Community Development and Community Organizing:

Apples and Oranges? Chicken and Egg?⁽¹⁾

Randy Stoecker

rstoecker@wisc.edu

February 2001

INTRODUCTION

Since the 1960s, as the presence and activity of community development corporations, or CDCs, in poor neighborhoods has grown, so has the debate surrounding them. Recently, community development analysts and practitioners have been trying to combine community development with the more politicized community organizing model.

This paper begins by defining and describing these two approaches. Next, it explores the extent to which they are complementary or contradictory--apples and oranges. Finally, it reviews ways of combining them, exploring their chicken-egg relationships.

Community development, in this paper, is defined as nonprofit organizations--CDCs--doing physical development of impoverished communities. CDCs are supposed to be "community-based," having some connection with the residents who live there. They are also expected to do "comprehensive development," creating jobs, housing, safety and other changes (though most emphasize housing). Finally, they are supposed to accomplish all this within the existing political economic system (Stoecker, 1997).

Critics of the CDC model, however, point out how CDCs often fail at projects that left their host neighborhoods in as bad or worse shape than when they started; fold under funding shortages that allow elites to both prevent real redevelopment and blame CDCs for failure; or disrupt neighborhood empowerment by purporting to speak on behalf of a community who they barely know (and who barely knows them) (Stoecker, 1994; 1997). These critics have led a call to bring back community organizing.

Community organizing works in local settings to empower individuals, build relationships and organizations, and create action for social change (Beckwith & Lopez, 1997, Bobo, Kendall & Max, 1991, Kahn, 1991; Alinsky, 1969; 1971). Community organizers have historically focused on building localized social movements in places as small as a single neighborhood. Consequently, the bulk of community organizing occurs "backstage" (Goffman, 1959), building relationships and networks in the quasi-private setting of the neighborhood community (Stall and Stoecker, 1998) that can create a larger social movement.

Community organizing has a much longer history than community development, including the early 20th century settlement house movement and other women-centered efforts (Stall and Stoecker, 1998), the Civil Rights Movement (Morris, 1984), and others. The most well-known influence was Saul Alinsky (1969; 1971) who, in the 1930s, created a community organizing model in Chicago's Back of the Yards neighborhood that was rowdy, bawdy, and confrontational (Finks, 1984). The Civil Rights Movement is the other crucial source of community organizing, though its influence on community organizing practice has been as profound as Alinsky's but has been historically neglected. The accepted founding event of the movement, the Montgomery Bus Boycott, was coordinated through local African American networks and organizations and created a model that would be used in locality-based actions throughout the south (Morris, 1984). Out of these efforts grew the Welfare Rights Movement (Piven and Cloward, 1979) and eventually the famous Association of Community Organizations for Reform Now (ACORN) (Delgado, 1986; Russell, 2000).

Today, community organizing is experiencing a resurgence, with an explosion of small efforts and the growth of better-publicized efforts by the Industrial Areas Foundation (IAF) (Tresser, 1999), by ACORN (1999) and the New Party (1997) in their Living Wage efforts, and by many other groups and networks (COMM-ORG, 2001) including the rapidly expanding National Organizers Alliance (2001) which is supporting and connecting independent and network organizers across North America.

COMMUNITY DEVELOPMENT--COMMUNITY ORGANIZING: APPLES AND ORANGES?

Can these two models--one that works within the system and the other that tries to change it--be combined? Or are they the proverbial apples and oranges? Steve Callahan et al. (1999) argue for combining project-based community development--which delivers social, economic, and housing services to poor communities--and power-based community development--which employs polarizing and militant tactics to develop the power of low-income people and hold officials accountable.

The challenge is that these models are rooted in fundamentally different theories of how society works, which sociologists refer to as functionalist and conflict models. The functionalist model argues that society tends toward natural equilibrium and its division of labor develops through an almost natural matching of individual talents and societal needs. For functionalists, healthy societies maintain some basic degree of equilibrium and place all of their members into the roles for which they are fit. The implication (though few today admit it) is that the poor and the oppressed are supposed to be poor and oppressed. Of course, those who don't belong there (i.e., those who are willing to work hard) are provided new roles. This theory also assumes that people have common interests even when they have different positions in society. Healthy, persistent societies are in a constant state of gradual equilibrium-seeking improvement. Thus, a group organizing to force change is actually unhealthy, as it can throw off equilibrium, and cooperation to produce gradual change is a better alternative (Eitzen and Baca Zinn, 2000). In this model, poor people only need opportunity, not power, and cooperation between the haves and the have-nots is the best means to provide opportunity. But because the model does not recognize structural barriers to equality, it can only provide opportunities determined by existing power holders.

