How Google and Social Entrepreneurs Perpetuate the Digital Divide Among Nonprofits
In the past 10 years, I have been working to address the digital divide, which is the gap between those who have access to and training with technology and those without. I serve as the Executive Director of TechMission, which runs the largest association of Black and Latino led nonprofits addressing the digital divide and manages UrbanMinistry.org, which is one of the most visited web portals of Black and Latino nonprofit leaders. During that time, I have seen many effective initiatives in addressing the digital divide. At the same time, I’ve seen many efforts that have been very well-intentioned, but in the end may have only made matters worse.
I grew up in the inner-city, but later went to MIT and co-founded the Internet Telecoms Consortium with one of the fathers of the Internet (David Clark) to study the social and business implications of the Internet. The most important thing I learned while studying complex systems at MIT is that often you make a change to the system and the result is exactly the opposite of what you were expecting.
Case in point: my wife and I went to Woodstock ’99—a complex system of people, alcohol, drugs, bands and unusually hot weather. At the concert, there was a nonprofit group wanting to promote peace, and it gave out thousands of “peace candles.” The result was the opposite of what they had intended because they had not thought through the consequences of giving candles to thousands of people who were inebriated and enraged from the extreme heat and exploitive prices at the concert. The result was that people used the candles to start hundreds of fires and burn everything in sight. A group wanting to promote peace at Woodstock unintentionally became a catalyst for riots, and in the worst circumstances, this is how complex systems work.
Addressing the digital divide and trying to help under-resourced communities is an extremely complex system. My assessment is that some of the largest efforts to address the digital divide by social entrepreneurs, including those at Google, may have unintentionally made matters worse among nonprofits.
Is Google Grants Hurting Black and Latino-Led Nonprofits?
In the past few years, Google has “given away” $300 million worth of free advertising to nonprofit organizations through Google Grants. Following Google’s value of employee input, Google leaves most of the grantmaking decisions to regular employees. In the United States, Google’s employees are disproportionately White and Asian and from elite schools. These employees then give grants to organizations with interest like theirs and people they know, which most often are also White or Asian and from elite schools.
This creates a bias in the process, but that is common in the nonprofit world. In fact, according to a report from the Greenlining Institute, foundations only give 3.6% of their funds to organizations led by people of color although people of color make up 52.4% of poverty in the USA. This bias is shown graphically in the adjacent diagram.
If that bias were the only effect Google Grants were having, then they would be no different than most foundations. The problem is how Google Grants works, by giving away “virtual money” that people can then bid on ads, they have essentially flooded the market for nonprofit ads with $300 million in virtual money.
Economists would say that this acts as a “tax” on other users, so that only $150 million is donated by Google, but the other $150 million comes from the increased bid cost for other advertisers, which are primarily nonprofits. Let’s say that in the USA, Google gives 95% of its grants to White and Asian led nonprofits, while only 5% to Black and Latino (or other) nonprofits (which is a reasonable projection based on the research we compiled). Ideally, grants would be distributed to more closely reflect the demographics of need, and 52.4% of grants in the USA would go to Black and Latino-led nonprofits since they represent 52.4% of poverty in the nation.
The end result of this bias is that Google Grants, rather than helping Black and Latino led nonprofits, is actually taxing them to give discounts primarily to White nonprofits. The model I’ve generated shows that the effect of Google Grants on Black and Latino-led nonprofits is that they are “taxed” by about $7.3 million by paying increased advertising fees to provide a subsidy to White-led nonprofits.
Google Grants is not the only example of starting with good intentions to change a system, but having unintended effects, as many tech-focused nonprofits serve as other examples.
Our organization did some research on the board and staff demographics of the leading technology-focused nonprofits, and found that on average about 80% of their staff are White (and often from elite schools). These organizations develop strategies, policies and values that reflect their staffing and boards. Our staff have gone to several of these conferences led by these organizations, and the conference attendees are also more than 90% White. We have followed their grantmaking processes and seen that more than 80% of their grants go to White-led organizations. These demographics of technology focused nonprofits based on our research are shown in the diagram below. The significant factor is that in every category, there is dramatically lower percentage of people of color as compared to the demographics of the low-income communities they are serving. The result is that most technology focused nonprofits are more reflecting the digital divide rather than transforming it. To really transform the digital divide, it requires representation of those on the other side of the divide at every level in the organization.
