View in EDS HTML Full Text PDF Full Text

Occupational Percentile Rank: A New Method for Constructing a Socioeconomic Index of Occupational Status

Saved in:
Bibliographic Details
Title: Occupational Percentile Rank: A New Method for Constructing a Socioeconomic Index of Occupational Status
Language: English
Authors: Xi Song (ORCID 0000-0003-3463-045X), Yu Xie (ORCID 0000-0002-3240-8620)
Source: Sociological Methods & Research. 2025 54(2):355-396.
Availability: SAGE Publications. 2455 Teller Road, Thousand Oaks, CA 91320. Tel: 800-818-7243; Tel: 805-499-9774; Fax: 800-583-2665; e-mail: journals@sagepub.com; Web site: https://sagepub.com
Peer Reviewed: Y
Page Count: 42
Publication Date: 2025
Document Type: Journal Articles
Reports - Research
Descriptors: Socioeconomic Status, Occupations, National Surveys, Statistical Data, Social Science Research, Sociology
Assessment and Survey Identifiers: American Community Survey
DOI: 10.1177/00491241231207914
ISSN: 0049-1241
1552-8294
Abstract: In this paper, we propose a method for constructing an occupation-based socioeconomic index that can easily incorporate changes in occupational structure. The resulting index is the occupational percentile rank for a given cohort, based on contemporaneous information pertaining to educational composition and the number of workers at the occupation level. An occupation may experience an increase or decrease in its occupational rank due to changes in relative sizes and educational compositions across occupations. The method is flexible in dealing with changes in occupational and educational measurements over time. Applying the method to U.S. history from the mid-nineteenth century to the present day, we derive the index using IPUMS U.S. Census microdata from 1850 to 2000 and the American Community Surveys (ACSs) from 2001 to 2018. Compared to previous occupational measures, this new measure takes into account occupational status evolvement caused by long-term secular changes in occupational size and educational composition. The resulting percentile rank measure can be easily merged with social surveys and administrative data that include occupational measures based on the U.S. Census occupation codes and crosswalks.
Abstractor: As Provided
Entry Date: 2025
Accession Number: EJ1473582
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwHqbt5tyZNmLVOqiXU8eJNLAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDH6QU7n2TKRXeL0I8AIBEICBm2zNKEKeL5drUzmt-u55x3qoXDL_dfHeB71WW-dXjybKpLdnSUU8w41PlNJGUS-GLxnK9vtDUpLCuMBNNuSReyOTHUWWmZO6A_40vhG4PCiBooMjlh8WsFFLdX5qmpp9u7mn9UsETg-XHijLjk9HGvnNzGTIa9ZGMCu0LT8K8O3U_tmrR5evMFmC5FtwyPnSByQf3O8Fe4i-RYYP
Text:
  Availability: 1
  Value: <anid>AN0184233957;som01may.25;2025Apr07.05:58;v2.2.500</anid> <title id="AN0184233957-1">Occupational Percentile Rank: A New Method for Constructing a Socioeconomic Index of Occupational Status </title> <p>In this paper, we propose a method for constructing an occupation-based socioeconomic index that can easily incorporate changes in occupational structure. The resulting index is the occupational percentile rank for a given cohort, based on contemporaneous information pertaining to educational composition and the number of workers at the occupation level. An occupation may experience an increase or decrease in its occupational rank due to changes in relative sizes and educational compositions across occupations. The method is flexible in dealing with changes in occupational and educational measurements over time. Applying the method to U.S. history from the mid-nineteenth century to the present day, we derive the index using IPUMS U.S. Census microdata from 1850 to 2000 and the American Community Surveys (ACSs) from 2001 to 2018. Compared to previous occupational measures, this new measure takes into account occupational status evolvement caused by long-term secular changes in occupational size and educational composition. The resulting percentile rank measure can be easily merged with social surveys and administrative data that include occupational measures based on the U.S. Census occupation codes and crosswalks.</p> <p>Keywords: occupation; socioeconomic status; percentile rank; continuous measure</p> <hd id="AN0184233957-2">Introduction</hd> <p>One of the key features of human society is the vast variability in social attributes. Not only is any social attribute highly heterogeneous at the individual level, but an individual's social attributes are also multidimensional in nature, manifested in education, occupation, income, wealth, personal reputation, community, and family background, among many other characteristics. Hence, developing quantitative measurements of an individual's social position is very difficult, as it is impractical to incorporate detailed measures of all these attributes. Many early studies in search of socioeconomic status indicators suggested occupation as a simple—yet arguably the single most important—indicator ([<reflink idref="bib5" id="ref1">5</reflink>]; [<reflink idref="bib14" id="ref2">14</reflink>]; [<reflink idref="bib33" id="ref3">33</reflink>]; [<reflink idref="bib110" id="ref4">110</reflink>]), a measure that is highly associated with one's ability, characteristics, and training, and from which others can infer one's social prestige ([<reflink idref="bib43" id="ref5">43</reflink>]; [<reflink idref="bib69" id="ref6">69</reflink>]; [<reflink idref="bib81" id="ref7">81</reflink>]). Compared to income and wealth, occupation is publicly known to others ([<reflink idref="bib41" id="ref8">41</reflink>]; [<reflink idref="bib53" id="ref9">53</reflink>]) and is often the only variable consistently collected in historical registers and records and widely available in social surveys.</p> <p>For more than a century, occupational measures have been widely used in both government statistics and social science research. Yet making good use of occupational data in sociological studies is fraught with methodological challenges. One difficulty is the assurance of measurement comparability across studies, populations, and time. Broadly speaking, the development of occupational measures has evolved along two major lines: (<reflink idref="bib1" id="ref10">1</reflink>) The class approach that groups occupations into categories (e.g., [<reflink idref="bib29" id="ref11">29</reflink>]; [<reflink idref="bib31" id="ref12">31</reflink>]; [<reflink idref="bib40" id="ref13">40</reflink>]; [<reflink idref="bib49" id="ref14">49</reflink>]; [<reflink idref="bib68" id="ref15">68</reflink>]; [<reflink idref="bib72" id="ref16">72</reflink>]; [<reflink idref="bib89" id="ref17">89</reflink>]; [<reflink idref="bib112" id="ref18">112</reflink>]; [<reflink idref="bib116" id="ref19">116</reflink>]); and (<reflink idref="bib2" id="ref20">2</reflink>) the gradational approach that represents occupations with a unidimensional, continuous scale based on occupational prestige or socioeconomic scores (e.g., [<reflink idref="bib21" id="ref21">21</reflink>]; [<reflink idref="bib24" id="ref22">24</reflink>]; [<reflink idref="bib37" id="ref23">37</reflink>]; [<reflink idref="bib38" id="ref24">38</reflink>]; [<reflink idref="bib51" id="ref25">51</reflink>]; [<reflink idref="bib52" id="ref26">52</reflink>]; [<reflink idref="bib56" id="ref27">56</reflink>]; [<reflink idref="bib73" id="ref28">73</reflink>]; [<reflink idref="bib76" id="ref29">76</reflink>]; [<reflink idref="bib93" id="ref30">93</reflink>]; [<reflink idref="bib97" id="ref31">97</reflink>]). Both approaches are widely accepted, tested, and debated for their strengths and limitations (see a review in, e.g., [<reflink idref="bib68" id="ref32">68</reflink>]).</p> <p>Yet most research thus far has focused almost exclusively on modern, industrialized societies. Very few researchers have developed occupational measures for past populations or transitional societies before or during industrialization. One exception is [<reflink idref="bib96" id="ref33">96</reflink>]), who matched an incomplete list of occupations observed in the U.K., U.S., Italy, and Nepal from the fifteenth to nineteenth centuries to the 1968 International Standard Classification of Occupations (ISCO). More recently, van Leeuwen, Maas, and their collaborators devised the Historical International Standard Classification of Occupations (HISCO)[<reflink idref="bib6" id="ref34">6</reflink>] and the Historical International Standard Class Scheme (HISCLASS) for occupations in preindustrial, agrarian societies in Western Europe from the eighteenth to the twentieth centuries ([<reflink idref="bib106" id="ref35">106</reflink>]; [<reflink idref="bib107" id="ref36">107</reflink>]).</p> <p>In this paper, we propose a new occupation-based measure of socioeconomic status. We align occupations on one dimension and summarize the socioeconomic status of an occupation with a single parameter based on percentile ranks. Each occupation's rank varies by birth cohort, depending on the number of occupational incumbents and their educational composition. This measure is closely related to Hauser and Warren's ([<reflink idref="bib52" id="ref37">52</reflink>]) occupational education score, but it is cohort-specific, with changing scores affected by the relative sizes and educational standings of occupations.</p> <p>The rest of the paper proceeds as follows. The "Occupational Measures: Historical Perspective" section provides a historical overview of qualitative and quantitative occupational measures developed in the sociological literature. The "Socioeconomic Status Indexes and Scales: An Overview" section focuses on distinctions among several widely used socioeconomic status indexes based on continuous measures. The "Limitations of Previous Occupational Indexes" section discusses the limitations of previous occupation-based socioeconomic indexes. We then introduce our new measure of occupational status based on percentile ranks in the "Methodology" section. The "Results" section describes historical changes in occupational percentile ranks using U.S. Census data from 1850 to 2000 and American Community Survey (ACS) data from 2001 to 2018. The "Conclusion" section concludes the paper.</p> <hd id="AN0184233957-3">Occupational Measures: Historical Perspective</hd> <p>While commonly used, occupation is among the most challenging and least agreed-upon measures in surveys and population registers. An "occupation" refers to the aggregation of inherently different jobs that are sufficiently similar with respect to requirements, duties, and responsibilities so as to be categorized together for statistical purposes.