Conflict theory sees no natural tendency toward anything but conflict over scarce resources. In this model society develops through struggle between groups. To the extent that stability is achieved, it's not because society finds equilibrium but because one group dominates the other groups. Conflict theory sees society as divided, particularly between corporations and workers, men and women, and whites and people of color. The instability inherent in such divided societies prevents elites from achieving absolute domination and provides opportunities for those on the bottom to create change through organizing for collective action and conflict.

The CDC model, rooted in the functionalist tenets of common interest and cooperation, can only work if functionalist theory is correct. In other words, there can be no barriers to poor communities rebuilding themselves. The problem is that, while individuals can lift themselves up and attain greatness, not all poor people can lift themselves up simultaneously because there are not enough better spaces available in society--not enough good jobs, not enough good housing. This problem is multiplied when the focus is trying to lift up poor communities, which can only occur if the people in those communities are simultaneously lifted up. If there's no space for all those individuals in the economy, there's no chance for that community. The simultaneous improvement of poor people everywhere requires a drastic redistribution of wealth, violating the fundamental tenets of functionalist theory, which argues that trying to create an artificial equality would actually upset equilibrium.

So what happens in the community development model is that people's need for a transformed economy providing a wealth of good jobs becomes replaced with training programs for people to compete for an extremely limited good job pool. People's need for affordable housing that is controlled by its occupants becomes replaced by a tradeoff between expensive home ownership and affordable rental housing. People's need for high quality health, daycare, and other services becomes translated into sporadic, overcrowded, and inefficient low quality stop-gap programs. Not only can a model emphasizing cooperation and denying class conflict not work to end poverty and oppression, it's not even supposed to work.

The community organizing model is much better suited for attacking the structural barriers that prevent poor communities from lifting themselves up. In a capitalist society, equal competitors make deals because each either has something to offer or something to take away. But when CDCs attempt to make deals with these power holders, they have nothing with which to bargain. They are in the powerless position of begging--for lower loan rates, reduced construction costs, more open hiring practices, etc. CDCs have little to offer as inducement for power-holders to say yes, and little to withhold if they say no. The community organizing model, however, substitutes the lack of money resources with people resources. The bargaining chip poor communities have is their cooperation. If they can collectively withhold their cooperation or, even more powerfully, can disrupt the activities of power holders, they have something to bargain with (Piven and Cloward 1979).

The community organizing model and its conflict theory underpinnings also has limits. When community organizations wrest concessions from corporations or government they often discover that those wins are only as good as the community's ability to implement them. When the Cedar-Riverside neighborhood beat back a government-developer coalition out to displace the neighborhood with massive high-rises, they were faced with the prospect of their existing housing being condemned unless they found resources to fix it up (Stoecker, 1994). When the Dudley Street Neighborhood Initiative won city approval of their neighborhood redevelopment plan, they had to find funding and eventually even do the development themselves (Medoff and Sklar, 1994). ACORN had to create a community development arm when it began winning housing through squatting and other tactics (Russell, 2000; ACORN, 1997). Because of the incompatibilities of the theories on which they are based, many community organizing groups make the transition to development gagging and retching. Some of them destroy themselves in the process (Stoecker, 1995).

Community organizing is necessary to get the power. Community development is necessary to keep it. So what do we do?

COMBINING ORGANIZING AND DEVELOPMENT: CHICKEN AND EGG?

Regardless of how hard it is to combine community organizing and community development, we must figure out how. This is the chicken and egg problem. Which is more important? Which is more powerful? Which comes first? Can you move from development to organizing and actually build power? Can you move from organizing to development without disrupting organizing?

There are two basic strategies. One is to combine organizing and development in a single organization. The other is to separate them into allied organizations.

1. Organizing in Development

The initial efforts to combine organizing and development came as effective organizing groups were forced into doing their own development. Because of the shift in funding during the 1980s from development to organizing, staff and directors of organizing groups found themselves forced into becoming developers. But many really wanted to be organizers and kept looking for ways to bring organizing back (Rubin, 2000).