The same pattern applies to much of the new social entrepreneurship movement. Most the emerging social entrepreneurs that are getting funded are led by White people from elite schools. They are establishing a culture and values in the social philanthropy sector that largely reflects White cultural values from elite business schools. The end result is that these social ventures get all the funding, squeezing out more grass-roots Black and Latino led organizations. This means that too often social ventures are perpetuating the same divides that they are supposedly addressing.
How Social Entrepreneurs Can Avoid Doing More Harm than Good
No one wants to try to do good, and then have the results only cause more harm than good. So what are social entrepreneurs to do? Am I promoting a Malcolm X philosophy that says that White people should just go home?
Of course not. The fact that I am White, and leading a nonprofit, shows that I believe that there is another way.
The reason why social entrepreneurs are encountering unintended consequences in bringing change to complex systems is that many do not understand one of the most important aspects of the system—the culture of at-risk communities. Social entrepreneurs need to strive to reflect the culture and diversity of those we are serving at all levels in nonprofits. At TechMission, the people of color are the majority at almost every level of our organization (board, senior staff, etc.). The result is that when you look at the demographics of those using our websites like UrbanMinistry.org, we have almost twice the percentage of people of color represented as our nonprofit counterparts (as shown above).
We have researched and written extensively on how nonprofits and funders can avoid this bias (see additional resources below). Based on that research and experience, we provide the following recommendations to social entrepreneurs:
1. Be intentional about reaching out to faith-based organizations doing social services.
Why? Our research indicates that about two-thirds of Black and Latino-led nonprofits are in churches or other faith-based organizations. Comparitively, about two-thirds of White-led nonprofits are secular. This means that on average, faith-based organizations are about twice as likely to be Black or Latino-led. This has important implications for funders, because if they will not fund faith-based organizations, then they are cutting their chances of funding a Black or Latino-led nonprofit by 50%. Most funders will support faith-based organizations if their focus is on social services and they do not discriminate in who they serve. Of more than 50 other tech companies that we have identified, Google Grants has, in effect, the most restrictive policy toward faith-based social service organizations. The faith-based social services sector represents about one-third of the total social services sector, but appears to only be less than 1% of Google Grant awardees. In the 1950’s there were literacy tests for voting that excluded a large number of Black voters. Today this has been replaced with religious tests for funding that is excluding the majority of Black and Latino-led nonprofits. Restrictions to not support faith-based organizations can communicate a paternalistic attitude that secularization is advancing the cause of good by “educating superstitious natives” to believe in the doctrines of elite academia. It is just the “bad missionary model” but with a new name. They fail to recognize that globally, 97.5% of people profess some faith. Often by having restrictions on faith, these organizations think that they are blocking right-wing religious groups like Jerry Fallwell, but in reality they are usually blocking Black and Latino religious leaders that could become the next Martin Luther King, Jr.
2. Have diversity measurements and affirmative action at every level in the organization.
Why? Values and culture are set by those in power in an organization. If a nonprofit has a significantly different set of values than the community it is serving, then that community will need to assimilate those values to get support. Religious missionaries in the 20th century have been replaced with nonprofit missionaries in the 21st century. If a nonprofit does not reflect the culture and values of the community it is serving, then it cannot help but be like the “missionaries” that teach the “natives” how to talk White, dress White, and respond to outcome measures reflecting values of the White community. The only solution is to have diversity measurements and affirmative action at every level. This enables diversity at every level (constituents, entry level staff, middle management, senior management, board and grantees) needed for a leadership pipeline.
3. Immerse yourself in the community you are serving and learn from it.
Why? The difference between strong faith-based leaders like Martin Luther King, Jr. and Mother Theresa and “bad missionaries” is that the former immersed themselves in the communities they were serving. At TechMission nearly all of our staff and board live in at-risk communities, and the majority come from low-income backgrounds. It is only when you “go native” to the point where at-risk communities become “your people” that you are not acting as a nonprofit missionary. Immersion means not only your geographic location, but also your friendships and community.
