Greater Boston 2005 Community Study

Sponsor(s): Combined Jewish Philanthropies of Greater Boston (CJP)

Principal Investigator(s): Leonard Saxe, Benjamin Phillips, Charles Kadushin, Graham Wright, Daniel Parmer

Study Dates: Interviewing: 2005. Preliminary Findings, Saxe et al. for CJP, 2006.

Katherine N. Gan, Harvard University; Patty Jacobsen (CJP), Gil Preuss (CJP) and Barry Schrage (CJP) issued report on Intermarried Families and Their Children, March 2008.

Population Estimates:

The Jewish Community of Boston (CJP service area) in 2005 included 208,500 Jewish persons (adults and children) living in 105,500 Jewish households; an additional 57,000 non-Jewish persons lived in these households for a total of 265,500 people in these households.  The report also estimates that another 2,000 Jews reside in institutional settings

In 1995, the North Shore of Boston was also included; the estimated number of Jewish persons was 18,400 (DataBank analysis of report which in 1995 included non-Jews living in Jewish households as part of the "Jewish population") who resided along with another 1,600 non-Jews (total of all people  =  20,000) in 8,300 Jewish households.

While the 2005 Study did not include the North Shore, as of July 1, 2013, the Jewish Federation of the North Shore has merged with Combined Jewish Philanthropies.  Using the 1995 North shore estimates, therefore, the total number of Jews in the combined CJP/North Shore Jewish community is estimated to be 226,900 Jews and 58,600 non-Jews (all people =  285,500) living in 113,800 Jewish households.

Updated information is highly likely to emerge with the planned Greater Boston 2015 Jewish community study, which would mark half-a-century of decade-spaced Jewish community studies since the first study in 1965.

Key Findings:

2005 Findings:

The 2005 Study focused only on the CJP service area in Greater Boston, as opposed to earlier studies which included many communities in the North Shore of Boston.

The 2005 report noted that the Jewish population of the CJP Greater Boston catchment area (not including the North Shore) increased significantly since 1995; total number of Jewish households increased from 86,000 to 105,500; total number of Jewish persons in these households increased from 177,000 to 208,500; total number of people in Jewish households increased from 209,500 to 265,500. The report also estimates that another 2,000 Jews reside in institutional facilities.

Jewish individuals are 7.2% of the Boston area population; the total number of people in Jewish households represent 9.1% of total CJP Boston area general population.

The Boston Jewish community is geographically dispersed; half of the Jewish population resided within Route 128 and half outside of it; the historically central areas of Newton, Brookline and Brighton continue to be home to the largest number of Jewish persons;

The Jewish community of Boston has a demographic bulge among baby boomers, ages 50-59.

While the community is relatively affluent, 6% of households report annual incomes under $15,000, and another 9% report incomes under $35,000; 9% of Jewish households are under 200% of the U.S. government measures;

Intermarriage has continued to increase in the Greater Boston area; number of intermarried Jewish households increased from 18,000 in 1995 to 35,500 in 2005; recent intermarriage rate is approximately 37% of married couples; the number of children in intermarried Jewish households is approaching the number of children in in-married households; however, the 2005 report indicates that 60% of children in intermarried households are being raised as Jews, 4% in Judaism and another religion, 8% in another religion and 28% without a religion;

Levels of Jewish connections are moderately high; 49% of Jewish adults report congregation affiliation. In terms of Jewish ritual practice: 79% of Jewish adults always/usually light Chanukah candles, 79% always/usually attend a Passover seder, 25% always/usually light Shabbat candles, and 10% strictly observe dietary laws; Jewish education is practically universal for Jewish children ages 9-13; Jewish education for Jewish children in intermarried Jewish households is almost universal also, but declines after celebrating bar or bat mitzvah more rapidly than among children in i-married Jewish households;

Israel travel was reported by almost half of all Jewish adults, but by relatively few in the five years preceding the survey; Israel occupies an "important but not central place in the minds" of Boston Jews; 38% feel a lot connected to Israel, another 38% feel somewhat attached;


Jewish households in the Boston CJP catchment area; maps included in the Research Methodology Report.

Again, the area studied in 2005 does not include some parts of the Greater Boston Jewish community study area used in 1995 and previous studies, largely the North Shore of Boston.

Sample Size: Sample size reported in Preliminary Findings (November 2006) was 1,766: 401 from RDD (random digit dialing frame) and 1,365 from List frames.

Please see next section "Sample Notes" re: sample size.

Sample Notes:

Research Methodology report (February 2007) indicates that in addition to the 401 RDD interviews, another 33 multiplicity frame interviews were completed; in addition to the 1,365 List-frame interviews, an additional Russian and web-based surveys were completed for a List-web-Russian total of 1,388.

Household data files in STATA and/or SPSS indicate 1,826 weighted interviews: 434 RDD and 1,392 from the List.