Given the funding constraints, however, traditional organizing that threatened funders' power was not to be and it is no mistake that community building, consensus organizing, and women-centered organizing are in the spotlight today. The avoidance of confrontation, the lack of focus on structural change, and the absence of conflict in these models makes them well-suited to CDCs.

Community building is defined as "projects which seek to build new relationships among members in a community and develop change out of the connections these relationships provide for solving member-defined problems." (Hess 1999). Linked to Kretzmann and McKnight's (1993) asset-based community development model, and to communitarianism, the emphasis in community building is creating and restoring relationships between community residents. The focus is internal, finding and building the community's own "assets" or "social capital" rather than confronting or negotiating with external power and resource holders (Smock, 1997).

Consensus organizing includes relationship-building but also focuses on moving people from welfare to work, improving school achievement, promoting inner-city reinvestment, and developing housing and businesses, among other things. This model specifically opposes the "us vs. them" model of community organizing (Eichler 1998). The purpose of consensus organizing, consistent with functionalist theory, is to build cooperative relationships between community leaders and business and government to improve poor communities (Consensus Organizing Institute, 2000).

The women-centered organizing model emphasizes relationship building that is not rooted in self-interest but in an understanding of mutual responsibility. And while it does see a structural division in society that holds women back, it also emphasizes that power is infinitely expandable rather than zero-sum, thus reducing the need for conflict. Like the community building model, women-centered organizing emphasizes small group development and has more of an internal problem-solving focus. The goal is as much the development of individuals as it is the development of communities (Stoecker, and Stall, 1998).

Some CDCs are now able to break free of these limited community organizing models through new funding sources. The largest and most well-known effort to help CDCs do community organizing is the $1.5 million Ricanne Hadrian Initiative for Community Organizing (RHICO), sponsored through the Massachusetts Association of CDCs and the Neighborhood Development Support Collaborative. The program supports and trains CDCs throughout Massachusetts to do community organizing (Winkelman, 1998, 1998b).

A similar project to promote community organizing through CDCs was the Toledo Community Organizing Training and Technical Assistance Program sponsored through the Toledo Community Foundation. Over a two year period, ACORN provided training and technical assistance to three CDCs, though one dropped out due to a lack of fit. The Organized Neighbors Yielding Excellence (ONYX) CDC adopted a combined organizing and development group model, where leadership and authority over the organizing effort remained vested in the CDC board of directors, though they gave tacit approval to developing an informal community organizing leadership structure. Conversely, the Lagrange Development Corporation established a relatively autonomous community organizing group, adopting a written code of principles to prevent the CDC from interfering in the organizing effort even while it paid the organizer's salary. The Lagrange Village Council, the relatively autonomous community organizing group, practices a more traditional Alinsky-style community organizing model, using actions and pressure tactics to close problem businesses in the community, improve trash collection, and manage a long drawn-out campaign against a predatory property speculator. ONYX has practiced a much more subdued approach consistent with the community development model, and with fewer subsequent victories.

What are the outcomes of this combined model? CDCs in the RHICO initiative have seen are more community involvement in CDC decision-making, less funder-driven project development, and more effective CDC advocacy efforts. (Winkelman, 1998). This also appears to be the case with the Lagrange Development Corporation and Lagrange Village Council in Toledo, which has gotten a number of problem businesses to shape up or leave, can turn out dozens of people for a demonstration, and has hundreds attend its annual meetings.

There are also important problems. The first problem is the potential restriction on militancy. One of the RHICO CDCs lost government funding when they moved to organizing. However, this CDC continued down the organizing path, weathering the cut and actually freeing itself from restrictive funding (Winkelman, 1998). Other groups are less able to make such bold moves. In Toledo, ONYX's organizing effort has been hindered by the fear of funding loss, and the organization has been threatened with government funding reductions.