Changing the Culture of Social Entrepreneurship Programs
Social entrepreneurs are often trying to change complex systems, but unfortunately, without understanding those systems they can often do more harm than good. The primary element of understanding that is missing here is an understanding of the culture of those being served by social entrepreneurs. For the social entrepreneurship movement to be successful, it must change its culture.
Currently the culture of the social entrepreneurship movement primarily reflects that of elite schools which have social entrepreneurship programs. This culture is extremely antagonistic to the cultures in most at-risk communities.
Organizations like Google and other social philanthropists need to make sure that those in their grantmaking process reflect the demographics and culture of those they are trying to reach. They also should consider eliminating restrictions that will not fund social service organizations if they are faith-based, while being intentional about creating a culture that welcomes religious organizations, where the majority of Black and Latino nonprofit leaders are.
Because Google Grants makes up about one-third of the traffic most of the tech nonprofits, the increased traffic (and resulting links from visitors) from that visibility increases their web search rankings. The end result is that the Google Grants process is that White nonprofits receive significantly higher search rankings than Black and Latino nonprofits. Given Google’s market share, overall this has a very significant affect of "Whitening" Google’s search rankings within the nonprofit sector. If Google does not bring its policies more in line with that of its peer technology companies, then it risk facing a public relations issues of being perceived as having an racial bias compare to other technology companies by "Whitening" its search rankings among nonprofits.
People of color already represent the majority globally, the majority of poverty in the USA, and the majority in many major US cities. Soon they will represent the majority in the nation as a whole. As these demographic shifts happen, if the social entrepreneurship movement does not shift its culture to more closely reflect the cultures of those it is serving, then it risks becoming relegated to the sidelines of history.
Sources:
http://www.aecf.org/upload/publicationfiles/executive_transition_survey_report2004.pdf
http://greenlining.org/publications/index.php?initiative=democratizing-philanthropy
http://www.slideshare.net/rosettathurman/race-matters-in-nonprofits-promoting-diversity-in-our-profession
http://www.independentsector.org/programs/research/Charitable_Fact_Sheet.pdf
http://answers.google.com/answers/threadview?id=96516
http://hirr.hartsem.edu/research/fastfacts/fast_facts.html
http://www.techsoup.org/stock/restrictions.asp
Additional Resources
http://www.urbanministry.org/foundationdiversity
http://www.urbanministry.org/fundingbias
http://www.urbanministry.org/onlinesegregation
http://www.urbanministry.org/reconciliation-across-social-class
About Andrew Sears
Andrew has been working to address the digital divide since 1998 and is the founder and executive Director of TechMission. He grew up in an inner-city environment, and later received his MS in Technology and Policy and MS in Computer Science at Massachusetts Institute of Technology (MIT). At MIT, he co-founded the Internet and Telecoms consortium, a multi-million dollar research group focused on determining the social and business implications of the Internet. He has also worked as an independent consultant on technology strategy with clients including St. Paul Venture Capital, Sprint and several Boston startups.
Special thanks to Natalie Parrague who provided research and editing support for this project.
Background on TechMission and this Issue
TechMission's national partners represent over 30,000 local organizations, 4 million volunteers serving over 40 million low-income individuals in the USA. TechMission ranks among the top 10 organizations focused on nonprofit technology, and among this group we have the highest representation from Black and Hispanic communities. Our websites received over 1.15 million unique visitors in the first half of 2009. We started doing this research because we saw that with the current economic crisis, the Black and Latino led nonprofits were the first to see a loss of funds, so we started investigating why that was. The first area we saw was the religious bias. Given that Google has 75% market share of search, and they have the most hostile attitudes towards faith-based organizations, we wrote Google letters (sent to Sadie Ferguson, Alana Karen, Larry Brilliant, Gregory Miller and the Google Grants office) on July 17, 2008, Feb. 11, 2009 and March 19, 2009. We never heard a response. As we continued our research, we realized that the issue was most common among most social entreprenuers in the tech sector and foundations, and that there was also a significant bias across race. Since then, we have written a series of articles and presentations (listed above) to raise awareness to this issue.