</p> <p>In the United States, occupation was first enumerated in the 1820 full-count Census, on the basis of families, rather than individuals. Three broad occupational classes were used: Agriculture, commerce, and manufactures. This item was dropped in the 1830 Census but was added back in 1840 and extended to seven classes: Mining, agriculture, commerce, manufactures and trades, navigation of the ocean, navigation of canals, lakes, and rivers, and learned professions and engineers. A major change occurred in 1850, when the U.S. Census Bureau decided to shift from family to individual enumeration, collecting 323 specific occupations for men over the age of 15. In 1860, women were also subject to occupational enumeration, and the number of detailed occupations was expanded to 584 ([<reflink idref="bib18" id="ref38">18</reflink>]; [<reflink idref="bib82" id="ref39">82</reflink>]).[<reflink idref="bib7" id="ref40">7</reflink>] We summarize changes in the U.S. Census Bureau's occupational classification in Appendix Table A1 in the supplemental material.</p> <p>Each occupation in the U.S. Census Bureau's classification system is an aggregation of more detailed occupational titles, which are in turn based on respondents' descriptions of job titles and/or types of work. Since 1900, the Census Bureau relied on clerical occupation coding staff, who converted write-ins for job titles and duties in the U.S. decennial Censuses and demographic surveys to a list of over 30,000 occupation titles in alphabetical order, known as the Alphabetic Index of Occupations ([<reflink idref="bib98" id="ref41">98</reflink>], [<reflink idref="bib99" id="ref42">99</reflink>]). These index titles were then aggregated into a system of Census Occupational Classifications, later known as the Standard Occupational Classification (SOC) after 2000, with more than 1000 detailed occupational categories.</p> <p>Another major U.S. government occupation classification system is the dictionary of occupational titles (DOTs), originally published as a reference manual for local offices of the U.S. Employment Service. The DOT listed 13,000–30,000 different job titles from the 1930s to the 1990s and provided linkages from these job titles to occupations ([<reflink idref="bib104" id="ref43">104</reflink>],[<reflink idref="bib105" id="ref44">105</reflink>], [<reflink idref="bib17" id="ref45">17</reflink>], [<reflink idref="bib100" id="ref46">100</reflink>], [<reflink idref="bib101" id="ref47">101</reflink>], [<reflink idref="bib102" id="ref48">102</reflink>]). The DOT was created and updated by professional job analysts who visited U.S. workplaces and recorded job requirements. It has been widely used as a reference manual for employment services (such as matching job applicants with jobs and guiding job training, vocational education, and career counseling) and for converting occupational coding obtained from surveys to detailed census categories ([<reflink idref="bib13" id="ref49">13</reflink>]; [<reflink idref="bib34" id="ref50">34</reflink>]).[<reflink idref="bib8" id="ref51">8</reflink>] Many job titles included in the Alphabetic Index of Occupations came from the DOT. The DOT was later replaced by an online system called the Occupational Information Network (O*NET), based on input from job incumbents and occupational experts with direct experience working in different occupations ([<reflink idref="bib103" id="ref52">103</reflink>]). The O*NET subsequently adopted an eight-digit O*NET-SOC classification system by subdividing the six-digit SOC occupations into more detailed occupations with two extended digits ([<reflink idref="bib77" id="ref53">77</reflink>]).</p> <p>The international version of the U.S. Census Bureau's occupational classification systems is known as the ISCO, developed by the International Labor Organization. The ISCO has been widely used for international comparisons of occupational data, particularly for countries that have not developed their own national classifications ([<reflink idref="bib38" id="ref54">38</reflink>], [<reflink idref="bib39" id="ref55">39</reflink>]; [<reflink idref="bib95" id="ref56">95</reflink>], [<reflink idref="bib97" id="ref57">97</reflink>]). The first version of the ISCO, later known as ISCO-58, was proposed in 1957 by the Ninth International Conference of Labor Statisticians. This version was later superseded by ISCO-68, ISCO-88, and ISCO-08. ISCO-08 includes ten major groups, 43 sub-major groups, 130 minor groups, and 436 4-digit unit groups in total.</p> <p>Detailed occupations are typically not directly used in analyses of census or survey data. They are often aggregated into a small number of broad categories or classes. For example, [<reflink idref="bib27" id="ref58">27</reflink>], [<reflink idref="bib28" id="ref59">28</reflink>]) proposed an occupational classification system based on 6 main groups and 12 socioeconomic subgroups, which later influenced the Census Bureau's development of occupational systems (see a review in [<reflink idref="bib30" id="ref60">30</reflink>]: 53–81). He first divided occupations into "hand" and "head" groups and then ranked the "hand" workers by the degree of skill and the "head" workers based on the degree of training required for the job and the level of prestige.[<reflink idref="bib9" id="ref61">9</reflink>] Blau and Duncan ([<reflink idref="bib5" id="ref62">5</reflink>], Table 2.1) developed 17 occupational groups based on major occupational groups used in the 1950 U.S. Census: professionals (self-employed), professionals (salaried), managers, salesmen (not in retail), proprietors, clerical workers, salesmen (in retail), craftsmen (in manufacturing), craftsmen (in construction), craftsmen (in other industries), operatives (in manufacturing), operatives (in other industries), laborers (in manufacturing), laborers (in other industries), farmers, and farm laborers. [<reflink idref="bib49" id="ref63">49</reflink>]) adopted a five-category occupational scheme, which includes upper nonmanual, lower nonmanual, upper manual, lower manual, and farming. In the U.K., [<reflink idref="bib44" id="ref64">44</reflink>]) proposed a seven-category standard classification.[<reflink idref="bib10" id="ref65">10</reflink>][<reflink idref="bib31" id="ref66">31</reflink>]) developed two versions of the well-known EGP class scheme. One issue with these categorical approaches is their high degree of aggregation—workers within a large category are all assumed to be relatively homogeneous, or at least interchangeable. To modify this assumption of strong homogeneity, Grusky, Weeden, and their collaborators developed a microclass occupational scheme and applied it to the study of occupational segregation, social mobility, and inequality ([<reflink idref="bib68" id="ref67">68</reflink>]; [<reflink idref="bib112" id="ref68">112</reflink>], [<reflink idref="bib113" id="ref69">113</reflink>]).</p> <p>Alternatively, researchers have converted original occupational categories into scalar measures of occupational status. One well-known example is Duncan's ([<reflink idref="bib24" id="ref70">24</reflink>]) socioeconomic index, discussed below, which is a weighted average of occupational education and occupational income. It is not our intention to evaluate the relative merits of many earlier approaches, as there are advantages and disadvantages associated with each. Judicious choices should be made within a concrete research context to meet particular theoretical, methodological, and practical needs. Instead, we propose a relative occupational measure, a version of the scalar approach, and show that our measure has several advantages over other existing alternatives, especially in its ease of use and interpretability in comparing occupations cross-nationally or across time.</p> <hd id="AN0184233957-4">Socioeconomic Status Indexes and Scales: An Overview</hd> <p>Scalar measures of socioeconomic status rest on the assumption that the vertical status hierarchy of different occupations can be characterized by a single, latent, continuous scale. It is important to recognize that, being latent, a scalar measure always needs to be normalized with two constraints: Location and scale. Location normalization involves setting zero; scale normalization defines the magnitude of the measure, say, with 100 at the maximum. This approach emphasizes gradational and quantitative differences, a view deeply rooted in the structuralist perspective of societies as hierarchical stratification systems ([<reflink idref="bib22" id="ref71">22</reflink>]).[<reflink idref="bib11" id="ref72">11</reflink>] In working with continuous measures of occupations, researchers take a data-driven approach to estimating the status grading of the occupational structure, rather than imposing an a priori occupational hierarchy. Previous continuous occupational measures fall into three categories: Occupational prestige, socioeconomic index, and percentile scores. We highlight significant developments in each of these categories below and summarize major occupational measures in the literature in Appendix Table A2 in the supplemental material.</p> <hd id="AN0184233957-5">Occupational Prestige</hd> <p>Occupational prestige, the extent of social deference or derogation conferred to incumbents of an occupation, was one of the earliest and remains one of the most widely used indicators of social standing. Measurements of occupational prestige typically rely on subjective evaluations of occupations by either population-representative or well-informed respondents. In one of the earliest studies on occupational prestige, [<reflink idref="bib21" id="ref73">21</reflink>]) asked students in the United States and Russia aged 12–17 to rank 45 occupations.[<reflink idref="bib12" id="ref74">12</reflink>] Most notably, the National Opinion Research Center (NORC) at the University of Chicago collected information on occupational prestige in four major national surveys: the NORC 1947 and 1963 Occupational Prestige Surveys and the 1989 and 2012 General Social Surveys (GSSs) ([<reflink idref="bib78" id="ref75">78</reflink>]; [<reflink idref="bib86" id="ref76">86</reflink>]). Because it is not practical to ask every respondent to evaluate the prestige of all occupations, typically only a subsample is given the task of evaluating a subset of occupations. Integrating information through statistical modeling across different subsamples enables researchers to derive occupational prestige for all occupations. For example, assume that each respondent was asked to rank nine occupations from the lowest to the highest. If we assume equal distances between two adjacently rated categories, the value of the <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>j</mi></math> </ephtml> th occupation can be calculated as follows ([<reflink idref="bib114" id="ref77">114</reflink>]): <ephtml> <math display="block" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>P</mi><mi>j</mi></msub><mo>=</mo><munderover><mo>∑</mo><mrow><mi>i</mi><mo>=</mo><mn>1</mn></mrow><mn>9</mn></munderover><mn>12.5</mn><mo stretchy="false">(</mo><mi>i</mi><mo>−</mo><mn>1</mn><mo stretchy="false">)</mo><msub><mi>X</mi><mrow><mspace width=".1em" /><mi>j</mi><mi>i</mi></mrow></msub></math> </ephtml></p> <p>Graph</p> <p>where <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>X</mi><mrow><mspace width=".1em" /><mi>j</mi><mi>i</mi></mrow></msub></math> </ephtml> is the proportion of rankings received by the <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>j</mi></math> </ephtml> th occupation with ranking <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>i</mi></math> </ephtml> , with <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><munderover><mo>∑</mo><mrow><mi>i</mi><mo>=</mo><mn>1</mn></mrow><mn>9</mn></munderover><msub><mi>X</mi><mrow><mspace width=".1em" /><mi>j</mi><mi>i</mi></mrow></msub><mo>=</mo><mn>1</mn></math> </ephtml> . The prestige estimate <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>P</mi><mi>j</mi></msub></math> </ephtml> is a weighted score normalized in the range of 0–100.