Using Data to Drive Jewish Communal Policy

As important as it is for Jewish communal leaders to use data to help shape Jewish communal policy, “counting” alone is not sufficient. Just as our communal problems are complex, research responses must be more than decontextualized numbers.

Photo by dirkcuys/Flickr
Photo by dirkcuys/Flickr

By Leonard Saxe

A dictum attributed to Albert Einstein quips, “Not everything that can be counted counts, and not everything that counts can be counted.” It’s a pithy statement of humility about collecting and using data.

As important as it is for Jewish communal leaders to use data to help shape Jewish communal policy, “counting” alone is not sufficient. Just as our communal problems are complex, research responses must be more than decontextualized numbers.

For nearly two decades, my colleagues and I at the Cohen Center and the Steinhardt Institute have tried to conduct contextualized, policy-relevant, research. We apply cutting-edge methodological techniques to understand the contemporary Jewish community and provide information that can inform communal decision-making. Although the use of data by the Jewish community often emphasizes counting – how many Jews, how many intermarriages, how many children go to day schools or summer camps – our focus explores the determinants of Jewish engagement and trends over time.

We have tackled issues including the relationship of millennial Jews to Israel, Jewish life on campus, summer camps, intermarriage, and demographic predictors of engagement with Jewish life.

I would like to think that our work on projects such as Birthright Israel, day schools, teen engagement, and the socio-demography of American Jewry has helped to deepen understanding of these issues, create more effective programs, and guide resource allocation.

Comparative Analyses

Analyses that focus on comparisons, rather than measurements of characteristics or outcomes, are central to almost all of the research we do. Although for some, measurement has become synonymous with systematic research, we believe it is only part of what is needed to develop useful policy-relevant research findings. No matter how well one measures a phenomenon, absent comparison with a baseline or other groups, the findings are not very useful and, in some cases, can be misleading. Measurement is also far more complex than most think – often not as much about the questions per se, but how different measures are combined.

Our research on the impact of Birthright Israel illustrates the importance of comparison and multivariate measurement. We collect data on hundreds of indicators that describe characteristics of applicants and the effects of participating in Birthright Israel’s ten-day Israel education programs.

But what makes the research useful is that we are able to compare participants with similar non-participants. Just knowing the number of Birthright Israel participants who are later involved in the Jewish community, advocate for Israel, or marry other Jews is not useful, absent comparative information for similar others who did not have the Birthright Israel experience. We need to know whether the high rate of Jewish involvement by participants is an artifact of only the most engaged individuals having applied to the program.

Predictive Modeling

In addition to comparison groups and the ability to compare to a baseline, the most useful policy-oriented research also includes predictive modeling. Statistical models are created from multiple indicators of individual characteristics and experiences and then used to identify the conditions associated with Jewish identity and engagement outcomes. For example, in recent research on intermarriage, we compare groups (e.g., children of inmarriage and intermarriage) and models of past experiences (e.g., Hebrew school as a child) to understand how Jewish education and experiences during college affect later Jewish connections.


For much of the last 50 years, the Jewish community has made its largest research investment with socio-demographic studies. A central goal of these studies is to estimate the size of the Jewish population, but their most important function is describing the characteristics of those who are Jewish. Understanding the attitudes and behavior of contemporary Jews enables the community to better address their needs. Note that counts of the Jewish population often have little meaning without comparative data from earlier periods. Such comparative data is often problematic due to changes in sampling methods over time. This challenge makes it even more important to focus primarily on understanding the characteristics of contemporary Jews and predictors of engagement.

Although there is much to learn from the past, the best research will take into account rapidly developing changes at the individual and communal levels. Just as Jewish identity formation is not limited to one particular developmental stage but continues as a life-long project, our research agenda must be similarly flexible. There is an urgent need to understand the ways in which we can influence the future. For that, Einstein’s comment about choosing carefully what should and should not be counted remains important advice. Rather than deter us from developing and measuring indicators of Jewish life, his observation should serve as a reminder that counting and understanding are inextricably linked.

Leonard Saxe is Klutznick Professor of Contemporary Jewish Studies at Brandeis University and Director of the Cohen Center for Modern Jewish Studies and the Steinhardt Social Research Institute.

This article was published in the Hornstein Program’s November 2016 issue of Impact; reprinted with permission. All rights reserved.