| Attachment | Size |
|---|---|
| Nonprofit_Digital_Divide.xls | 178.5 KB |
- Andrew Sears
- Andrew Sears
- Black Christians
- Culture/Ethnic Identity
- Fundraising
- Computers/Technology
- Racial Reconciliation
- Social Class
- Latino Christians
- Entrepreneurship
- Faith Free Content
- Microfinance
- Reconciliation & Culture
- Business
- Company Customer
- Human Interest
- Labor
- Law
- Person Career
- Person Education
- Social Issues
- Technology
- USD
RSS
Response to Comments
Since sending this article, I've received quite a few comments. I have been encouraged that about 80% of them have been positive, such as, "You seem to be revealing some dirty little secrets about Google and by extension about the philanthropic community in general that many in the foundation community may not want revealed." About 20% have been negative, of which most can be summarized by "Don't talk about race because if you talk about it, you only make it worse."
With any argument, people will try to poke holes in it, but most of the people trying to find holes are not actually looking at the data. While only 3.6% of foundation funding goes to nonprofits led by people of color, people of color make up 16.5% of nonprofit leaders. The percentage of nonprofit leaders is still more than 4.5 times the percentage getting grants. This ignores the fact that if you don't fund nonprofits led by people of color, that will result in a lot less nonprofits led by people of color. So it becomes a cycle. There needs to be affirmative action in funding to address that cycle and to address the cultural bias. I'm not talking about "affirmative action" that gives handouts. I'm talking about affirmative action that offsets the ongoing cultural and other bias that exists in the process. You cannot be color blind and have affirmative action.
While not all nonprofits are serving the poor, it is my understanding that nonprofits serving the poor is the largest segment of the nonprofit sector. I can see where one might argue that the 52.4% figure is a little too high given all the nonprofits that don't serve the poor. But given the number of nonprofits serving the poor, a just number is much closer to 52.4% than it is to the existing 3.6%.
As for Google's response that they don't pay attention to race--that is exactly the point. I never said that they intentionally discriminate. I said that there seems to be an unintentional cultural bias in the process. Google states that the goal of their grants program is to serve "the globally diverse audience that our search engine and tools serve." They are not doing that.
There is a shortage of data and studies that have explored how the demographics of nonprofit leadership affects the effectiveness of the organizations. I have looked for them, and unfortunately it is not an area that is very well studied emperically. My assessment is that indegenous organizations are more effective, which comes from several sources.
What is interesting to me is the difference in response that I'm getting from people based on their cultural context. The comments I've heard from people who live in an urban community and who have people of color as the majority of their friends (regardless of their race or class background) have all responded overwhelmingly favorably with a "Thank God someone is saying this." The response from others is mixed.
The difference is that if you are in the 'hood every day then you see the same things that I see that make it obvious:
Because of what I've witnessed first hand, I started using my MIT background to do research to document what I've seen first hand with more thorough evidence. The data is pretty compelling, and it is hard to argue with data. While some of the arguments presented could arguably push the data I presented 10-20% in one direction or the other, they ignore that there is a 450% bias between grantees and nonprofit leaders and a 1450% bias between grantees and poverty statistics. Even if you cut those numbers in half, there is still a major bias. For people who would respond critically, I would post these three questions.
Thoughts?
I like my arguments without fallacies.
Mr Sears has written an interesting article. However I like my arguments to be logical and not just presentations of wishful thinking. The article is flawed in it's analysis in a number of key ways. Here are the three largest flaws as I see them:
1. The single largest flaw is the key metric in which he's using a "Begging the Question" style of argument.
As he writes in his "cover letter" message and throughout the paper, "...the issue of funding bias in that foundations only give 3.6% of their funds to organizations led by people of color although people of color make up 52.4% of poverty in the USA."
The hidden flaw here is the assumption that all funding is going towards poverty related issues, and that therefore the ratios of funding as a whole should match his poverty demographic target of 52.4%. Obviously not true.
