BERKELEY
  MORTALITY
    DATABASE

Introduction

Overview

User Agreement

Data by Country:
   France
   Japan
   Sweden
   United States



home
   
e-mail

Overview of the BMD


This database contains life tables for national populations and, whenever available, the raw data used in constructing these tables. The raw data generally consist of birth and death counts from vital statistics, plus population counts from periodic censuses. Both general documentation and the specific steps followed in constructing life tables are given below. More detailed information, including data sources, can be found by following the links within the sections for each country on the main page.

Raw data and estimations

Whenever possible, the database includes raw data on birth, death, and census counts (i.e., original data without any adjustments or estimations). Raw data are presented here as whole numbers without a decimal point. However, when the final value represents an estimate of any sort, a decimal point and trailing digits are included as well.

The most common adjustment of raw data is to distribute persons of unknown age (in either death or census counts) proportionately across the age range. A less common adjustment is to split aggregate data into finer categories. In all cases, adjustments to raw data (i.e., birth, death, or census counts) are described in the Notes section for each country.

All population or exposure estimates, as well as calculated death rates, depend on various estimations and thus are presented here in decimal format. Life table quantities are rounded to whole numbers or presented in decimal form following standard conventions.

Male / Female / Total

Life tables and all data used in their construction are available for males and females separately or together. In some cases, a single file contains columns labeled "male," "female," and "total." In others, separate files contain data for these three groups.

Age and time intervals

Age and time are categorized into 1-, 5-, and 10-year intervals. The dimensions of a matrix of death rates (or some other quantity) are denoted by 1x1, 5x1, 5x10, etc. In this notation, the first number refers to age intervals, and the second refers to time intervals. For example, 1x10 denotes a configuration with single years of age and 10-year time intervals. By convention, the first five-year age interval separates the first year of life (age 0) from the rest of its 5-year age group (ages 1-4). Thus, a 5x1 configuration contains data for single years of time with the following age intervals: 0, 1-4, 5-9, 10-14, …, 95-99, 100+ (or …, 105-109, 110+).

Steps for computing life tables

  1. Births

    Birth counts are usually available for individual calendar years. For some countries (e.g., Japan), birth counts for selected months and years are also included (corresponding to births that occur before or after the national census date, e.g., October 1).

  2. Deaths
  3. In this database, death counts are provided at the finest level of detail available. Thus, whenever possible, deaths are organized by individual triangles of the Lexis diagram. In other situations, however, death counts are provided only for 1x1 Lexis squares or 5x1 Lexis rectangles.

    A common adjustment is to distribute deaths of unknown age proportionately across the age range. A less common adjustment is to split some aggregated deaths into finer categories. Such "disaggregated" death counts are shown here only if data for adjacent years is available at the finer level of detail (e.g., as occurs for France during certain years at ages 100 and above), thus permitting a uniform presentation. Such adjustments are described in the Notes section for each country.

  4. Census

    Census counts are derived from national tabulations. The data given here contain the maximum level of detail by age available in published sources. The exact reference date (within the year) for a given census differs from country to country and, in some cases, from census to census within a country (see the Notes section for each country). When necessary, persons of unknown age were distributed proportionately into other age groups.

  5. Population estimates

    Estimates of population size on January 1 of each year are derived using a combination of three methods: intercensal survival, extinct generations, and survival ratios. Population estimates above age 90 are derived by the method of extinct generations for all cohorts that are  extinct (defined arbitrarily as attaining age 110 or more by the end of the last year of available death counts). For all ages below 90, population estimates are attained by the method of intercensal survival with an adjustment for migration if migration data exist. The survival ratio method is used for cohorts that are almost extinct (generally defined as being at least age 90 by the end of the observation period). The exact combination of these three methods may differ by country depending on the availability and quality of existing data. See the Notes section for each country for more information.

  6. Exposure estimates

    Estimates of the population exposed to the risk of death (during some interval of age and time) are computed based on the January 1 population estimates. The exposure estimation used here contains a small correction for death counts during the interval.

  7. Death rates

    Death rates are simply the ratio of death counts to population exposure in the relevant interval of age and time.

  8. Life tables

    Death rates are converted to life tables by standard methods.