organized arts and culture organizations that sought grants from public and private grantmakers.  In other words, selection bias (and I mean this in the value neutral, social science definition of the term) is baked right into the data collection process.  Again, this is not an indictment of CDP or the data they collect.  It’s a basic description of what’s driving that dataset. 

 

What does this mean?  Well, for one it means that the question should not be how representative CDP data is of all arts organizations.  The more pertinent question is how representative is CDP data of formally organized grant seeking arts and culture organizations.  That, as most of you know, is a very different lens.  The distribution of formally organized arts and culture organizations is not a clean bell-shaped curve.  It is skewed with many, many small entities that bottom end of the distribution, that, for a variety of reasons may not be seek grant support from state-level public agencies or large foundations.  The get a good picture of ALL formally organized arts and culture organizations we are going to need to look at a variety of different data collection techniques.  We can’t hold out hope for one organization or one data collection method to capture and represent such a varied and dynamic set of organizations.  It’s just not going to happen.  That’s not the data world we live in any more.  Data come in many different forms from many different sources. We need to map our data collection efforts to the realities of the world we live in.  And that means drawing in the best data we can from wherever we can get it.

 

This is something that IMLS recently did this for museums with our partner Neville Vakharia at Drexel’s Westphal College of Media Arts and Design.  We pulled together data from IRS, Foundation Center, IMLS administrative data and data aggregation service providers like Factual to come up with a new estimate of operating museums in the US. It’s a first pass and I we have been praised and criticized for putting the file out there but you’ve got to put a stake in the ground somewhere if you are going to build a data resource. 

 

I think that the same thing needs to happen for arts and culture organizations generally.  Right now I know of no researchers or research centers out there that are collecting data on an ongoing basis to monitor the size and scope of the arts and culture sector.  But I will readily admit that I’m not the best authority on this issue (and actively looking for these numbers).  Do others know of an ongoing data collection effort of this sort?  Was this something that Sustain Arts was trying to do?  Are they doing it?  Zannie, is this something that you've mapped out for the National Center for Arts Research?

 

-Carlos

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 

C. Arturo Manjarrez

Director, Office of Planning, Research and Evaluation

Institute of Museum and Library Services

9^th Floor

1800 M Street NW

Washington, DC 20036

 

cmanj...@imls.gov

[Research Center for Arts and Culture]

Hi All, I have enjoyed this discussion. Perhaps it would help us all to think about an auspice for all this collective research which clearly would need to explain methodologies, sampling and collection techniques, etc. but would enable us all to go to a central source. With  CPANDA now at NEA, the obvious question is, "Could NEA expand to serve this function, not only for US researchers, but for international ones as well?"(Sorry to put you on the spot, Sunil.)
Best, Joan Jeffri
,

--

[Ian David Moss]

Hi Carlos et al,

Greetings from Athens, Greece. Great analysis of appropriate researcher expectations for the CDP, and I Iove the idea of a “known universe” data file for museums. Kudos!

 

Regarding your question at the end, the upcoming Sustain Arts resources in Southeast Michigan and the SF Bay Area will indeed provide a “known universe” count of arts and culture nonprofits in those regions. The data sources previously mentioned (principally NCCS/Guidestar/IRS) are triangulated with other information from the organizations’ websites, grant histories, and Google search results to get a better sense of what each individual organization actually does and whether it belongs within SA’s discipline-based definition of arts and culture. The capability to extend this method and database nationally exists, but it is a resource-intensive process to get to the level of data quality being sought, and even with the additional due diligence there is still no promise of 100% accuracy. If having this kind of canonical census of arts nonprofits on a national scale is something our field decides is of value, it’s going to need to be willing to invest in making it happen – there is no cheap/quick fix that I’m aware of.

 

Happy to answer further questions about this, but as I’m on vacation for the next week…it might not be for a little while. J

--

[Zannie Voss]

Hello Colleagues,

 

Quite a lively discussion!

 

Carlos, in response to your question related to NCAR and "researchers or research centers out there that are collecting data on an ongoing basis to monitor the size and scope of the arts and culture sector," NCAR isn't in the primary data collection business. We aggregate and integrate data from as many good data sources as we can identify in order to examine organizational health and the factors that influence it. That task is currently taking all of our energy and resources. Our primary organization-level data sources are NCCS, CDP, and Theatre Communications Group.