</p> <p>Using survey data, [<reflink idref="bib84" id="ref78">84</reflink>]) developed the first set of prestige scores for all 1960 U.S. Census occupations. [<reflink idref="bib51" id="ref79">51</reflink>]) and [<reflink idref="bib23" id="ref80">23</reflink>]) updated the prestige scores of occupations in the 1970 U.S. Census. These scores were subsequently updated for the 1980, 1990, and 2010 Census occupations ([<reflink idref="bib63" id="ref81">63</reflink>]; [<reflink idref="bib73" id="ref82">73</reflink>]; [<reflink idref="bib92" id="ref83">92</reflink>]). The IPUMS project recently added the Siegel prestige score on the basis of the 1950 Census occupational scheme and the Nakao-Treas prestige score on the basis of the 1990 scheme in its harmonized U.S. Census and ACS data. In addition to these prestige ratings for the United States, [<reflink idref="bib95" id="ref84">95</reflink>], [<reflink idref="bib97" id="ref85">97</reflink>]) developed the Standard International Occupational Prestige Scale using the ISCO.</p> <p>One remarkable finding emerging from previous research is that occupational prestige, measured by averaging individual-level survey responses, is highly stable across populations and time ([<reflink idref="bib15" id="ref86">15</reflink>]; [<reflink idref="bib16" id="ref87">16</reflink>]; [<reflink idref="bib115" id="ref88">115</reflink>]). [<reflink idref="bib93" id="ref89">93</reflink>]) showed that subgroups by sex, age, place of residence, and occupation tended to give highly consistent prestige ratings of different occupations. [<reflink idref="bib56" id="ref90">56</reflink>]) found a correlation coefficient of 0.99 for occupational prestige estimated from the 1947 and 1963 NORC surveys. [<reflink idref="bib50" id="ref91">50</reflink>]) showed that prestige ratings for occupations in a few historical studies of mid-nineteenth-century American cities (Philadelphia, Pennsylvania; Hamilton, Ontario; Kingston, New York; Buffalo, New York; and Poughkeepsie, New York) are highly correlated with those in the 1964–1965 NORC surveys. [<reflink idref="bib65" id="ref92">65</reflink>]) revealed a pairwise correlation coefficient in occupational prestige in the range of 0.74–0.97 between six industrialized countries. [<reflink idref="bib57" id="ref93">57</reflink>]) and [<reflink idref="bib95" id="ref94">95</reflink>]) further showed an average correlation coefficient of 0.81 among occupations across 55 countries in the 1970s. Attributing this finding to Treiman's ([<reflink idref="bib97" id="ref95">97</reflink>]) dissertation-based book, [<reflink idref="bib62" id="ref96">62</reflink>]) called this property of prestige scores the "Treiman constant."[<reflink idref="bib13" id="ref97">13</reflink>]</p> <p>As a highly visible and stable social attribute, prestige is one of many features of an occupation. As [<reflink idref="bib70" id="ref98">70</reflink>]) argued, "people living at different times, persons living in different societies, and members of different groups in the same society may evince remarkably similar evaluations of occupations" when they are asked to grade, score, or rank occupations using a single, ordered dimension. The simple implementation of this measure also reflects its weakness. Individuals may vary substantially in the criteria they adopt when assigning prestige ratings for occupations ([<reflink idref="bib20" id="ref99">20</reflink>]; [<reflink idref="bib47" id="ref100">47</reflink>]). Prestige scores are thus not always highly correlated with other occupation-level variables such as political behavior, social participation, health, and other wellbeing outcomes. [<reflink idref="bib52" id="ref101">52</reflink>]) called this problem "the low criterion validity of occupational prestige."</p> <hd id="AN0184233957-6">Socioeconomic Indexes</hd> <p>In 1961, Otis Dudley Duncan published his work on "a socioeconomic index for all occupations," the Duncan Socioeconomic Index (hereafter, the Duncan SEI), which quickly became the most popular continuous occupational status measure. The main difference between prestige scores and socioeconomic indexes like the Duncan SEI is that the latter are composites based on objective indicators at the occupation level and thus do not require data on personal evaluations of occupations. Specifically, the Duncan SEI is a weighted sum of the average education and average income of incumbents within an occupation.[<reflink idref="bib14" id="ref102">14</reflink>]</p> <p>[<reflink idref="bib24" id="ref103">24</reflink>]) derived the weights for the Duncan SEI using public opinion data for a small set of occupations from the 1947 NORC Occupational Prestige Survey and occupational data from the 1950 U.S. Census (National Opinion Research Center [NORC] 1947, 1948). He first regressed the occupational prestige of 45 occupational titles in the 1947 survey on the age-specific average education and income of matched occupations in the 1950 Census and obtained the following equation:[<reflink idref="bib15" id="ref104">15</reflink>] <ephtml> <math display="block" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mover><mi>Y</mi><mo stretchy="false">^</mo></mover></mrow><mrow><mn>1</mn><mi>j</mi></mrow></msub><mo>=</mo><mn>0.59</mn><mo>*</mo><msub><mi>Y</mi><mrow><mn>2</mn><mi>j</mi></mrow></msub><mo>+</mo><mn>0.55</mn><mo>*</mo><msub><mi>Y</mi><mrow><mn>3</mn><mi>j</mi></mrow></msub><mo>−</mo><mn>6.0.</mn></math> </ephtml></p> <p>Graph</p> <p>In the original NORC survey questionnaire, respondents were asked to choose one rating for each occupation from the choices of "excellent," "good," "average," "somewhat below average," and "poor." <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>Y</mi><mn>1</mn></msub></math> </ephtml> refers to the percentage of "excellent" or "good" ratings.[<reflink idref="bib16" id="ref105">16</reflink>] <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>Y</mi><mn>2</mn></msub></math> </ephtml> refers to the percentage of male occupational incumbents who earned incomes of $3,500 or more in 1949, excluding those who did not answer the income question or reported no income. <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>Y</mi><mn>3</mn></msub></math> </ephtml> refers to the percentage of male incumbents in the 1950 U.S. Census who were high school graduates.[<reflink idref="bib17" id="ref106">17</reflink>] Duncan further adjusted the Duncan SEI for age differences among occupations using the indirect standardization method. Specifically, he treated the age-specific distribution for the entire male experienced civilian labor force aged 14 and above as the standard and used it to adjust the actual age distribution of a particular occupation to yield an overall expected proportion of high school graduates or of incomes of $3,500 or more for incumbents of that occupation. Duncan devised the method to predict the prestige levels of all occupations in the 1950 Census even though direct prestige ratings were unavailable for most of them. The predicted value of <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mrow><mover><mi>Y</mi><mo stretchy="false">^</mo></mover></mrow><mn>1</mn></msub></math> </ephtml> obtained from the age-adjusted education and income values for a given occupation in the Census is the Duncan SEI.</p> <p>The original Duncan SEI was based on the 1950 Census, but Duncan later updated the SEI scores using the 1960 Census occupations and used these resulting scores to analyze data from the "Occupational Changes in a Generation" (OCG-I) study ([<reflink idref="bib5" id="ref107">5</reflink>]). [<reflink idref="bib51" id="ref108">51</reflink>]) and [<reflink idref="bib91" id="ref109">91</reflink>]) updated the Duncan SEI for the 1970 Census occupations. [<reflink idref="bib90" id="ref110">90</reflink>]) updated the Duncan SEI for the 1980 Census occupations. [<reflink idref="bib52" id="ref111">52</reflink>]) further re-estimated the Duncan SEI using the occupational prestige ratings from the 1989 NORC GSS and occupational education and income data from the 1990 Census.</p> <p>Socioeconomic indexes are attractive because they generalize survey-based prestige ratings for a small set of occupations to a wider range of occupations based on occupation-level objective attributes.[<reflink idref="bib18" id="ref112">18</reflink>] Yet, as Duncan ([<reflink idref="bib24" id="ref113">24</reflink>]) noted, the two objective characteristics, i.e., occupational income and occupational education, are not the only social determinants of prestige ratings. Other factors may also influence individuals' judgment of prestige values. [<reflink idref="bib32" id="ref114">32</reflink>]), [<reflink idref="bib55" id="ref115">55</reflink>]), and [<reflink idref="bib94" id="ref116">94</reflink>]) provided extensive discussions about the strengths and limitations of prestige versus socioeconomic index measures.[<reflink idref="bib19" id="ref117">19</reflink>][<reflink idref="bib52" id="ref118">52</reflink>]) demonstrated that occupational education and occupational income scores represent two correlated, but not always aligned, aspects of occupational status and would be better analyzed separately rather than being combined into a composite, as in the Duncan SEI.</p> <p>As discussed earlier, the Duncan SEI was built upon the 1947 NORC occupational prestige scores. Yet, another strand of research has also developed "prestige-free" measures of socioeconomic status. [<reflink idref="bib6" id="ref119">6</reflink>]) and [<reflink idref="bib8" id="ref120">8</reflink>]) created the standard scores of occupational statuses based on income and years of schooling in the Canadian Census. [<reflink idref="bib58" id="ref121">58</reflink>], [<reflink idref="bib59" id="ref122">59</reflink>], [<reflink idref="bib60" id="ref123">60</reflink>]) and [<reflink idref="bib61" id="ref124">61</reflink>]) developed the ISP based on two or more social factors related to educational attainment, occupation, quality of the neighborhood of residence, marital status, and sex, sometimes also known as the Hollingshead Two-Factor, Three-Factor, and Four-Factor indexes.