After all, I ask you, how does he account for the fact that 0% of foundation grants are given to organizations headed by Dolphins, Whales, or Trees, when they are so clearly the "population" being served by many foundations? I'm being tongue in cheek, but many non-profits are geared towards medical and environmental issues (just to name two larget issue groups that come to mind quickly) which don't fit into the poverty issue. So, his 3.6% number is meaningless to this discussion unless you remove all these other fundraising issues from the calculation.
That flaw sort of makes most of the rest of the calculations moot.
2. The next flaw is of the "Confusion of Cause and Effect" type.
The contention is "because most Google grants are going to non-people of color lead non-profits, Google is perpetuating the divide". This is flawed in it's analysis. Specifically, it ignores the basic fact that most non-profits are not led nor run by people of color. It would be correct to say "most grants go to non- Black and Latino led non-profits because there are far fewer of those non-profits than white led non-profits. Google isn't a factor.
(The Elite schools point is similar. One would expect more non-profits to be led by people who went to better schools because they are (in theory) better equipped to start, run, be on the board, etc. of non-profits.)
3. Lastly, the argument put forth to promote grants to faith based social service groups is a "Red Herring" argument.
The proposition is put forth that "Our research indicates that about two-thirds of Black and Latino-led nonprofits are in churches or other faith-based organizations. Comparitively, about two-thirds of White-led nonprofits are secular. This means that on average, faith-based organizations are about twice as likely to be Black or Latino-led." This ratio is then used to urge people to donate to non-secular nonprofits.
However while this is an interesting number it's not relevant. The statement made a few sentences lower: "The faith-based social services sector represents about one-third of the total social services sector..." voids the argument. Using Mr. Sears goal of parity, it's clear that at best no more than one third of giving should be going to faith based organizations regardless of if they are POC or White led. Perhaps within that third one could consider if they are People of Color led or not, but that leads us back to argument 2.
In short, there are numerous other fallacies and conclusions drawn from false premises. (For example, using a statistic that 97.5% of people profess some sort of faith to validate increasing grantmaking to faith based organizations. That's another Red Herring argument, the numbers are unrelated in that I could turn it around and say "since 97.5% of people have faith, it's clear that we're already giving most money to nonprofits who are faith based.)
I don't doubt Mr. Sears passion, but this feels to me like an article meant to say "Give my group and groups like me more money", and uses Google for "advertising the article".
This is not to say that many faith-based, people of color, white led, (and all variations) aren't doing excellent work, I believe they are. It's just that I believe arguments are best if they are actually rigorous and true.
To accuse Google, with it's Faith-blind, Color-blind policies of making things worse is fallacious at best. And while this isn't related, it ignores the benefits of such policies (which I won't go into but which I believe are large.)
Thanks, Steve.
E-mail from "a Google spokesperson."
Just to give an update, I received this E-mail from Google in response. It sounds like a fairly standard PR department response...
"The Google Grants program provides free, AdWords advertising to non-profit charitable organizations around the world. Launched in 2003, the program is designed to help non-profit groups reach a global audience online and increase awareness, visibility and website traffic of their various causes. To date, Google Grants has awarded free, in-kind advertising to thousands of non-profit groups whose areas of focus range from science and technology to the environment and the arts.
In accordance with legal requirements, U.S. organizations must have current 501(c)(3) status, as assigned by the Internal Revenue Service, to be considered for a Google Grant. We also employ a strict set of guidelines when reviewing applications to ensure that the process is consistent and objective. Factors such as the ethnicity, education, or any other demographic information about an organization’s ownership and staff are neither collected nor considered when reviewing applications.
For more information about our eligibility criteria, please visit http://www.google.com/grants/details.html#eligibility.
Clarifications
Andrew: You make some really important and valuable points in your post and study. There is no doubt that the tech world lacks the presence and active participation of a more diverse community. And I too often notice the lack of diversity at tech events and among social enterprise organizations "representing" poor and underserved communities - both on a local, national, and global scale. It's a REAL problem that needs to be addressed. but I would take issue on a couple of points relative to your Google study.
First, I know one of the people that started the Google Grants program and they are highly ethical. that's not to say that what your report indicates is not true, but have you given Google a chance to explain their decision making process and/or respond to your points directly?