A second related problem is the internal conflict that the combination can produce. The East Toledo Community Organization (ETCO), an Alinsky style community organization in Toledo, Ohio, turned to development to support its staff during the 1980s when funding shifted from organizing to development. ETCO began conducting home energy audits, providing advice on how to reduce energy costs. They took on city contracts to board up vacant houses. They got a grant to start a jobs bank. And the organization imploded as infighting between organizing proponents and development proponents broke into open warfare (Stoecker, 1995; 1995b). The Dudley Street Neighborhood Initiative in Boston began by reclaiming a city park from drug dealers, closing area garbage transfer sites, curbing illegal dumping. They also developed a plan to build new housing in their community, fighting off a government redevelopment plan that would have wiped them off the map. They won government and foundation support for their plan, and found a developer who would do the project. But they had problems finding a reliable development partner and ended up doing the development themselves. The time consuming technical and financially risky aspects of managing housing construction badly distracted the organization. (Medoff and Sklar, 1994).

2. Organizing and Development

In this model, organizing and development are separated into different organizations. In a 1997 article I argued that the ideal type model was a small locality-based community organizing group partnering with a high-capacity multi-local CDC. The reason was that CDCs, to be successful, needed technically-skilled (and thus expensive) staff and enough capitalization to do development in higher risk situations. Larger CDCs would be more likely to have those qualities. But because those characteristics would also increase the separation of the CDC from the community, small neighborhood-based community organizing groups were necessary to maintain community control of development. I have had great difficulty finding examples of this model.

The Cedar-Riverside neighborhood redevelopment movement is the source of my model, as they consciously kept their organizing and development activities separate, gaining a combination of political power and redevelopment resources that I have not seen matched since. There were very concerned (and it was from them that my own thinking developed) about the compromised politics of CDCs that might prevent them from truly following the neighborhood's direction and abiding by their demands. At one point the neighborhood's community organizing group brought in a private developer to build housing when its own CDC was behaving too insensitively (Stoecker, 1994).

In contrast to my model of a highly-capitalized CDC partnering with local community organizing groups, more common are cases of large community organizing networks partnering with small development and service organizations. The Toledo Community Organizing Training and Technical Assistance Program discussed above has expanded to a partnership between the two local participating CDCs and an independent ACORN organizing effort now in the city. The Sacramento Valley Organizing Community (SVOC), an organization of nearly 30 predominantly Latino Catholic and African-American Protestant churches across a three-county area in Northern California, is part of the Industrial Areas Foundation community organizing network founded by Saul Alinsky. In one instance SVOC brought 1,800 members to a meeting with area health system officials, successfully demanding 200 jobs. To implement the win, SVOC partnered with the Private Industry Council (PIC), the county welfare department, and a community college to do job training and preparation (Callahan et al., 1999). The famous Communities Organized for Public Service (COPS) in San Antonio practiced a similar model, and ended up directing a large proportion of San Antonio's CDBG budget over a number of year to COPS-defined project. But COPS refused to do any of the development themselves to preserve their organizing focus (Cortes, 1995; Warren, 1995).

Perhaps the most famous case is the NorthWest Bronx Community and Clergy Coalition. The NWBCCC, with a 25-year history, organizes with ten neighborhoods and approximately twenty local religious communities in the Northwest Bronx area of New York City and has spawned a number of CDCs. Because NWBCCC is an affiliate of groups, different sub-coalitions can work on issues they have in common. They consciously put organizing first, understanding the technical constraints placed on development. Consequently, they came up with the idea of "Neighborhood Improvement Plans ... as opposed to fitting into existing programs, leaders were asked to think about what they wanted to see in the area and then we would try to figure out how to get there." (Buckley, n.d.). Two of the CDCs formed through NWBCCC--Fordham Bedford Housing Corporation and Mount Hope Housing Company--are highly capitalized, multi-local CDCs with hundreds of employees and thousands of housing units. The NWBCCC's housing committee, or neighborhood groups, determine projects and then negotiate with one of the CDCs about how to implement it (Dailey, 2000).

The problem with this multi-organizational model is not its theoretical desirability but its practicality. In many poor communities, even with the resurgence of community organizing, we are more likely to find a neighborhood CDC than an organizing group. Corvallis Neighborhood Housing Services, in Corvallis Oregon, is looking for community organizing groups to partner with so they can "direct development to organized neighborhoods instead of building a project and then organizing around it." But the neighborhoods in Corvallis, with rare exception, are not organized (Smith, 2000). It makes no sense to tell CDC staff with a knowledgeable commitment to community organizing that they shouldn't do it just because they don't fit the ideal model.

There are also risks in separating organizing and development into independent organizations. When separate organizations forget their complementarity, they can compete rather than cooperate. In a single neighborhood, the infighting that can occur within a CDC trying to do both organizing and development can also occur between a neighborhood-based CDC and a neighborhood-based organizing group.

So how do we decide whether to implement the organizing in development or the organizing and development model? Based on lessons from RHICO (Winkleman, 2001) and the Toledo Community Organizing Training and Technical Assistance Program, here are some guidelines.

First, it is important to assess what exists in the neighborhood already. If there is an organizing group and a CDC, it is probably counterproductive for the CDC to also do organizing. Instead, they should find ways to partner with the organizing group. If there is a CDC but no organizing group in the neighborhood, it is important to assess the CDC's readiness and capacity to do community organizing. How knowledgeable is the executive director about organizing in general and different organizing models? Is there anyone else highly skilled in organizing on staff? Is there an organizer in place and, if so, what do they know about organizing in general and which organizing model do they prefer? Who is or would be responsible for supervising the organizer? How are leaders (board members) identified/recruited? Are leaders elected or appointed (elected is better for organizing)? What do leaders know about organizing in general and different organizing models? How do organization leaders and director respond to a series of organizing vs. development dilemmas (such as doing an action against a bank that also gives loans to the CDC projects)? What procedures are in place for replacing staff and leadership without losing internal organizing culture? Does the CDC have a broad mission statement that could easily include community organizing? In general, the more skill, knowledge, and support for community organizing, the more successfully a CDC will be able to develop its organizing capacity.

Second, it is important to assess what exists beyond the neighborhood. If the neighborhood has neither a CDC nor an organizing group, are there high-capacity CDCs or culti-local organizing networks working in the area? If there are both, what are their histories of working cooperatively across the organizing-development divide? If there is a neighborhood CDC, but it is not structured to support organizing, what is its history of working cooperatively with other organizations? Also, what is its level of power? Numerous neighborhoods have small CDCs which do little or nothing, and would be better replaced with a combination of community organizing and high-capacity community development.

It is clear there is no single correct way to combine organizing and development. It is also clear that the two models, contradictory as they are, are inseparable. So we must continue the search for ways of combining them where organizing will not be compromised and development will not be limited.

REFERENCES

ACORN. 1997. Capital & Communities: A Report to the Annie E. Casey Foundation on ACORN's Work to Revitalize Low and Moderate Income Communities. http://www.acorn.org/ACORNarchives/studies/c-and-c/capital-and-communities.html

Alinsky, Saul. 1969. Reveille for Radicals. New York: Vintage

Alinsky, Saul. 1971. Rules for Radicals. New York: Vintage

Beckwith Dave, with Cristina Lopez. 1997. "Community Organizing: People Power from the Grassroots." COMM-ORG Working Papers Series, 1997 Working Papers. http://comm-org.wisc.edu/papers.htm

Bobo, K., Kendall, J. & Max, S. (l991). Organizing for social change: A manual for activists in the l990s. Washington, DC: Seven Locks Press.

Buckley, Jim. n.d. statement, http://www.nwbccc.org/History/Stories/Buckley.html)

Callahan, Steve, Neil Mayer, Kris Palmer, and Larry Ferlazzo. 1999. Rowing the Boat With Two Oars. paper presented on COMM-ORG: The On-Line Conference on Community Organizing and Development. http://comm-org.wisc.edu/papers.htm.

COMM-ORG: The On-Line Conference on Community Organizing and Development. 2001. CO Groups and Networks. http://comm-org.wisc.edu/orgs.htm

Consensus Organizing Institute. 2000. http://www.consensusorganizing.com

Cortes, Ernesto. 1995. Remarks as panelist, Special Session on "The Legacy of Saul Alinsky" at the American Sociological Association Annual meetings.

Dailey, Mary. 2000. message posted on COMM-ORG: The On-Line Conference on Community Organizing and Development. http://coserver.sa.utoledo.edu/pipermail/colist/2000-December/001212.html

Delgado, Gary. 1986. Organizing the Movement. Philadelphia, Temple University Press, 1986.

Eichler, Michael. 1998. Organizing's Past, Present, and Future. Shelterforce Online. http://www.nhi.org/online/issues/101/eichler.html

Eitzen, Stanley, and Maxine Baca Zinn. 2000. In Conflict and Order: Understanding Society, 9e. Boston: Allyn and Bacon.