 

We wholly support the partners that collect and provide us data, such as CDP, and strongly encourage any and every arts and cultural organization to submit its data to CDP. We follow a similar process as Sustain Arts in terms of verifying certain organizational factors when necessary. We have data on roughly 50,000 unique organizations currently; for many we have several years of data.

 

In TCG's Theatre Facts, we annually report estimates of some basic financial, attendance, and employment data on nonprofit professional theatres in the report's Universe Theatres section. It's only one arts sector, though and, like all such efforts, no doubt leaves out some very small oragnizations. The 2013 report will be coming out later this summer.

 

I wholeheartedly concur with Ian's comments, both about data accuracy and about the gargantuan task of attempting to create a census. Hope you have a fabulous time in Greece, Ian!

 

Enjoy the weekend,

Zannie

[Sunil Iyengar]

Joan: I'll bite! I too have enjoyed following this thread, noting in particular Carlos' desire to see how different and (differently purposed) data collections about arts organizations map to each other.

 

For any single research team/s or center/s to begin to "monitor the size and scope of the arts and cultural sector," the first hurdle to clear is consensus on variables and definitions--and elucidation of the gaps that permeate extant datasets whose value to this enterprise is clear, but whose contents sprang from other motivations, other constructs.

 

We at the NEA could do more to merge administrative and commercial data with government datasets to obtain a clearer view of arts organizations in general.

 

Consistent with our federal vantage, however, and in keeping with our agency's mission and goals, we've made more of a priority of assembling national, publicly accessible data collections (often in partnership with larger federal entities) that can help researchers to investigate the value and impact of the arts.

 

This means that increasingly the data collections we sponsor are purposed for answering specific research questions. Arts organizations are key actors in our system of inquiry, but like many people on the Cultural Research listserv, we've been obliged for this topic to use extant sources already named.

 

Still, I'd like to point to two data sources that might not have been considered by everyone: for all their limitations, the U.S. Economic Census and the new Arts & Cultural Production Satellite Account (ACPSA) can be profitably examined with respect to certain questions about arts and cultural organizations, for-profit and non-profit alike.

 

Also, please be aware that the NEA's Research: Art Works program represents an opportunity to establish new data collections that might help address questions raised on this listserv. Application guidelines will be out this summer. As for obtaining administrative data about NEA grantee organizations, we encourage researchers to do so if useful. (See http://arts.gov/freedom-information-act-guide.) We also have established an online search function for the NEA grants database, to help with certain kinds of requests: http://apps.nea.gov/GrantSearch/.

 

Finally, it's correct that through an agreement with Princeton University, the NEA has acquired the CPANDA datasets. We're currently working to transfer them to a special collection that will be available at the Interuniversity Consortium for Political & Social Research (ICPSR), free of charge. In the interim, those datasets still may be accessed via cpanda.org.

 

In tandem with this service, the NEA is producing online Arts Data Profile pages (at the rate of roughly one per quarter) to illumine arts-related datasets through summary statistics, macro-data, and basic visualizations.

 

Thanks for allowing me to be part of this dialogue!

 

Kind regards,

 

Sunil

 

Sunil Iyengar

Director, Research & Analysis

National Endowment for the Arts

 

 

From: cultural...@googlegroups.com [mailto:cultural...@googlegroups.com] On Behalf Of Research Center for Arts and Culture
Sent: Thursday, June 26, 2014 2:51 PM
To: cultural...@googlegroups.com
Subject: Re: [cultural-research] Re: New report finds lack of association between NYC arts orgs and disadvantaged neighborhoods

 

Hi All, I have enjoyed this discussion. Perhaps it would help us all to think about an auspice for all this collective research which clearly would need to explain methodologies, sampling and collection techniques, etc. but would enable us all to go to a central source. With  CPANDA now at NEA, the obvious question is, "Could NEA expand to serve this function, not only for US researchers, but for international ones as well?"(Sorry to put you on the spot, Sunil.)
Best, Joan Jeffri
,

[Andrew Taylor]

I’m forwarding the message below on behalf of Ann Markusen:

……..