[<reflink idref="bib20" id="ref125">20</reflink>] The IPUMS project recently added the Duncan SEI on the basis of the 1950 Census occupational scheme and the Hauser-Warren SEI on the basis of the 1990 scheme to harmonized microdata from the U.S. Census/ACS data.</p> <hd id="AN0184233957-7">Percentile Scores</hd> <p>In a series of papers, [<reflink idref="bib75" id="ref126">75</reflink>], [<reflink idref="bib76" id="ref127">76</reflink>]) and [<reflink idref="bib74" id="ref128">74</reflink>]) developed an occupational percentile score measure to capture substantial changes in occupational statuses based on Census occupational categories. The rationale behind this measure is that high-status occupations have grown, in both number and size, over time, whereas low-status occupations have shrunk. Nam and Powers rated occupations using the average percentile of their incumbents in the cumulative distribution of workers across occupations after the occupations were ranked by median education and median income, respectively.</p> <p>Nam and Powers ([<reflink idref="bib76" id="ref129">76</reflink>]:127–142) first sorted occupations according to the educational level of men over 14 years old in the civilian labor force for the 1970 Census data. They then derived the education-based distribution by accumulating the proportion of workers employed in each occupation from the lowest-educated occupation to the highest-educated occupation. They used the same procedure to create a cumulative distribution by occupational income. They averaged the midpoints of the two cumulative distributions of workers in a given occupation and converted the raw score to a percentile status score for the occupation, which necessarily lies between 0 and 100. Each score indicates the approximate cumulative percentage of workers who are in "occupations having combined average levels of education and income below that for the given occupation."</p> <p>[<reflink idref="bib32" id="ref130">32</reflink>]) adopted a similar approach when they compared the Duncan SEI, Siegel's NORC prestige score, and Treiman's international prestige index. Specifically, they ranked occupations by these raw scores for the 1962 OCG-I data and then calculated and compared the percentile scores for all occupations in the 1970 Census based on the three different ranking criteria. However, their goal was not to create a new occupational measure but to use the percentile scores as a normalization method to compare these three socioeconomic status scales.</p> <p>The Nam-Powers-Boyd score encompasses several desirable properties. First, the score can be interpreted as representing an occupation's relative standing. Second, the score is scale-free, naturally scaled between 0 and 100 as percentiles, representing the position of male workers in an occupation relative to all other male civilian workers in the entire labor market.[<reflink idref="bib21" id="ref131">21</reflink>] Third, and most importantly, the score changes in response to the evolving occupational distribution, even in the absence of changes in workers' characteristics within an occupation. [<reflink idref="bib25" id="ref132">25</reflink>]) postulated that occupational status changes with a glacial speed, and thus a constant score can be reasonably assigned to each occupation. This view may be correct for an observation period spanning a few decades, but not from a longer term perspective.[<reflink idref="bib22" id="ref133">22</reflink>] As the score reflects a relative status measure, the social standing of a group of workers depends on their own characteristics as well as those of workers in other occupations (see [<reflink idref="bib48" id="ref134">48</reflink>] and [<reflink idref="bib80" id="ref135">80</reflink>] for discussions on the advantage of the Nam-Powers-Boyd score). However, Hauser and Warren ([<reflink idref="bib52" id="ref136">52</reflink>]:193–194) pointed out that this score has the same problem as the Duncan SEI, by averaging occupational income and occupational education percentiles for a given occupation.</p> <hd id="AN0184233957-8">Index of Job Desirability</hd> <p>A potential limitation of previous occupational status measures is that they do not reflect desirability of jobs. Incorporating Weberian views of social stratification, [<reflink idref="bib67" id="ref137">67</reflink>]) developed the Index of Job Desirability (IJD) that intends to measure all three dimensions of inequality: Class, status, and power. In practice, the index combines 13 nonmonetary job characteristics with occupational earnings and weighs the relative importance of these characteristics in a score that varies between 41 and 689. These nonmonetary job characteristics include working hours per week, vacation weeks, on-the-job training, risk of job loss, educational requirements, the proportion of repetitive work, dirty work conditions, control of one's own hours, frequent supervision, union contract, federal employee, state or local employee, and whether having a boss. [<reflink idref="bib42" id="ref138">42</reflink>]) developed a similar measure called the general desirability of occupations for the United Kingdom.</p> <p>Compared to prestige scores and socioeconomic indexes, the development of the IJD required substantially more effort in data collection. The original IJD was derived from 14 closed-ended questions asked in a survey with 621 respondents. Given the small sample size, the occupations of these respondents covered only a subsample of all occupations in the population. The number of occupations included in the sample would affect the IJD normalization, so that the range of this index score changes when more or fewer occupations are considered. Although the original IJD scores are not widely used in empirical research, it has influenced the subsequent development of the U.S. Government Employment and Training Administration's O*NET, which periodically collects data related to occupational characteristics based on input from job incumbents and occupational experts. O*NET consists of a content model with hundreds of descriptions of work and workers organized into domains, such as skills, knowledge, tasks, work activities, and work values.[<reflink idref="bib23" id="ref139">23</reflink>] Yet, the O*NET content model does not provide direct measures of occupational statuses (see a review in [<reflink idref="bib46" id="ref140">46</reflink>]).</p> <hd id="AN0184233957-9">IPUMS Occupational Scores</hd> <p>IPUMS recently released its (<reflink idref="bib1" id="ref141">1</reflink>) occupational education score, (<reflink idref="bib2" id="ref142">2</reflink>) occupational income score, and (<reflink idref="bib3" id="ref143">3</reflink>) occupational earnings score using 1950 and 1990 Census occupational coding systems. The IPUMS occupational education scores rely on the percentage of employed civilian workers aged 16 and above in each occupation with one or more years of college education. The score is calculated for the 1950 and 1990 Census occupations, separately. Because detailed education questions were not included in the Federal Population Census until 1940, IPUMS assigns scores calculated from the 1950 data for years before 1950. In calculating the occupational earnings score, the median earnings were first standardized—namely, by subtracting the mean earnings of all occupations and dividing the difference by the standard deviation of occupational earnings (i.e., deriving the z-score)—and then converted into a percentile rank. In the calculation of the occupational income score, the unstandardized median total income of workers within each occupation measured in hundreds of dollars was used instead. The income variable includes workers' wages, business income, and farm income. A given occupation receives the same score in each Census before 1950 due to the lack of income information, but the score is based on updated income information across Census years after 1950.</p> <hd id="AN0184233957-10">Limitations of Previous Occupational Indexes</hd> <p>From the very beginning of social stratification research, scholars have been concerned with the comparability of occupations across time or generation. Occupations evolve, in ways big or small, in response to economic changes, with the emergence of new occupations and the obsolescence of old ones. In recent decades, many of these changes were driven by the influx of new technologies, educational upgrading, occupational licensing, and the restructuring of the local labor market due to globalization and offshoring. As a result, the relative socioeconomic standing of the same occupation may vary over time as the overall occupational structure evolves, presenting challenges for studies of intergenerational mobility. For example, if parents and offspring stay in the same occupation, but the relative status of the occupation itself changes, should we consider such a scenario as exemplifying intergenerational mobility or immobility?</p> <p>Social stratification researchers have long noticed this problem and considered whether or not separate occupational standing schemes should be applied to individuals born into different birth cohorts. [<reflink idref="bib25" id="ref144">25</reflink>]) argued that the same occupational score scheme could be used for intercohort analyses as long as the occupational structure changes at a slow pace. Duncan's own calculations suggest that correlation coefficients in occupational prestige scores or occupational mean income between two consecutive survey years in 1925, 1940, 1947, and 1963 were between 0.95 and 0.99. The correlation coefficient between prestige scores measured 38 years apart was as high as 0.934. Follow-up evidence also shows that the interannual correlation coefficient in occupational prestige ([<reflink idref="bib56" id="ref145">56</reflink>]) and occupational education score ([<reflink idref="bib54" id="ref146">54</reflink>]) can be as high as 0.97. Thus, Duncan ([<reflink idref="bib25" id="ref147">25</reflink>]:708) concluded, "the structure, if it continued to evolve under these conditions, would gradually drift away from its initial configuration to one which bore no resemblance thereto ...Yet, on the estimates now available, this would take quite a long time."