Sampling design described utilized two independent, complementary sampling frames: (1) a Jewish Community List compiled from 84 Jewish organizations, including CJP, and (2) an RDD frame from which all numbers on the List frame had been removed, and which was stratified by Jewish density modeled after distributions on the List frame.

Both the Appendix to the Preliminary (Summary) Report and the massive Methodology Report provide methodological details, including a discussion of the elimination of some areas from the sampling design, since Jewish incidence in these areas would have been exceptionally low and interviewing costs exceptionally high. Population estimates for these un-enumerated areas were included in the overall Jewish household and population estimates for the study, but not in the data file.

Response rate for study reported as 40% (AAPOR RR2); 50% in List frame, 34% in RDD frame. Sample disposition included in the Methodology Report.

2005 Preliminary Report authors recalculated 1995 data for comparisons in the 2005 Report, so some reporting discrepancies between the published 1995 report and the 1995 data reported in the 2005 reports may exist. Readers and researchers reviewing the 1995 and earlier studies of the Boston Jewish community should focus on the geographic areas studied, since the 2005 Preliminary Report focuses on the CJP catchment area only.

In addition, the 2005 Boston Jewish Community Study reported the number of Jewish persons and non-Jewish persons separately, unlike the 1995 Report (and several prior studies) in which Jewish persons and non-Jewish persons were combined as either the "Jewish population" or as "Jews."

Study Notes: The Steinhardt Institute researchers believe that Stata (see indented note below for further details) should be used for the analysis of the Boston 2005 data - and that the Stata ".do" files should be consulted before data analysis. They have graciously agreed, however, that SPSS portable versions of the data files could also be made available.

The Brandeis researchers note that Stata (when used with svy commands) adjusts for the special sampling design used in the study; SPSS users need to be cautious in interpreting data and assuming statistical significance of results, since the Brandeis researchers note that SPSS users will underestimate sampling design error, unless using the SPSS Complex Samples module.

    "The 2005 Boston Jewish Community Study used complex sample design. Consequently, analyses that do not account for sample design will return incorrect standard errors and tests of statistical significance. Calculated confidence intervals will also be invalid. Accordingly, it is necessary to use a software package capable of adjusting for sample design (e.g. Stata, SPSS with Complex Samples module, WesVar, or SUDDAN) and to use those capabilities. The included Stata .do files contain svyset commands that correctly set up the data for analysis. Information on sample design is available from the methodological report, including the procedures used to compress weights to adjust for extreme variables."

The Preliminary Report utilized a confidence interval range of 80%. The data files were transmitted to the Data Bank on behalf of CJP by Katherine Gan, and the Data Bank was requested to include the following note: "As the data for the 2005 Boston Jewish Community Study were to be used for policy and planning purposes, it was decided that the conventional 95% confidence intervals would have unduly prioritized certainty (avoiding Type I errors in statistical terms) over utility (avoiding Type II errors), as greater levels of statistical confidence are accompanied by a wider interval in which the true value of a statistic could be found. The 80% percent confidence interval was felt to be an appropriate balance between these competing considerations. The breadth of the confidence intervals was increased by the design effect of the study--the extent to which the complex sampling scheme used increased levels of sampling error compared to a simple random sample."

DATA FILES. Both household and individual data files exist for Boston 2005 in both Stata and SPSS format. In addition, the Stata ".do" files are included with the Stata files since they describe all analyses used in the Preliminary Report. An Appendix in the Intermarriage Report also includes Stata commands and comments. SPSS users are advised to read these files as they would read SPSS syntax.

Survey weights used are described in the Methodology Report, although the names may have been simplified for the final data files. Using "mainhhwt" in Stata there are 1,826 weighted interviews: 434 non-List and 1,392 List interviews. The mean weight is 64.19809, so the weighted total is 117,825 - the number as SPSS users should find using "mainhhwt" in the household file.

The data files contain all cases in the sample, including non-Jewish HH and non-contacts. The only weighted cases are completed interviews, so researchers wanting to only use a completed interview subset might want to subset where "mainhhwt" is greater than zero; 1,826 cases should emerge.

A compressed version of the household weights exists in both files, and is discussed at length in the Research Methodology report. Essentially, the weight adjusts for extremely high weights assigned after screening to Jewish households (RDD sample) in very low incidence areas where few interviews are completed. In Stata, this variable is "mainhhwtcomp"; in SPSS, it is labeled "mainhhw0." Other compressed weights exist in the data files; the chapter on weighting should be consulted.

Language: English


Survey Reports

» Intermarriage Report

» Summary Report

Slide Sets

» Overview Slide Set

Documentation, Questionnaires and Frequencies

» Methodology Report

» Questionnaire

Data Files and Data Definitions

» Zipped Package includes SPSS & STATA Data Files

Bookmark and Share