Dear CRN colleagues (and thanks to Andrew for posting),

I’d like to weigh in on why it is important that data sources like the CDP ensure that users of its data for arts advocacy purposes understand the bias in the data set and acknowledge it in their publications. In an early use of CDP data, the Philadelphia Cultural Alliance produced detailed graphics of the composition of the arts sector in the metro that show a roughly “normal” distribution of arts organizations with more medium-sized than either very large or small, and their decomposition by arts discipline was similarly flawed. Does anyone know of other uses of the CDP for such purposes? I believe I saw a report from an arts advocacy organization in the southeastern states (or a metro therein) whose use of the CDP data was similarly problematic, but I can’t find it at the moment.

Here are two published papers/articles where my co-authors and I review this mis-use of CDP data in the Philadelphia report and offer some reflections on why it happens and why it matters:

From Noah Isserman and Ann Markusen. 2013. “Shaping the Future through Narrative: the Third Sector, Arts and Culture.” International Regional Science Review, Vol. 36, No. 1: 115-36.  http://www.annmarkusen.com

After demonstrating the severe undercounting of smaller arts organizations and particularly those in ethnic, immigrant and multi-disciplinary categories, we review a case of advocacy misuse of the CDP Data:

“These concerns with the bias and use of CDP data are not mere researcher nitpicking. In the first major use of CDP data for arts advocacy purposes, the Greater Philadelphia Cultural Alliance study (2008) showcased CDP data that are similarly selective. Despite a data note acknowledging an undercount of smaller organizations, the study’s graphical depiction of area nonprofit arts organizations by size and discipline (GPCA, 2008: Chart 2, p. 10) depicts medium and large-sized (not counting very large) organizations as more numerous than smaller ones, highly unlikely given bias in the CDP survey respondents. We estimate that California organizations under $250,000 account for 85% of all arts and cultural nonprofits, compared to 39% in the Greater Philadelphia study. The same chart displays only 30 community arts and cultural organizations, 30 museums, galleries and visual arts organizations, and 77 dance, theatre and other performing arts—again, highly unlikely given our benchmarked results. Yet the 2008 Portfolio claims to provide “an accurate measure of the current health and vibrancy of the cultural environment” (p. 8.) Thus the CDP data, its collection and use are, to date, reinforcing the notion that larger, urban arts and cultural organizations are dominant and by inference, superior, providers of arts and cultural experiences.”

Citation: Greater Philadelphia Cultural Alliance. 2008. 2008 Portfolio. Philadelphia, PA: Greater Philadelphia Cultural Alliance. http://www.philaculture.org/research/reports/2008-portfolio

From Amy Kitchener and Markusen, Ann. 2012. “Working with Small Arts Organizations: How and Why it Matters.” Grantmakers in the Arts Reader, 23:2: 5-12. http://www.giarts.org/article/working-small-arts-organizations, http://www.annmarkusen.com

“Similarly, California’s ethnic, folk arts and multi-disciplinary organizations (of all sizes) are greatly underrepresented in the CDP data, where coverage is best in performing arts (22%) and music organizations (19%) (Figure 3). The CDP captured only 4% of California arts nonprofits specializing in humanities, legacy, and non-visual arts museums. The CDP also over-estimated the share of California arts nonprofits in the large urban agglomerations like the Bay Area and Los Angeles while under-estimating those in the valleys and mountains north and east of Sacramento, the San Joaquin Valley, and Inland Empire. Since organizations don’t take the CDP survey if they are not applying to funders that require it, the under-count is caused both by underfunding of small nonprofits and by fewer funders working outside of major metro areas. 

Under-counting often reinforces under-appreciation. In the first major use of CDP data for arts advocacy purposes, the Greater Philadelphia Cultural Alliance  (2008) used CDP data to depict the area’s nonprofit arts. Despite a note acknowledging an undercount of smaller organizations, the study’s graphs of nonprofit arts organizations by size and discipline (GPCA, 2008: Chart 2, p. 10) depicts medium and large-sized (not counting very large) organizations as more numerous than smaller ones. We found that California organizations under $250,000 comprise 85% of all arts and cultural nonprofits, compared to 39% in the Greater Philadelphia study. Yet the 2008 Portfolio claims to provide “an accurate measure of the current health and vibrancy of the cultural environment” (p. 8.) 

Undercounting and under-appreciation reflect and perpetuate under-funding. For the US as a whole, Sidford (2011) shows, using NCCS data, that philanthropic funding heavily favors large arts organizations over small and ethnically-specific and social justice groups. Those with budgets greater than $5 million receive 55% of all contributions, gifts, and grants, even though they account for only 2% of total arts and cultural nonprofits (p. 8).