</p> <p>Duncan's conclusion is limited to the comparison of birth cohorts only a few decades apart. However, we do not know if commonly used occupational scales are stable for long-term historical changes over many decades or centuries. It is time to revisit the comparability of occupations in terms of classifications and scales in light of the growing availability of historical and contemporary administrative and survey data ([<reflink idref="bib83" id="ref148">83</reflink>]; [<reflink idref="bib87" id="ref149">87</reflink>]).</p> <p>Building on previous work, we develop an occupational measure for relative occupational status over long periods of time or even across multiple generations. Substantial changes in occupational status occur not only because of the increasing diversity of work or the changing nature of activities under the same occupational titles but also because of historical changes in the proportions of workers employed in different occupations. As Nam and Powers ([<reflink idref="bib75" id="ref150">75</reflink>]:165) observed, "the relative status levels of an exceedingly high percentage of occupations were lower in 1960 than in 1950, owing to a general depression of the status structure brought about by decreasing relative numbers of persons in low-status occupations and corresponding increasing relative numbers in high-status occupations." In recent decades, the expansion of the technology sector and manufacturing job losses have resulted in relatively more persons being situated at the top of the occupational status scale, and fewer at the bottom. As a result, many previously high-status jobs have become less privileged, and low-status jobs have become even more disadvantaged in status rankings. To account for occupational restructuring over time, we construct an occupational percentile measure for each birth cohort, assigning possibly different status scores to workers in the same occupation but who entered it at varying times.</p> <hd id="AN0184233957-11">Methodology</hd> <p></p> <hd id="AN0184233957-12">Data and Variables</hd> <p>We constructed occupational percentile ranks using the IPUMS U.S. Population Censuses from 1850 to 2000 and ACSs from 2001 to 2018.[<reflink idref="bib24" id="ref151">24</reflink>]Appendix Table A3 in the supplemental material summarize these data sources. When full-count Census data were available (e.g., 1850–1940), we used full-count data over Census samples, except for 1890, for which the original data were damaged. When both 1-percent and 5-percent samples were available (e.g., 2000), we chose the larger sample.[<reflink idref="bib25" id="ref152">25</reflink>]</p> <p>Our analyses pooled individuals who were born in the same year but were observed in different Census years. We first restricted the sample to men and women aged 25–64 and then generated cohort-specific occupational percentile ranks based on the literacy rate or educational distribution within an occupation. Birth cohorts are defined by a 10-year interval centered on the midpoint year. For example, the 1890 birth cohort refers to workers who were born between 1886 and 1895. Appendix Table A4 in the supplemental material shows the availability of data by birth cohort.</p> <p>The education variable is measured by the highest year of schooling or degree completed, as asked in the population Censuses. The harmonized IPUMS Census coding includes 0 (N/A or no schooling), 1 (Nursery school to grade 4), 2 (Grade 5, 6, 7, or 8), 3 (Grade 9), 4 (Grade 10), 5 (Grade 11), 6 (Grade 12), 7 (1 year of college), 8 (2 years of college), 9 (3 years of college), 10 (4 years of college), and 11 (5+ years of college). Given that some of these groups contain very few observations for some birth cohorts, we further collapse them into six groups: 1 (No schooling), 2 (1–8 years of schooling), 3 (9–11 years of schooling), 4 (12 years of schooling), 5 (13–15 years of schooling), and 6 (16+ years of schooling). The educational attainment variable was not included in the U.S. Census until 1940. Workers were only asked to report literacy—whether they could write or read—before 1940. Therefore, we used the literacy variable to generate the percentile ranks of occupations for birth cohorts born between 1790 and 1880.</p> <p>Although educational measures are different before and after 1940, percentile ranks measured by literacy rates and detailed levels of education yield very similar results for cohorts who were observed in multiple census years and for whom both measures are available (i.e., cohorts born between 1880 and 1910). The differences in percentile ranks as measured by these two education variables are minimal for most occupations. We conducted sensitivity analyses to assess potential measurement errors caused by the switch from literacy to educational attainment measures in the 1940 U.S. Census. These robustness check results are presented in online Appendix Figure A4 in the supplemental material. For all the analyses reported in this paper, we used estimates of occupational percentile ranks based on educational attainment when both measures are available.</p> <hd id="AN0184233957-13">The Construction of Percentile Ranks</hd> <p>Four steps are adopted in the construction of occupational percentile ranks:</p> <hd id="AN0184233957-14">(1) Defining a Consistent Occupational Grouping over Time</hd> <p>For ease of interpretation and consistency in measurement over Census years for the same cohorts, the researcher may wish to have the same occupational classification over time. Ideally, we classify workers into a parsimonious number of occupational categories distinct from one another while also maintaining relative homogeneity within each category. The DOT scheme discussed earlier provides the most detailed categorization of occupations currently available, but there has been no administrative effort to collect information pertaining to incumbents under each occupational title. The detailed occupational list in modern decennial Censuses provides a more tractable set of occupations, retaining more than 300 occupational groups with a fair degree of within-group similarity. It is worth noting that the generation of occupational percentile ranks depends on the coding of occupations. More detailed occupation classification would capture more nuanced differences between occupational groups.</p> <p>We decided to map historical occupations measured in different Census years to one of the three Census Bureau occupational classifications (1950, 1990, or 2010) because these harmonized occupation codes already exist in all years of IPUMS Census and ACS data, with the variable names OCC1950, OCC1990, and OCC2010, respectively. To illustrate our method, we use OCC1950 as an example.[<reflink idref="bib26" id="ref153">26</reflink>] The original 1950 occupational classification consists of 269 valid occupational categories.[<reflink idref="bib27" id="ref154">27</reflink>] However, not all occupations in the 1950 Census scheme are consistently measured across the period 1850–2018. As new occupational titles emerged and old titles split and recombined, some Census 1950 occupation categories became more heterogeneous over time.[<reflink idref="bib28" id="ref155">28</reflink>] Also, for some occupations with few workers (i.e., fewer than 100 observations), their percentile ranks are undefined but inconsequential to our task of measuring workers' social status, as very few workers would be found in these occupations. Note that the final step is optional, as the rest of the methodology can be applied without it.</p> <hd id="AN0184233957-15">(2) Rating and Ranking Occupations</hd> <p>The previous step results in a dataset that contains 5,380 observations (269 occupations <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mo>×</mo></math> </ephtml> 20 cohorts), each of which refers to a 1950 Census occupation for a certain birth cohort born between 1790 and 1980. Other variables in the dataset include the number of workers within each occupation and the number of persons with varying levels of education. The detailed education variable was not available until the 1940 Census. Below we illustrate the method using the education variable as example. For Census years prior to 1940, we generate occupations' literacy scores from a dichotomous variable (0 = illiterate; 1 = literate—that is, can both read and write).</p> <p>Next, we measure occupational statuses based on the educational distribution within each occupation. For occupation <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>i</mi></math> </ephtml> , its status score is the weighted average of the educational percentile: <ephtml> <math display="block" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>S</mi><mrow><mi>i</mi><mi>t</mi></mrow></msub><mo>=</mo><munder><mo>∑</mo><mi>x</mi></munder><mi>P</mi><mo stretchy="false">(</mo><mi>x</mi><mo>∣</mo><mi>i</mi><mo>,</mo><mi>t</mi><mo stretchy="false">)</mo><mo>⋅</mo><msub><mi>Q</mi><mi>t</mi></msub><mo stretchy="false">(</mo><msup><mi>r</mi><mi>x</mi></msup><mo stretchy="false">)</mo><mo>,</mo></math> </ephtml></p> <p>Graph</p> <p>where <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>P</mi><mo stretchy="false">(</mo><mi>x</mi><mo>∣</mo><mi>i</mi><mo>,</mo><mi>t</mi><mo stretchy="false">)</mo></math> </ephtml> is the proportion of educational level <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>x</mi></math> </ephtml> in occupation <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>i</mi></math> </ephtml> and birth cohort <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>t</mi></math> </ephtml> and <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>Q</mi><mi>t</mi></msub><mo stretchy="false">(</mo><msup><mi>r</mi><mi>x</mi></msup><mo stretchy="false">)</mo></math> </ephtml> is the percentile rank of educational level <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>x</mi></math> </ephtml> in birth cohort <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>t</mi></math> </ephtml> . For example, assume we have four educational groups ranked from 1 (low) to 4 (high) and varying in size from 40, 30, and 20 to 10 in a general population that contains 100 individuals in total. The percentile rank of group 4 is 95—the midpoint of the 90th percentile and the 100th percentile. Likewise, the percentile ranks of groups 1, 2, and 3 would be 20, 55, and 80, respectively. Assume that for a specific occupation <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>i</mi></math> </ephtml> , the proportions of each educational group from 1 to 4 are 0.1, 0.35, 0.3, and 0.25, respectively. Thus, this occupation's status score is 69 (= 0.1*20 + 0.35*55 + 0.3*80 + 0.25*95). Note that the term <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>Q</mi><mi>t</mi></msub><mo stretchy="false">(</mo><msup><mi>r</mi><mi>x</mi></msup><mo stretchy="false">)</mo></math> </ephtml> , the education percentile score, normalizes educational attainment across cohorts. Specifically, an occupation with relatively more college-educated workers would have a higher status than an occupation with relatively fewer college-educated workers, all other things being equal. However, due to the expansion of higher education over time, the relative status of the college-educated group per se has declined. An occupation in which 20 percent of its workers are college-educated in 1940 would have a relatively higher status than an occupation with the same proportion of college-educated workers in 2000 because there was a larger proportion of college-educated workers in the overall labor force in 2000 than in 1940. To calculate the size of each educational group, we rely on a weighted person count because not everyone was asked to report education during some Census years.