topbrandingbar
corner.gif
Government Information Departments and Agencies NJ Business Portal MY New Jersey NJ people NJ Home Page

CHS Home Page CHS Data CHS Reports List of CHS tables and reports CHS topics from A-Z CHS Links CHS Frequently Asked Questions Search the CHS pages

MUNICIPALITY-LEVEL OVERCOUNTS AND UNDERCOUNTS IN
NEW JERSEY VITAL EVENTS DATA

UPDATED: APRIL 2006

New Jersey has a long standing problem with the assignment of municipality of residence on vital records. This is due primarily to the differences between postal boundaries and municipal boundaries. Throughout most of New Jersey , the town/city where mail is delivered does not necessarily have the same name as the municipality in which the person resides. When stating one’s residence address, most people give their mailing address. This presents difficulties when analyses are performed at the municipality level.

Since the street address of the mother is entered into the electronic files of birth records, the Center for Health Statistics (CHS) has been able to correct improperly assigned residence data through the use of geocoding software. However, street addresses are not available on the electronic files of death records before 2004; therefore geocoding correction is not possible on pre-2004 death data. CHS has a policy of not releasing death data for municipalities which are known to have poor quality residence assignment.

An updated analysis has been performed on five years of birth data to reassess the degree to which each municipality’s events are under- or overcounted. By comparing the original residence code to the corrected residence code on the birth file, it is possible to determine the net overcount or undercount of events that each municipality experiences without geocoding correction. The shift in residence on birth records is a reasonable proxy for what occurs with death records. Therefore, the net over- and undercounts based on birth data have been applied to death data.

The table below gives a listing of all municipalities along with the estimated net percentage over/undercount without geocoding correction. These percentages can be applied to death data to give a very rough estimate of how much the reported data under- or overcounts the true number of deaths in a given municipality. For example, the figure of 82.8 percent given for Absecon means that results based on uncorrected data represent a 82.8 percent overcount of the true number of events occurring to residents of Absecon. Alternatively, the –1.0 percent for Brigantine means that there is a 1.0 percent undercount of events occurring to Brigantine residents.

CHS has adopted a policy of not releasing death data for any municipality which has more than a 10 percent overcount or undercount. By this criterion, 341 of New Jersey ’s 566 municipalities have releasable death data. Deaths to residents of these 341 municipalities comprise 78% of all deaths of New Jersey residents in 1999-2003. All municipality-level death data generated by CHS are based on the assignment of residence as stated on the death certificate. No adjustments for over- or undercounts have been made to the published data.

UPDATED BASED ON 1999-2003 BIRTH DATA
Table of Overcounts and Undercounts by Municipality
Over- and undercounts of New Jersey resident births by municipality.

Municipality-Level Births and Deaths by Selected Characteristics: Tabulation of births by age, race/ethnicity, and marital status of the mother; sex and birth weight of the child; and trimester in which prenatal care began by mother's municipality of residence. Tabulation of deaths by age, race/ethnicity, sex, and cause of death by municipality of residence.

 
State Privacy Notice legal statement DOH Feedback Page New Jersey Home

 
department: njdhss home | index by topic | programs/services
statewide: njhome | my new jersey | people | business | government | departments | search

Copyright © State of New Jersey, 1996-2004
Department of Health
P. O. Box 360
Trenton, NJ 08625-0360

Last Updated: