Social Explorer Logo
Documentation: Census 1960 Tracts Only Set
you are here: choose a survey survey document chapter
Publisher: U.S. Census Bureau
Document: National Data Use and Access Laboratories (DUALabs)
Social Explorer; U.S. Census Bureau; Census Tract-Level Data, 1960 [Computer file]. ICPSR version. Washington, DC: U.S. Dept. of Commerce, Bureau of the Census [producer], 1971. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2007-12-13. doi:10.3886/ICPSR07552
National Data Use and Access Laboratories (DUALabs)
I. Technical Conventions for the 1960 Census Tract Summary Tape
A. Technical Characteristics of Magnetic Tape Recording
Characteristic Type of Tape
  IBM-Compatible 7-Track IBM-Compatible 9-Track
1. Computer on which users tape copy created IBM 360/65 IBM 360/65
2. Tape width, reel size, and tape length .5 inch, 10.5 inch diameter; maximum 2,400-foot length .5 inch, 10.5 inch diameter; maximum 2,400-foot length
3. Recording mode NRZI NRZI
4. Language BCD EBCDIC
5. Density 556 or 800 CPI 800 or 1600 CPI
6. Parity Even Odd
7. Recorded tape representation One 6-bit character per tape frame 1 8-bit character (byte) per tape frame
8. Logical record size 10,800 characters or 6 physical records 10,800 characters or 6 physical records
9. Blocking factor or physical record size 1,800 characters 1,800 characters
10. Interblock spacing .75 inch .6 inch (nominal)
11. Volumes    
12. Labels Standard IBM 0S labels or unlabeled Standard IBM 0S labels or unlabeled

B. Conventions for Data Records
Block (Physical Record) Size
Block-size for this summary file follows 1970 census summary tape conventions and has been fixed for the entire file at one block-size which is an even multiple of 120 characters. No data item is ever split across a 120-character sub-block or a physical record although the practice of splitting tables across these units often occurs.
The block-size for the 1960 tract file is 1,800 characters.

Logical Record Size
A logical record consists of the geographic identification and all data fields for one summary matrix comprising a single geographic area.
The logical record size for this file is 10,800 characters of 6 physical records. Each logical record consists of 207 characters of geographic identification and padding followed by 1,177 nine-character fields containing population and housing summary data and padding.

Total File Size
Approximately 23,000 logical records for the United States.

Sequence of Files
Knowledge of the pattern of data organization, often referred to as the sort sequence, is necessary to locate and perhaps further aggregate census data on the summary tapes. The file described in this document consists of records for 1960 census tracts. The file is in sort (ascending sequence) by tract (prefix, basic, and suffix) within county within state. Records for split tracts (i.e., tracts that have been divided into two or more portions because they cross corporate or other boundaries) occur together in the file within their respective counties. A diagram illustrating this sequence is shown on the following page.

Diagram #1: Schematic Representation of 1960 Summary File Sort Sequence
Unit Summarized Hierarchical Structure
1960 Census Tract
Prefix Basic Suffix County 1 State A
AAA 10 0AA    
AAA 10 0AB
AAA 11 0
AAB 5 0
AAB 6 0
AAB 7 0
AAB 8 0
0 1 00A
0 1 00B
0 2 000 (part)
0 2 000 (part)
0 2 000 (part)
0 4 0
BBB 12 0 County 2  
BBB 13 0    
CCC 15 0
CCC 16 0
0 3 AAA County3  
0 3 BBB    

Conventions for Fields
Geographic Identification
The geographic codes which identify a logical record are contained in the first 207 characters of the first block of the logical record. (Field sizes for these codes are variable.) This identification is not repeated in the subsequent blocks of the record.

The designation padding has been applied broadly are refers both to characters that are unused or which contain information that is not relevant at the census tract level of summarization.

Geographic fields so identified may be blank, contain zeros, or numeric or alphanumeric representations. In all instances these fields may be ignored.

Data Fields
Data fields contain integers only and are a uniform size of 9 charcters each. Padding convention for a data fields are as described below.

Suppression Rules
Complete Count Data (Tables 1-16)
1. If total population is less than 5; tables 1, 8, and 11 through 16 will be suppressed and contain zeros.
2. If total housing units is less than 5; tables 2 through 7 will be suppressed and contain zeros.
3. If total housing units is 5 or more but
a) if total owner-occupied units reporting value is less than 5; then table 7 will be suppressed and contain zeros.
b) if total renter-occupied units reporting rent is less than 5; table 10 will be suppressed and contain zeros.

Sample Data (Tables 17-63)
If Nonwhite (Spanish surname/Puerto Rican) population is less than 400; then sample detail for these universes will not be shown and all pertinent fields in tables 17 through 63 will be filled with zeros. No other suppression rules apply to sample tables 17 through 63.

C. Label Conventions
Header and trailer labels on DUALabs 1960 Census Tract Summary Tape copies are standard OS labels.

Labels are classified into two required types: volume labels, where Volume coincides with a physical tape reel, and File labels, where File has the normal meaning of a collection of related data records, more than or less than one reel.
A Label is always an 60-character physical record.
In the conventions for the 1960 tapes, labels are restricted to the following:

1. One Volume Header (VOL1) at the beginning of each physical tape reel.
2. Two Volume Headers (BOR1,2) at the beginning of each file and at the beginning of each tape reel (following the Volume Header) within file.
3. One Volume Trailer (EOV1 ) if end of tape does not coincide with end of file.
4. Two File Trailers (EOF1,2) at end of file. If end of tape and end of file coincide, use only File Trailers.

The rules for tape mark are:
1. One tape mark separates labels and data, occurring after all labels which precede data, and before all labels which follow data.
2. Two tape marks appear at the end of tape after EOV or EOF records.
3.If a file ends within a volume and another file is written on that volume, then one tape mark both precedes and follows the end of file labels; the next file headers appear and are separated from their data by a tape mark.

Label Content
As noted, the labels on DUALabs tape copies are standard IRM OS labels. The File Identifier (Data Set Name) which appears beginning in character 5 of the First File Header is as follows:
1. State-separate and -merged compressed copies:
DSN=STAAC60, where AA=1960 State code.
2. State-separate and -merged straight copies:
DSN=STAAX60S, where AA=1960 State code.
Note that all state-merged versions retain each States data as a separate file.

Unlabeled Tape Copies
Unlabeled copies simply contain data followed by a tape mark. In the case of State-merged files, where each States data constitutes a separate file and there is more than one file on a reel, a tape mark separates each States data.

D. Character Set
The character set is restricted to those characters which are generally accepted and which are represented by a unique code in any particular coded decimal language. Analysis of the IBM binary code decimal tape code recognized by 8 computer manufacturers on approximately 15 different computer types reveals a consensus on the 10 decimal digits, the 26 alphabetic symbols, 6 special characters, and the blank or space. Also, each of these systems recognizes the tape mark.

The character conversion chart on the following page contains the 43 character graphics and their coded representation in BCED and EBCDIC. The 6-bit character code is shown in octal representation; the 8-bit character code is shown in hexadecimal.

The six special characters which may be used in data intended for interchange are indicated below. They also appear on the character conversion chart shown on the next page.

Character Graphic BDC* (Octal)* EBCDIC* (Hexadecimal)*
- (minus) 40 60
, (comma) 33 6B
. (period) 73 4B
$ (dollar sign) 53 5B
* (asterisk) 54 5C
/ (slash or virgule) 21 61
Note: *See footnotes to following character conversion chart

Character Graphic BCD1 (Octal)2 EBCDIC3 (Hexadecimal)4
Blank (space) 20 40
- (minus) 40 60
0 12 F0
1 1 F1
2 2 F2
3 3 F3
4 4 F4
5 5 F5
6 6 F6
7 7 F7
8 10 F8
9 11 F9
, (comma) 33 6B
. (period) 73 4B
A 61 C1
B 62 C2
C 63 C3
D 64 C4
E 65 C5
F 66 C6
G 67 C7
H 70 C8
I 71 C9
J 41 D1
K 42 D2
L 43 D3
M 44 D4
N 45 D5
O 46 D6
P 47 D7
Q 50 D8
R 51 D9
S 22 E2
T 23 E3
U 24 E4
V 25 E5
W 26 E6
X 27 E7
Y 30 E8
Z 31 E9
$ (dollar) 53 5B
* (asterisk) 54 5C
/ (slash or virgule) 21 61

1This is a 6-bit code known as binary coded decimal, used originally in earlier model IBM computers and peripheral devices.
2This is an 8-bit code, known as extended binary coded decimal interchange code, which originated with IBM Systems 360.
3In the octal representation of each 6-bit code, the three high order bits are represented in the most significant digit of the octal code, and the three low order bits are represented in the least significant digit of the octal code. The range of each octal digit is 0-7.
4In the hexadecimal representation of the 8-bit code, the four most significant bits are represented in the most significant digit of the octal code, and the four least significant bits are represented in the least significant position on the hexadecimal code. The value range of each hexadecimal character is 0-15 where the values 10, 11 --- 15 are defined by the letters A, B, --- F, respectively.
5Octal 20 for the blank, in BCD, appears only on tape. Internally this code is 00 and difference between the internal and external representations is handled automatically by the hardware. Where BCD is generated internally in the UNIVAC 1107 or 1108, and written to tape without the hardware translation, the octal 20 must be created in core.