 

Ethnic, non-Euro-American, and low-income community-serving arts organizations are markedly underfunded. Sidford estimates that three-quarters of US arts nonprofits have budgets under $250,000 (excluding those under $25,000). Using 1stAct Silicon Valley data on 659 active arts, culture and humanities organizations in 2008, she finds that 70% of the region’s groups are less than 20 years old, and 30% of these are ethnic-specific, focusing on the cultural traditions of India, Mexico, Japan, Korea, China, the Philippines, among others (p. 13). In the largely rural San Joaquin Valley, where ACTA often works, most arts organizations are small and play a crucial role. While our illustrations in what follows are drawn from ethnic, folk arts, and multi-disciplinary organizations, our characterizations apply to the entire range and diversity of small arts nonprofits.”

Citation: Sidford, Holly. 2011. Fusing arts, culture and social change: High impact strategies for philanthropy. Washington DC: National Center for Responsible Philanthropy, October. http://www.ncrp.org/paib/arts-culture-philanthropy

See also my comments on the fascinating Barry’s Blog about the brewing battle in San Francisco over funding for small versus large arts organizations: http://blog.westaf.org/2014/06/a-potential-deep-divide-in-arts-sector.html#comment-form

Really appreciate the debate on this, and want to underscore, as I’ve written before, that with proper benchmarking, the CDP offers us some wonderful windows into dimensions that we don’t otherwise have, such as the balance between contract and employment status of people paid for by arts nonprofits by occupational group (insight in California: 57% of all those paid for work are artists, the vast majority of whom are working on contract and part-time…). See our California Arts and Cultural Ecology study and especially our technical appendix that explains how we benchmarked using the NCCS:

Ann Markusen, Anne Gadwa, Elisa Barbour, and Bill Beyers. 2011.  California’s Arts and Cultural Ecology.  San Francisco, CA: The James Irvine Foundation, September. http://irvine.org/news-insights/publications/arts/arts-ecology-reports

Regards,

Ann Markusen

mark...@umn.edu

annmarkusen.com

……..

..............................................................

Andrew Taylor <ata...@artsaxis.com>

[Nicholas Crosson]

The Greater Philadelphia Cultural Alliance objects to the misrepresentation of our research on arts & culture. Our “Portfolio” reports do not purport to be a census. They present the best understanding of organizational health in the Philadelphia region, and an excellent source for data to do this is CDP. In our 2011 Portfolio, our findings illuminated in detail the difficulties arts organizations faced in the Great Recession; the detailed financials in CDP allow a very thorough examination of revenue sources, increases in expenses by type, organizational liquidity, etc.

 

We do a careful census of all arts organizations in our service area every two years, combining data from NCCS, CDP, and local listings of arts and cultural organizations. In our 2012 census, we counted 2,569 organizations known to us – 82% under $250,000 – obviously the vast majority are small, neither Portfolio nor our other publications claim otherwise.

http://www.philaculture.org/research/21143/number-arts-cultural-organizations-greater-philadelphia-2012

http://www.philaculture.org/research/briefs/number-arts-cultural-organizations-greater-philadelphia-2010

 

Any broad census of arts organizations necessarily captures a more limited set of data. (How many questions is even the US Census able to ask each household? 12?) An in-depth analysis like Portfolio REQUIRES THE DETAILS provided by systems like CDP. Before CDP started in 2004, for example, it was almost impossible to get a good idea of total Government Contributions to arts organizations in Greater Philadelphia, given the vagaries of government budgets, and the fact that not all money came from government arts agencies. While we “only” capture 400+ organizations in our area through CDP, we have a better, more detailed understanding of government funding mechanisms and trends than one could ever get with a more limited system.

 

Like all others on these posts, we are trying to do careful analysis on moving targets using datasets that are not always perfect.

 

 

Nicholas Crosson

Research Analyst

Greater Philadelphia Cultural Alliance

ni...@philaculture.org

215-399-3514 direct

 

 

 

 

From: cultural...@googlegroups.com [mailto:cultural...@googlegroups.com] On Behalf Of Andrew Taylor

Sent: Monday, June 30, 2014 8:04 PM
To: cultural...@googlegroups.com