[<reflink idref="bib29" id="ref156">29</reflink>] Individuals with no occupation information are excluded from the estimation in this step.</p> <p>Following the recommendation of [<reflink idref="bib52" id="ref157">52</reflink>]), we derive occupational status on the basis of occupational education alone. We do not use occupational income because income information is either absent or very crude in historical data. [<reflink idref="bib53" id="ref158">53</reflink>]) showed that socioeconomic status scores based on occupational education and occupational income are not always consistent with each other. For example, women's occupational education has exceeded or trailed that of men for recent birth cohorts, but their occupational income largely falls behind that of men (see, e.g., [<reflink idref="bib12" id="ref159">12</reflink>]). Thus education- or prestige-based occupational measures typically yield higher socioeconomic standings for women than do income or wage-based measures ([<reflink idref="bib9" id="ref160">9</reflink>]; [<reflink idref="bib111" id="ref161">111</reflink>]).</p> <p>Gender has long been a concern in critiques of socioeconomic-index-based studies. The original NORC surveys excluded female-dominated occupations from prestige ratings (Reiss [<reflink idref="bib81" id="ref162">81</reflink>]:5), and these data were further restricted to male workers in Duncan's prestige regression for the construction of the Duncan SEI. [<reflink idref="bib36" id="ref163">36</reflink>]) showed that the exclusion of female workers or female-dominant occupations in prestige surveys had little impact on the estimation of occupational prestige ratings. However, [<reflink idref="bib10" id="ref164">10</reflink>]) compared socioeconomic indexes using only men ([<reflink idref="bib8" id="ref165">8</reflink>]), only women ([<reflink idref="bib7" id="ref166">7</reflink>]), and all members of the labor force for Canadian Census data and found that a socioeconomic index based on both sexes is preferable for studying occupational attainment.[<reflink idref="bib30" id="ref167">30</reflink>][<reflink idref="bib52" id="ref168">52</reflink>]) also recommended a socioeconomic index with workers of both sexes (see also [<reflink idref="bib111" id="ref169">111</reflink>]). For these reasons, we include both men and women in the calculation of the occupational status scores.</p> <p>With resulting status scores for all occupations for a given cohort, <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>S</mi><mrow><mi>i</mi><mi>t</mi></mrow></msub></math> </ephtml> from equation (<reflink idref="bib2" id="ref170">2</reflink>), we rank the occupations from the lowest to the highest in the 1950 classification. At the occupational level, [<reflink idref="bib97" id="ref171">97</reflink>]) showed the relative stability of occupational prestige ranks over time and across societies. We thus refer to the ranking at the occupation level as "Treiman's rank."</p> <hd id="AN0184233957-16">(3) Converting Occupational Ranks to Percentile Ranks</hd> <p>We further convert occupational ranks into percentile ranks at the workers' level by aggregating the number of incumbents engaged in each occupation into a cumulative distribution from the lowest- to the highest-ranked occupations within each cohort. Compared to the Treiman's ranks, the percentile ranks are less stable over time because they respond to changes in occupational sizes. For example, if a high-status occupation expands dramatically in size without changing its educational composition, the relative status of this occupation and occupations below it would decline. In summary, this step of normalization yields occupational percentile ranks at the workers' level, from 0 to 100, with a higher value indicating a higher occupational status.</p> <hd id="AN0184233957-17">(4) Percentile Rank Smoothing</hd> <p>Because we only have a 1-percent or 5-percent sample for the Census years after 1950, the number of observations within some occupations varies wildly across years. Also, some occupations were very small during the 19th century. For example, Statisticians and Actuaries (code 83) have only 2 observations for the 1800 cohort, 3 for the 1810 cohort, 8 for the 1820 cohort, and 25 for the 1830 cohort. To smooth out fluctuation caused by small samples, we use the moving average method. Specifically, we average the first two lagged values, the present value, and the first two forward values of the percentile rank series, with each value in the average receiving a weight of 1. The adjusted percentile for birth cohort <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>t</mi></math> </ephtml> is <ephtml> <math display="block" xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mrow><mi mathvariant="normal">pcrank</mi></mrow><mi>t</mi><mrow><mi>a</mi><mi>d</mi><mi>j</mi></mrow></msubsup><mo>=</mo><mrow><mfrac><mn>1</mn><mn>5</mn></mfrac></mrow><mo>*</mo><mo stretchy="false">(</mo><msub><mrow><mi mathvariant="normal">pcrank</mi></mrow><mrow><mi>t</mi><mo>−</mo><mn>2</mn></mrow></msub><mo>+</mo><msub><mrow><mi mathvariant="normal">pcrank</mi></mrow><mrow><mi>t</mi><mo>−</mo><mn>1</mn></mrow></msub><mo>+</mo><msub><mrow><mi mathvariant="normal">pcrank</mi></mrow><mi>t</mi></msub><mo>+</mo><msub><mrow><mi mathvariant="normal">pcrank</mi></mrow><mrow><mi>t</mi><mo>+</mo><mn>1</mn></mrow></msub><mo>+</mo><msub><mrow><mi mathvariant="normal">pcrank</mi></mrow><mrow><mi>t</mi><mo>+</mo><mn>2</mn></mrow></msub><mo stretchy="false">)</mo></math> </ephtml></p> <p>Graph</p> <p>This step is also optional. The smoothing adjustment is unnecessary when each occupation contains a sufficiently large number of workers for each birth cohort. We also construct 95-percent bounds for the percentile ranks based on 1000-bootstrap samples. Our appendix data files in the supplemental material include both smoothed and unsmoothed percentile rank estimates and their 95-percent confidence bounds.</p> <hd id="AN0184233957-18">Results</hd> <p>In Table 1 we present correlations across birth cohorts, both in Treiman's ranks, shown in the lower triangle, and percentile ranks, shown in the upper triangle. The shaded region refers to cases where correlations in Treiman's ranks are higher than corresponding ones in percentile ranks. Note that to be conservative, we report correlations in unsmoothed percentile ranks across cohorts, as correlations in smoothed percentile ranks would be larger.</p> <p>Table 1. Treiman's Rank and Percentile Rank Correlation Matrices</p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /></colgroup><thead><tr><th align="left" /><th align="left" colspan="20">Birth Cohort</th></tr><tr><th align="left" /><th align="left">1790</th><th align="left">1800</th><th align="left">1810</th><th align="left">1820</th><th align="left">1830</th><th align="left">1840</th><th align="left">1850</th><th align="left">1860</th><th align="left">1870</th><th align="left">1880</th><th align="left">1890</th><th align="left">1900</th><th align="left">1910</th><th align="left">1920</th><th align="left">1930</th><th align="left">1940</th><th align="left">1950</th><th align="left">1960</th><th align="left">1970</th><th align="left">1980</th></tr></thead><tbody><tr><td>1790</td><td /><td>0.89</td><td>0.86</td><td>0.82</td><td>0.77</td><td>0.73</td><td>0.68</td><td>0.63</td><td>0.59</td><td>0.55</td><td>0.52</td><td>0.5</td><td>0.47</td><td>0.45</td><td>0.44</td><td>0.41</td><td>0.39</td><td>0.37</td><td>0.37</td><td>0.34</td></tr><tr><td>1800</td><td>0.86</td><td /><td>0.88</td><td>0.85</td><td>0.8</td><td>0.76</td><td>0.71</td><td>0.66</td><td>0.61</td><td>0.56</td><td>0.52</td><td>0.5</td><td>0.48</td><td>0.46</td><td>0.44</td><td>0.43</td><td>0.4</td><td>0.38</td><td>0.37</td><td>0.35</td></tr><tr><td>1810</td><td>0.83</td><td>0.85</td><td /><td>0.87</td><td>0.84</td><td>0.79</td><td>0.73</td><td>0.69</td><td>0.63</td><td>0.58</td><td>0.54</td><td>0.5</td><td>0.49</td><td>0.47</td><td>0.45</td><td>0.44</td><td>0.42</td><td>0.39</td><td>0.38</td><td>0.36</td></tr><tr><td>1820</td><td>0.79</td><td>0.82</td><td>0.85</td><td /><td>0.86</td><td>0.82</td><td>0.77</td><td>0.71</td><td>0.66</td><td>0.6</td><td>0.55</td><td>0.51</td><td>0.49</td><td>0.47</td><td>0.46</td><td>0.46</td><td>0.43</td><td>0.41</td><td>0.39</td><td>0.37</td></tr><tr><td>1830</td><td>0.76</td><td>0.78</td><td>0.82</td><td>0.84</td><td /><td>0.84</td><td>0.81</td><td>0.75</td><td>0.69</td><td>0.63</td><td>0.57</td><td>0.52</td><td>0.49</td><td>0.47</td><td>0.46</td><td>0.46</td><td>0.45</td><td>0.43</td><td>0.41</td><td>0.38</td></tr><tr><td>1840</td><td>0.72</td><td>0.75</td><td>0.78</td><td>0.81</td><td>0.83</td><td /><td>0.83</td><td>0.79</td><td>0.73</td><td>0.66</td><td>0.61</td><td>0.54</td><td>0.51</td><td>0.48</td><td>0.46</td><td>0.47</td><td>0.47</td><td>0.45</td><td>0.44</td><td>0.41</td></tr><tr><td>1850</td><td>0.69</td><td>0.72</td><td>0.75</td><td>0.77</td><td>0.8</td><td>0.83</td><td /><td>0.81</td><td>0.78</td><td>0.71</td><td>0.64</td><td>0.58</td><td>0.54</td><td>0.51</td><td>0.48</td><td>0.47</td><td>0.48</td><td>0.47</td><td>0.48</td><td>0.45</td></tr><tr><td>1860</td><td>0.65</td><td>0.68</td><td>0.72</td><td>0.75</td><td>0.77</td><td>0.8</td><td>0.82</td><td /><td>0.8</td><td>0.76</td><td>0.69</td><td>0.61</td><td>0.58</td><td>0.54</td><td>0.5</td><td>0.5</td><td>0.47</td><td>0.48</td><td>0.5</td><td>0.49</td></tr><tr><td>1870</td><td>0.62</td><td>0.65</td><td>0.68</td><td>0.72</td><td>0.74</td><td>0.76</td><td>0.79</td><td>0.81</td><td /><td>0.78</td><td>0.74</td><td>0.67</td><td>0.61</td><td>0.57</td><td>0.53</td><td>0.51</td><td>0.5</td><td>0.47</td><td>0.49</td><td>0.49</td></tr><tr><td>1880</td><td>0.59</td><td>0.61</td><td>0.64</td><td>0.68</td><td>0.71</td><td>0.73</td><td>0.75</td><td>0.77</td><td>0.79</td><td /><td>0.76</td><td>0.72</td><td>0.66</td><td>0.61</td><td>0.56</td><td>0.53</td><td>0.5</td><td>0.49</td><td>0.47</td><td>0.48</td></tr><tr><td>1890</td><td>0.55</td><td>0.58</td><td>0.61</td><td>0.64</td><td>0.67</td><td>0.7</td><td>0.72</td><td>0.73</td><td>0.76</td><td>0.78</td><td /><td>0.75</td><td>0.72</td><td>0.66</td><td>0.6</td><td>0.58</td><td>0.55</td><td>0.49</td><td>0.51</td><td>0.45</td></tr><tr><td>1900</td><td>0.53</td><td>0.55</td><td>0.58</td><td>0.62</td><td>0.64</td><td>0.67</td><td>0.7</td><td>0.71</td><td>0.72</td><td>0.75</td><td>0.78</td><td /><td>0.76</td><td>0.74</td><td>0.68</td><td>0.66</td><td>0.64</td><td>0.61</td><td>0.59</td><td>0.56</td></tr><tr><td>1910</td><td>0.5</td><td>0.53</td><td>0.55</td><td>0.58</td><td>0.61</td><td>0.64</td><td>0.66</td><td>0.68</td><td>0.7</td><td>0.71</td><td>0.75</td><td>0.78</td><td /><td>0.75</td><td>0.73</td><td>0.68</td><td>0.66</td><td>0.64</td><td>0.62</td><td>0.61</td></tr><tr><td>1920</td><td>0.47</td><td>0.49</td><td>0.52</td><td>0.55</td><td>0.57</td><td>0.61</td><td>0.63</td><td>0.65</td><td>0.67</td><td>0.68</td><td>0.7</td><td>0.75</td><td>0.76</td><td /><td>0.73</td><td>0.73</td><td>0.68</td><td>0.65</td><td>0.65</td><td>0.61</td></tr><tr><td>1930</td><td>0.45</td><td>0.46</td><td>0.49</td><td>0.53</td><td>0.54</td><td>0.58</td><td>0.6</td><td>0.61</td><td>0.63</td><td>0.65</td><td>0.67</td><td>0.7</td><td>0.72</td><td>0.74</td><td /><td>0.73</td><td>0.72</td><td>0.67</td><td>0.64</td><td>c0.65</td></tr><tr><td>1940</td><td>0.42</td><td>0.44</td><td>0.45</td><td>0.5</td><td>0.52</td><td>0.55</td><td>0.57</td><td>0.58</td><td>0.59</td><td>0.61</td><td>0.63</td><td>0.67</td><td>0.68</td><td>0.71</td><td>0.73</td><td /><td>0.71</td><td>0.7</td><td>0.65</td><td>0.62</td></tr><tr><td>1950</td><td>0.43</td><td>0.41</td><td>0.43</td><td>0.45</td><td>0.49</td><td>0.54</td><td>0.54</td><td>0.55</td><td>0.57</td><td>0.58</td><td>0.61</td><td>0.64</td><td>0.66</td><td>0.67</td><td>0.69</td><td>0.71</td><td /><td>0.71</td><td>0.67</td><td>0.61</td></tr><tr><td>1960</td><td>0.41</td><td>0.41</td><td>0.39</td><td>0.43</td><td>0.44</td><td>0.5</td><td>0.52</td><td>0.52</td><td>0.54</td><td>0.55</td><td>0.56</td><td>0.61</td><td>0.63</td><td>0.66</td><td>0.66</td><td>0.68</td><td>0.7</td><td /><td>0.7</td><td>0.66</td></tr><tr><td>1970</td><td>0.44</td><td>0.41</td><td>0.42</td><td>0.41</td><td>0.44</td><td>0.48</td><td>0.51</td><td>0.52</td><td>0.51</td><td>0.52</td><td>0.55</td><td>0.57</td><td>0.6</td><td>0.62</td><td>0.65</td><td>0.64</td><td>0.63</td><td>0.68</td><td /><td>0.67</td></tr><tr><td>1980</td><td>0.41</td><td>0.43</td><td>0.41</td><td>0.44</td><td>0.4</td><td>0.47</td><td>0.5</td><td>0.51</td><td>0.53</td><td>0.51</td><td>0.52</td><td>0.57</td><td>0.59</td><td>0.6</td><td>0.63</td><td>0.62</td><td>0.56</td><td>0.6</td><td>0.64</td><td /></tr></tbody></table> </ephtml> </p> <p>1 <emph>Data sources:</emph> IPUMS United States Population Censuses 1850–2000 and American Community Survey (ACS) 2001–2018. See a summary of the data sources in Appendix Table A3 in the supplemental material.</p> <p>2 <emph>Note:</emph> The lower triangular part of the correlation matrix refers to a series of correlations between Treiman's ranks at times <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>t</mi></math> </ephtml> and <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>t</mi><mo>+</mo><mi>n</mi></math> </ephtml> . The upper triangular part of the correlation matrix refers to correlations between unsmoothed percentile ranks at times <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>t</mi></math> </ephtml> and <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>t</mi><mo>+</mo><mi>n</mi></math> </ephtml> . The shaded region refers to higher correlations between years in Treiman's ranks than in percentile ranks.</p> <p>Overall, the majority of cells below the diagonal are highlighted, suggesting that occupational statuses measured by Treiman's ranks are more stable over time than percentile ranks. For example, the correlation between Treiman's ranks of occupations for the birth cohorts 1790 and 1850 is 0.69, and the corresponding correlation in percentile ranks is 0.68. As we discussed earlier, this pattern is due to the fact that Treiman's ranks reflect only changes in the relative order of occupations, whereas percentile ranks are a function of both the relative order of occupations and relative sizes of occupations. However, we also note that most cells at the top left corner and the bottom right corner are not highlighted, deviating from the general pattern.</p> <p>For illustration, we provide a graphic summary of occupational percentile changes for some major occupations in Figures 1–4. Figure 1 shows a select group of 1950 Census occupations in the broad category of professional, technical, and managerial occupations. Most top-level professional occupations (codes 000–095), such as accountants, architects, lawyers, dentists, and scientists, have remained constant in their occupational percentile ranks. For example, dentists have been among one of the most prestigious occupations, with the percentile ranks in the range of 97.4–99.9 for birth cohorts between 1800 and 1980 (see Appendix Figure A1, supplemental material). By contrast, the percentile ranks of low-skilled technological occupations (codes 200–290), such as postmasters, purchasing agents, pilots and pursers, and railroad conductors, have declined. For example, the percentile ranks of railroad conductors used to be higher than 95 for cohorts born before 1880 but dropped to 68.5 for the 1890 birth cohort and further to 36.8 for the most recent 1980 cohort. A possible explanation is rising automation, with much work that used to be performed by humans being replaced by machines, programs, software, and robots and a corresponding massive increase in professional and technical occupations over time ([<reflink idref="bib4" id="ref172">4</reflink>]; [<reflink idref="bib66" id="ref173">66</reflink>]).</p> <p>Graph: Figure 1. Trends in Percentile Ranks across Birth Cohorts for a Select Group of Professional and Managerial Occupations. Data source : IPUMS United States Population Censuses full count 1850, 1860, 1870, 1880, 1900, 1910, 1920, 1930, 1940; 1% sample, 1950; 5% samples, 1960, 1980, 1990, 2000; 6% sample, 1970; American Community Survey (ACS) 2001–2018. Notes : The numbers in the subtitles refer to 1950 U.S. Census occupational codes. See Appendix Figure A1 in the supplemental material for the full list of occupations. The shaded areas refer to 95% confidence bounds constructed from 1000-bootstrap samples.</p> <p>Graph: Figure 2. Trends in Percentile Ranks across Birth Cohorts for a Select Group of Clerical, Sales, and Service Occupations. Data source : IPUMS U.S. Full-Count Population Censuses 1850, 1860, 1870, 1880, 1900, 1910, 1920, 1930, 1940; 1% sample: 1950; 5% samples: 1960, 1980, 1990, 2000; 6% sample: 1970; American Community Survey (ACS) 2001–2018. Notes : The numbers in the subtitles refer to 1950 U.S. Census occupational codes. See Appendix Figure A1 in the supplemental material for the full list of occupations. The shaded areas refer to 95% confidence bounds constructed from 1000-bootstrap samples.</p> <p>Graph: Figure 3. Trends in Percentile Ranks across Birth Cohorts for a Select Group of Craftsmen and Operative Occupations. Data source : IPUMS U.S. Full-Count Population Censuses 1850, 1860, 1870, 1880, 1900, 1910, 1920, 1930, 1940; 1% sample: 1950; 5% samples: 1960, 1980, 1990, 2000; 6% sample: 1970; American Community Survey (ACS) 2001–2018. Notes : The numbers in the subtitles refer to 1950 U.S. Census occupational codes. See Appendix Figure A1 in the supplemental material for the full list of occupations. The shaded areas refer to 95% confidence bounds constructed from 1000-bootstrap samples.</p> <p>Graph: Figure 4. Trends in Percentile Ranks across Birth Cohorts for a Select Group of Farming Occupations. Data source : IPUMS U.S. Full-Count Population Censuses 1850, 1860, 1870, 1880, 1900, 1910, 1920, 1930, 1940; 1% sample: 1950; 5% samples: 1960, 1980, 1990, 2000; 6% sample: 1970; American Community Survey (ACS) 2001–2018. Notes : The numbers in the subtitles refer to 1950 U.S. Census occupational codes. See Appendix Figure A1 in the supplemental material for the full list of occupations. The shaded areas refer to 95% confidence bounds constructed from 1000-bootstrap samples.</p> <p>Figure 2 shows that percentile ranks of most occupations in the clerical, sales, and service occupations have trended downward, and this secular change has in some instances reordered the relative rankings of occupations, i.e., the Treiman's ranks. For example, the percentile ranks of bill and account collectors (code 321) used to be close to 99 for those born around the 1800s but dropped to the level of 38 for the most recent birth cohort. In contrast, the percentile rank of policemen and detectives (code 773) was 97 for the 1800 birth cohort, lower than that of bill and account collectors, but experienced less decline and stabilized around the 73rd percentile for the recent cohorts. Several occupations stand out as outliers: The status of housekeepers and stewards (code 764) increased from the 31st percentile for the 1800 cohort to the 86th percentile for the 1860 cohort, declined for the 1870–1920 cohorts, and increased again for the 1930–1980 cohorts. Waiters and waitresses (code 784) show a similar N-shaped trend, with their occupational status first increasing from the 11th percentile to the 65th percentile for cohorts between 1800 and 1870, then declining to the 14th percentile for the 1940 cohort, and rising to the 34th percentile for the most recent cohort. Cooks (code 754) used to be one of the lowest-status occupations, ranked below the 1st percentile for the 1800 cohort, but its status grew dramatically in subsequent cohorts and peaked at the 49th percentile for the 1890 cohort. Even though its occupational status dropped for cohorts born after 1900, its percentile rank for the 1980 cohort, at 15th, is still higher than that for the 1800 cohort.</p> <p>Given the relative nature of our measure, an occupation would decline in social status when its size increases, or when sizes of higher-ranked occupations increase. For example, the percentile score of "clerical and kindred workers" (code 390 in Figure 2) was as high as 94.4 for the 1860 birth cohort but declined to 45 for the 1980 birth cohort, reflecting both the increase of its occupation size as well as the size increases of higher-ranked occupations (e.g., professional and technical occupations coded 0–99). This feature of our measure is in sharp contrast to what is commonly known as the "Treiman constant" ([<reflink idref="bib62" id="ref174">62</reflink>]), meaning that occupational prestige more or less stays the same over time and across societies. In Figure A5 in the supplemental material, we compare trends in occupational percentile ranks and Treiman's ranks (rescaled to lie between 0 and 100) across birth cohorts based on 1950 occupational codes. As shown in the figure, our percentile ranks depict more temporal fluctuations than Treiman's ranks, as the former accounts for shifts in both the education associated with each occupation and the relative sizes of the occupations.