Finks, P. David. 1984. The radical vision of Saul Alinsky. New York: Paulist Press.

Goffman, E.(1959).The presentation of self in everyday life. GardenCity, NY: Anchor.

Hess, Doug. 1999. Community Organizing, Building and Developing: Their Relationship to Comprehensive Community Initiatives. Paper presented on COMM-ORG: The On-Line Conference on Community Organizing and Development. http://comm-org.wisc.edu/papers.htm.

Kahn, S. (1991). Organizing: a guide for grassroots leaders. Silver Springs: MD. NASW Press.

Kretzmann, John P. and John L. McKnight. 1993. Building Communities From the Inside Out. Evanston, IL: Center for Urban Affairs and Policy Research Neighborhood Innovations Network, Northwestern University.

Medoff, Peter, and Holly Sklar. 1994. Streets of hope : the fall and rise of an urban neighborhood. Boston: South End Press.

Morris, Aldon. l984. The origins of the civil rights movement: Black communities organizing for change. New York: Free Press.

National Organizers Alliance. 2001. http://www.noacentral.org/

New Party. 1997. Living Wage and Campaign Finance Reform Initiatives http://www.newparty.org/reforms.html

Piven, Frances Fox, and Richard Cloward. 1979. Poor people's movements: Why they succeed, how they fail. New York: Vintage.

Reitzes, Donald C. and Dietrich C. Reitzes. 1987. The Alinsky legacy: Alive and kicking. Greenwich, CT: JAI Press.

Rubin, Herb. 2000. Renewing hope within neighborhoods of despair: the community-based development model. Albany : State University of New York Press.

Russell, Dan. 2000. Roots Of A Social Justice Movement (1970-75). http://www.acorn.org/history-content.html

Smith, Ken. 2000. message posted on COMM-ORG: The On-Line Conference on Community Organizing and Development. http://comm-org.wisc.edu/pipermail/colist/2000-December/001213.html .

Smock, Kristina, 1997. Comprehensive Community Initiatives: A New Generation of Urban Revitalization Strategies. Paper presented on COMM-ORG: The On-Line Conference on Community Organizing and Development. http://comm-org.wisc.edu/papers.htm.

Stall, S. and Stoecker, R. (1998). Community organizing or organizing community? Gender and the crafts of empowerment. Gender and Society, 12, 729-756.

Stoecker, Randy. 1997. "The Community Development Corporation Model of Urban Redevelopment: A Critique and anAlternative." Journal of Urban Affairs 19:1-23.

Stoecker, Randy. 1995. "Community Organizing and Community Development in Cedar-Riverside and East Toledo: A Comparative Study." Journal of Community Practice, 2:1-23.

Stoecker, Randy. 1995b. "Community, Movement, Organization: The Problem of Identity Convergence in Collective Action." The Sociological Quarterly 36:111-130

Stoecker, Randy. 1994. Defending Community: The Struggle for Alternative Redevelopment in Cedar-Riverside. Temple University Press.

Tresser, Thomas. 1999. The Work of the Industrial Areas Foundation--Building Strong Citizens Organizations for Power --Action --Justice. http://my.voyager.net/ttresser/iaf1.htm

Warren, Mark Russell. 1995. Social Capital and Community Empowerment: Religion and Political Organization in the Texas Industrial Areas Foundation. unpublished PhD dissertation, Harvard University.

Wessel, Paul. 2000. message posted on COMM-ORG: The On-Line Conference on Community Organizing and Development. http://comm-org.wisc.edu/pipermail/colist/2000-December/001212.html .

Winkelman, Lee. 1998. Organizing Renaissance: Twin Cities CDC Leads Exploration of Organizing by Massachusetts CDCs. Shelterforce. September/October. http://www.nhi.org/online/issues/sf101.html

Winkelman, Lee. 1998b. Organizing: An Investment That Pays. NeighborWorks Journal. Volume 16, #2, Winter. http://www.nw.org/resources/pub/nwjournal/1998/vol16no2/

NOTES

1. This paper is prepared for an edited collection by Ron Hayduk and Ben Shepard From ACT UP to the WTO: Urban Protest and Community Building in the Era of Globalization under review with Verso. It is part of a larger project studying the relationship between community organizing and community development, supported by a grant from the University of Toledo Urban Affairs Center. Many thanks to Lee Winkelman for comments on an earlier draft.