Robert Hanneman : ResearchEarlier projects:
Simulation and theory construction: Numerical simulation is an under-utilized approach to understanding in Sociology that falls between quantitative/statistical generalization on one hand, and qualitative/deductive reasoning, on the other hand. Simulation models of concrete systems, or of highly abstract theories, allow us to conduct systematic thought experiments to understand the implications and dynamics of systems. I have used, primarily, the "systems-dynamics" approach to build non-linear differential equation systems representing a range of phenomena from social psychological processes of stress buffering to the dynamics of rise-and-fall of imperial systems.
Methodology: In the course of teaching a variety of courses in statistics and social network analysis at the undergraduate and graduate levels, I've developed a number of web-sites, and a couple text books. This is not an area of original research; rather, it emphasizes application on a fairly wide array of quantitative methods to a fairly wide array of problems.
Medical care and education delivery systems: The ways that the organizational systems delivering social services are structured may be very consequential for their performance. Working closely with Jerald Hage and J. Rogers Hollingsworth, a number of my earlier publications focus on the structure (e.g. degree of centralized control over revenues) and performance (e.g. equality of access to services) of medical care and educational systems in several Western European nations, and the United States from the late 19th century to the late 20th century.
Welfare state: Variability in the development of state structures in the early-industrializing nations has been both a cause and a consequence of variability in the material and institutional structures of those nations. In some of my work, I explore (often with Jerald Hage) some of the causes of variation in welfare state development (esp. differences in the structure of education and welfare programs) and some of the consequences (variation in levels of income inequality and taxation.
Economic and organizational studies: Cooperatively owned enterprises sometimes act like "capitalist" enterprises by employing non-member labor. Working with Raymond Russell, the growth and variation in this tendency among Israeli Kibbutzim was explored using organizational theory. Also with Russell, patterns of use of part-time and contract labor in Russian firms (public, mixed, and private) was examined as a function of size, industry, and sector.
The military and politics: The nature of the relationship between defense contractors and the U.S. DOD was examined using dependency-exchange theory in my Master's thesis. More recently, the role of the military in political regimes for all world nations between 1950 and 1990 was examined.
Current projects:
Economic Sociology
The functions of trade/employers associations: An analysis of change in the activities of the Salt Producer's Association (later: Salt Institute) from 1915 until the mid-1990s. Using minutes of the meetings of the trade association, the goal is to seek to explain the changing functions of the association over time. In early years, for example, there was a great deal of emphasis on industrial governance; in later years, the primary focus has been on legislative and regulatory lobbying and public relations. The changing functions of the association are seen as a consequence of changes in the structure of the industry, as well as changes in the environments of the producers.
Long-term population ecology of economic organizations: A description and analysis of the population dynamics of establishments and organizations in the United State's dry salt industry from colonial times until 2000. This is a monograph-length project that describes the rise and fall of dry salt production at various locales in the United States, and examines changes in the organizational structure of individual producer's communities and the industry as a whole. The spatial distribution of demand and technology have set broad limits on industry structure; competition and changing organizational forms (e.g. the rise of the professionally managed, multi-establishment enterprise) have determined the demography of establishments and enterprises in the industry.
Higher education
Perceptions of niche in higher education: Using data collected by Steven Brint (who is the lead author on this project, which includes Mark Riddle as a co-investigator) we examine data from the chief executives of a sample of U.S. institutions of higher education. The data report which other institutions each executive regards as "similar" to his/her own, and which institutions the executive would like his/her own to resemble in five years. The perceptions of executives are used to see if the commonly used Carnagie classification, and several others, correspond well to the way executives think about their niche and institutional aspirations.
Institutional effects on research productivity: In this study (with Steve Brint as the lead co-PI), we have proposed to study whether institutional pressures may account for an apparent "flattening" of the rate of growth in scientific journal research publication by American scientists since the early 1990s. The study will combine data on the research productivity of scholars at a sample of research universities (from the "web of science" database) with Brint's institutional data archive that describes numerous attributes of institutions.