Second, does Google know the color/ethnicity of Google Grant applicants? My understanding is that it is a fairly anonymous process where grants are approved to an organization without knowledge of the racial/ethnic composition of applicants?
Third, in my experience just having a more diverse community and better representation from underrepresented groups does not always guarantee that those most in need and the "unvoiced" are truly heard and benefited. It's important to move toward full representation, as heroic programs like Affirmative Action have taught us so well. But we also need to imbue organizations with an open, representative, and responsive consciousness - not just fill seats for the sake of appearances and a false sense of diversity.
Complex problems, another perspective
Andrew, with your MIT background you've provided a thorough analytical overview of this problem. However, I would invite you to look at this issue using a couple of tools I've been working with. I use maps, created with Geographic Information Systems (GIS) databases, to show where poverty is concentrated in Chicago, where poorly performing schools are located, and where non-school tutor/mentor programs are located in the Chicago area. We launched an interactive version of this in March, which allows users to zoom into different parts of the region and build their own map analysis. You can view this at http://www.tutormentorprogramlocator.net. Using a tool like this, leaders could determine if a needed service is even available in a particular zip code. If a program is in a neighborhood, all sorts of people could be using their networking and marketing to help those programs get the consistent flow of volunteers and donors it takes to build a constatnly improving organization. If there neighborhood does not have a needed service, the maps will show this void. A set of asset maps, can then be used to identify businesses, faith groups, colleges, hospitals and others who might help fill these voids with new programs. Using Geographic maps could help build a better distribution of resources, and needed services. However, another type of mapping should be developed to help service providers, and donors, begin to follow "blueprints" that would show actions needed to repeat over many years to achieve a desired result. I've been using a free CMAP tool to build blueprints of my ideas. If you visit the strategy map at http://tinyurl.com/tmc-strategy-map you can see that my map shows the ultimate aim of a tutor/mentor program ought to be that a youth is in a job/starting a career by age 25. If the kids in an existing program are only in elementary school, funding and volunteering would need to be provided for another 10-15 years for those kids to be age 25 and working instead of in prison, or worse. These strategy maps show a role that business should be taking to help make this happen. In this type of analysis I'm not talking about the race or mix of people in the organizations who do this work. I'm talking about a creating a system that makes a distribution of needed services, such as volunteer-based tutor/mentor programs, available in more places where they are needed. I think such tools might do a lot more to convince decision makers at Google and other places to provide time, talent and resources into poverty neighborhoods, and might even be used to connect faith communities from affluent parts of a region, with faith communities and non profits working in the high poverty areas of the same region. I write about this on my http://tutormentor.blogspot.com blog and hope we can exchange ideas in ways that increase the number of people who use these discussions to decide where and how to be involved.
Whose judgment do we trust?
So many good issues raised! * Our church's youth just returned from a short-term mission to a different city's inner city, where they happily followed the neighborhood's guidance, pulled tree-size "crack weeds" (so called because they're where the dealers hide) and created a community garden. Would they have taken direction to create such an admittedly short-term solution in our own city, I wondered -- or would their parents have helped them "understand" that focusing on such short-term perspectives were part of the problem in "these" neighborhoods, and scotched the effort? * I work in a neighborhood that's up for a Neighborhoods of Promise grant. The people driving the plan, whether black or white, are sufficiently unfamiliar with the neighborhood that they can report, straightfaced, that in a community meeting one of the assets this community said the community would like is a dry cleaner. This in a neighborhood where the AMI is under $30K, fewer than half the students are at grade level in any subject, and two-thirds of the African American males don't graduate from high school. They aren't talking to the community; they're talking to the people beginning to gentrify the neighborhood who are comfortable talking with them. * In this county and state, we have invested huge amounts of money since 1994 in quality preschool education for 4 and 5 year olds -- allegedly to make sure that all children are "ready for and succeeding in school." But by grade 3, any gains demonstrated at kindegarten are lost. Completely. So is this really a program to help poor children succeed in school ... or is it to make sure that university educated women know that their children will be in quality day care centers when they go to their professional jobs? I could go on ... Thanks for the email, Andrew.
Post new comment