Nested spatial hierarchies of places
The central-place hierarchy in the western United States, 1997: This study examines data from the economic census of 1997 on the presence or absence of organizations in each of eleven major economic sectors (e.g. manufacturing, retail trade, accommodations and restaurants) in census economic places in the Mountain and Pacific regions. The joint distribution of sectoral functions across places is examined for "nestedness." Nestedness in a notion from bio-ecology that supposes that niches (analogous to economic places) can be arrayed in a nested hierarchy according the species present (analogous to sectoral functions). Perfect "nesting" in the joint distribution of human communities and the economic functions they support would support predictions of "central-place" theory from human ecology. The empirical reality looks to be somewhat more complex and variable.
The hierarchy of human communities and organizational communities in New Mexico: This study uses the same general human ecological theory and "nestedness" methodology from bio-ecology as the study above. But, in this study we examine a much wider range of human settlements (ranging from small "census designated places" that may have no organizations beyond the household level; to large metropolitan areas). We also capture the organizational ecology of these places in much more detail by using business directory information to classify organizations into some 300 4-digit NAICS functions. While a clearly nested central place hierarchy emerges, there is considerable complexity -- and many deviations from a simple pattern. Some of the co-variates of community ranking (e.g. size, organizational density, transport network embeddedness) may be identified, and some of the variations on the mixes or organizational communities commonly present in settlements may be identified.
Nested hierarchy in human and organizational communities: Building on the two projects described above, a wider study has been proposed to N.S.F. If funded, this project would expand the New Mexico detailed study to include a number of states that are relatively more concentrated in agriculture (e.g. North Dakota), manufacturing (e.g. Indiana) and services. The basic theories of nested hierarchy can be tested across a much wider range of populations with the proposed data; the community ecologies of organizational populations can also be examined with a large (14 million organizations) database.
Multi-scale modeling of spatial diffusion
The diffusion of genetically-modified corn: Norman Ellstrand and several other principle investigators at U.C.R. and U.C.S.B. have received a $1.5 million, multi-year N.S.F. Biocomplexity grant to develop formal models to assist in understanding the diffusion of genetically-modified plants in the world-system. My work is as one of a number of principle scientists on two teams that will develop the theory and a multi-scale simulation model of the non-human and human processes driving diffusion. Using a multi-level cellular automaton approach, we will attempt to integrate processes of diffusion at very small scales (meters) with those at larger spatial scales -- including the world-system. At micro levels, diffusion is driven primarily by physical processes (wind dispersion, competition from native species, etc.); at the meso-level, community social networks and human agents enter the process; at the macro-level, world trading patterns, the activities of multi-national corporations and governmental bodies become critical determinants of the spread (or not) of modified genes.
Multi-scale pollution diffusion processes in the "Inland Empire" of southern California: Working with a number of U.C.R. faculty, primarily from mechanical engineering and from environmental engineering, a proposal for an IGERT training program has been developed. We envision a multi-year program that will integrate teaching and research in the areas of the diffusion of pollution (automotive emissions, grass-fire, water pollution) with social-science theory. At the core of the proposal is adding social science insights to engineering research, and training social scientists in engineering approaches. Central to to proposed program is the integration of dynamic systems modeling of human processes with physical systems modeling, and connecting both to geographical information systems to understand pollution-diffusion dynamics across small (meters) to fairly large (kilometers) scales in the inland southern California region.
Innovation in organizations
Innovating and Implementing: Diffusion of new practices in the Israeli Kibbutz: For several years, work with Raymond Russell and Shlomo Getz has been focusing on the process of adoption of about 50 new organizational processes by Israel's approximately 240 non-religious Kibbutzim over the period from the early 1990s to the early 2000s. Over this period, many of the traditional institutional features of the Kibbutz have been challenged, and in large numbers of cases, replaced by new practices more in line with those of profit-making economic enterprises. We are seeking to understand the effects of institutional pressures, organizational performance pressures, and other factors on the likelihood of consideration and adoption.
Evolution of degeneration of cooperative enterprises: The research underlying the paper described above will also form much of the empirical basis of two or more chapters of a monograph by Rusty Russell on the evolution of the Kibbutz. Russell is one of the world's leading experts on cooperative enterprises, and the study of the Kibbutz is part of a larger agenda of attempting to understand the conditions under which employee ownership and management of economic production is viable, and when it is likely to fail or degenerate into more conventional capitalist enterprise forms. The book chapters will offer a more in-depth statistical analysis using pooled cross-section and time-series models; and, the chapters will offer an in-depth historical and qualitative understanding to the recent crises and change in the Israeli kibbutz.
Social Networks
Version 2.0 of Introduction to Social Network Methods: A free on-line and downloadable textbook that introduces the basic ideas of social network methods was published a few years ago. The text also serves as an introduction to the use and interpretation of many of the algorithms contained in the UCINET software package. With Mark Riddle, the text is being revised and updated to include new sections on graphing (using WebDraw) networks, data handling, and some new features that have appeared in UCINET since version 1.0. Version 2.0 will be made available for free use and download.
Applying graph theory to causal theories/models: A number of useful theories in sociology involve considerable numbers of constructs connected by propositions of covariation or determination that can be represented as simple or directed graphs. With Hiroko Inoue, basic social network concepts (e.g. centrality, structural equivalence) are applied to graph representations of theories involving 20-30 variables, with concepts as nodes, and relations of determination as relations. The same methods that are commonly used to examine sociometric data yield some interesting summary information about the structures of causal arguments that may not have been apparent from simple inspection.
Political democracy
Institutionalizing democracy in Latin America: During the 1990s, many nations in Latin America made transitions from more authoritarian to more liberal-democratic political systems. In this research, we (Shaun Bowler and John Hiskey of the Political Science department are co-authors) are seeking to understand variability in support for political democracy across a general population samples of persons, nested within a number of nations. The work uses attitudinal and demographic information from surveys in a number of nations, along with institutional features of the government and history of the nations in a two-level ordered logistic regression model to predict individual and national differences in support for democratic institutions.
Financial support of ballot initiatives in California: Shaun Bowler has prepared a database of contributors in support and opposed to the numerous initiatives in California, using information reported under campaign contributions laws. We will seek to cluster contributors in terms of similarity in the profiles of the initiatives that they support or oppose; simultaneously, we will seek to cluster the initiatives in terms of similarity in the profiles of those who support and oppose them. The two-mode clustering approach ("correspondence" analysis) aims for the creation of typologies of both supporters and initiatives that are based on the relational character of the data, rather than the attributes of either the actors or the initiatives -- which can then be used as explanatory covariates to predict patterns of support.
Systems-dynamics simulation modeling
Durkheim's ecological theory of differentiation: Jon Turner has provided a number of very useful formalizations of macro-level theories from classical writers. Turner moves the formalization of these classical theories from the author's text to causal diagrams. My goal is to take the causal diagrams and turn them into non-linear, simultaneous differential equation systems that can be simulated to explore the dynamics implied by the theories. This particular piece of work is, in many ways, a continuation of my earlier efforts of this type that seek to contribute to better critical understanding of some widely used sociological theories. In the current case, though, I'm experimenting with developing simulations that readers may themselves experiment with as learning modules to support a more active style of learning about sociological theory.
Spatial grid systems models: The systems-dynamics tradition of simulation modeling tends to focus on whole closed systems (often with internally differentiated sub-systems). The newer agent-based and cellular automata simulation approaches tend to use large numbers of coupled sub-systems, where each sub-system (actor) is, itself, very simple, but embedded in an environment of other agents. My work in this area is seeking to create models of multiple interacting actors, where each actor is itself a complex non-linear system. For example, we may model the rates of birth and death in one community or society as a non-linear system; however, it is more useful to model the demographic systems of multiple populations that are connected by migratory, trade, and information flows among them. Models of spatial interaction, and models of interaction in social networks form two main types of "sub-strates" on which complex actors interact.