</p> <p>Figure 3 reveals more substantial long-term declines in occupational percentile ranks among craftsmen and operative occupations. The occupational statuses of bookbinders (code 502), boilermakers (code 503), compositors and typesetters (code 512), machinists (code 544), and painters and maintenance workers (code 564) all declined from higher than the 90th percentile in cohorts earlier than 1900 to lower than the 37th percentile for the 1980 birth cohort. This trend results from occupational upgrading in general and technological innovation in these occupations ([<reflink idref="bib11" id="ref175">11</reflink>]). The introduction of sophisticated machinery, accompanied by the decline in industrial profit margins, routinized job tasks and deskilled many blue-collar craftsmen. Using compositors and typesetters in the printing industry as an example, [<reflink idref="bib108" id="ref176">108</reflink>]) argued that "deskilling" meant that many traditional skills, such as the judgment of operators in setting linotype machines, became antiquated in the modern composing room, and the advent of teletypesetting technology also significantly diminished the training time for apprentice jobs. As a result, skill levels for many craftsmen and operative occupations declined, leading to the erosion of the social statuses of craftsmen occupations.</p> <p>Figure 4 shows the trends for agricultural occupations. The social statuses of farmers and farm laborers have been consistently low throughout history. The occupational percentiles of farmers (code 100) dropped from the 43rd percentile for the 1810 cohort to the 36th percentile for the 1980 cohort. The percentiles of farm laborers (code 820) changed from the 6th to the 1st percentile over the same period. Part of this decline was driven by the changing size of the agricultural sector: The farming population accounted for 51 percent of the total labor force in the 1810 birth cohort but only 1 percent in the 1980 cohort. The statuses of farm managers (code 123) and farm foremen (code 810) first increased and then decreased over time, suggesting a possible skill upgrading and then downgrading in the course of industrialization and technological improvement.</p> <p>The figures illustrate a general trend: declining occupational percentile ranks among routine and manual occupations and relative stability for occupations at either the top or the bottom of the social hierarchy. Over time, relative occupational statuses could decline for two potential reasons. First, differential educational upgrading across occupations could lead to declining statuses of some occupations relative to others. For example, the educational ranking of sports instructors and officials (code 91) used to be very low, only ranking 17 out of the total 222 occupations for the 1800 cohort, but the rank improved steadily over time and reached 134 for the 1980 cohort. By contrast, the educational rank of surveyors (code 92) was higher than that of sports instructors for the 1800–1880 cohorts but fell behind afterward, falling to 117 for the 1980 cohort.[<reflink idref="bib31" id="ref177">31</reflink>] Second, the relative status of an occupation declines as higher-status occupations expand. The proportion of professional, technical, and managerial occupations (codes 0–290) excepting farmers (code 100) and farm managers (code 123) accounts for 9.1 percent of the whole labor force for the 1800 birth cohort and increases to 18.1 percent for the 1900 birth cohort and further to around 40 percent for birth cohorts after 1950. The increasing share of workers in these top-ranked occupations would lead to a fall in the ranks of middle- and lower-status occupations, such as those in services and manufacturing, and a widening in status distances between occupations ranked at different levels.[<reflink idref="bib32" id="ref178">32</reflink>]</p> <p>In Appendix Figure A1 in the supplemental material, we present the full list of percentile ranks by occupation and birth cohort using the 1950 Census classification. Appendix Figures A2 and A3 in the supplemental material provide percentile ranks by occupation and birth cohort using 1990 and 2010 Census occupational classification schemes. Because historical data are not coded in 1990 and 2010 occupations, earlier cohorts are not covered in the series presented in Appendix Figures A2 and A3 in the supplemental material.[<reflink idref="bib33" id="ref179">33</reflink>]</p> <p>How do the newly constructed occupational percentile ranks compare to other standard occupational measures over the long term? In Table 2, we present estimated correlations of the percentile ranks with the Duncan SEI, Hauser-Warren SEI, Siegel prestige scores, Nakao-Treas prestige scores, and Nam-Powers-Boyd occupational scores.[<reflink idref="bib34" id="ref180">34</reflink>] Note that our percentile ranks are occupation- and cohort-specific, whereas the other indexes, except for the Nam-Powers-Boyd scores, are assumed to be constant over time within occupations. Overall, our percentile ranks reveal a stronger correlation with socioeconomic indexes than with prestige scores. This result is not surprising, as we created the percentile ranks based on occupational education, which is also a core component of socioeconomic index measures. Our percentile ranks are also strongly correlated with the Nam-Powers-Boyd occupational scores derived from the average occupational income percentiles and occupational education percentiles for years after 1950.</p> <p>Table 2. Correlations ( <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>ρ</mi></math> </ephtml> ) between Different Socioeconomic Status Measures with Occupational Percentile Ranks by Birth Cohort</p> <p>Graph</p> <p> <ephtml> <table><colgroup><col align="left" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /><col align="center" /></colgroup><thead><tr><th align="left" /><th align="center" colspan="5">Correlations between percentile ranks and</th></tr><tr><th align="left">Birth</th><th align="center">Duncan</th><th align="center">Hauser-Warren</th><th align="center">Siegel prestige</th><th align="center">Nakao-Treas</th><th align="center">Nam-Powers-Boyd</th></tr><tr><th align="center">Cohort</th><th align="center">SEI</th><th align="center">SEI</th><th align="center">score</th><th align="center">Prestige score</th><th align="center">status score</th></tr></thead><tbody><tr><td>1790</td><td>0.483</td><td>0.368</td><td>0.407</td><td>0.356</td><td>0.431</td></tr><tr><td>1800</td><td>0.506</td><td>0.438</td><td>0.462</td><td>0.375</td><td>0.531</td></tr><tr><td>1810</td><td>0.528</td><td>0.403</td><td>0.461</td><td>0.359</td><td>0.535</td></tr><tr><td>1820</td><td>0.616</td><td>0.492</td><td>0.531</td><td>0.439</td><td>0.594</td></tr><tr><td>1830</td><td>0.584</td><td>0.459</td><td>0.495</td><td>0.403</td><td>0.575</td></tr><tr><td>1840</td><td>0.606</td><td>0.501</td><td>0.540</td><td>0.446</td><td>0.612</td></tr><tr><td>1850</td><td>0.702</td><td>0.588</td><td>0.641</td><td>0.547</td><td>0.695</td></tr><tr><td>1860</td><td>0.737</td><td>0.613</td><td>0.639</td><td>0.566</td><td>0.717</td></tr><tr><td>1870</td><td>0.772</td><td>0.648</td><td>0.668</td><td>0.595</td><td>0.724</td></tr><tr><td>1880</td><td>0.818</td><td>0.716</td><td>0.715</td><td>0.658</td><td>0.753</td></tr><tr><td>1890</td><td>0.842</td><td>0.752</td><td>0.746</td><td>0.709</td><td>0.721</td></tr><tr><td>1900</td><td>0.902</td><td>0.815</td><td>0.824</td><td>0.783</td><td>0.803</td></tr><tr><td>1910</td><td>0.924</td><td>0.846</td><td>0.839</td><td>0.798</td><td>0.829</td></tr><tr><td>1920</td><td>0.923</td><td>0.853</td><td>0.839</td><td>0.805</td><td>0.853</td></tr><tr><td>1930</td><td>0.908</td><td>0.874</td><td>0.823</td><td>0.809</td><td>0.836</td></tr><tr><td>1940</td><td>0.871</td><td>0.883</td><td>0.795</td><td>0.804</td><td>0.834</td></tr><tr><td>1950</td><td>0.878</td><td>0.916</td><td>0.811</td><td>0.819</td><td>0.870</td></tr><tr><td>1960</td><td>0.883</td><td>0.912</td><td>0.810</td><td>0.821</td><td>0.876</td></tr><tr><td>1970</td><td>0.875</td><td>0.900</td><td>0.802</td><td>0.800</td><td>0.879</td></tr><tr><td>1980</td><td>0.875</td><td>0.894</td><td>0.806</td><td>0.796</td><td>0.883</td></tr></tbody></table> </ephtml> </p> <p>3 <emph>Note:</emph> The five socioeconomic indexes and prestige scores are downloaded from the IPUMS USA project. The Duncan SEI was developed using the 1947 North-Hatt prestige scores and the 1950 Census data; the Hauser-Warren SEI was developed using 1990 Census data; the Siegel prestige scores were developed using 1963, 1964, and 1965 NORC prestige scores and 1960 Census data; the Nakao-Treas prestige scores were developed using the 1989 National Opinion Research Center(NORC) General Social Survey prestige scores and 1980 Census data; and the IPUMS Nam-Powers-Boyd status scores were estimated using the 1950 Census data. Detailed descriptions of these indexes can be found in the section on Socioeconomic Status Indexes and Scales: An Overview and Appendix Table A2 in the supplemental material. <emph>Data sources:</emph> IPUMS United States Population Censuses 1850–2000 and ACS 2001–2018. See a summary of the data sources in Appendix Table A3 in the supplemental material.</p> <p>We are particularly interested in changes in the correlations over time. The strongest correlations between the Duncan SEI and our percentile rank scores are observed for birth cohorts 1910 and 1920 ( <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>ρ</mi><mo>=</mo><mn>0.92</mn></math> </ephtml> ). This is because the Duncan SEI was developed using the 1950 Census data, in which most prime-age workers were born around the 1910s and 1920s. While generally lower than those between percentile ranks and the Duncan SEI, the highest correlations between percentile ranks and the Siegel scores are observed also for the 1910 and 1920 birth cohorts ( <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>ρ</mi><mo>=</mo><mn>0.84</mn></math> </ephtml> ).[<reflink idref="bib35" id="ref181">35</reflink>] The IPUMS Nakao-Treas prestige scores and the Hauser-Warren SEI were both developed using the 1990 Census occupational scheme as the basis.[<reflink idref="bib36" id="ref182">36</reflink>] The strongest correlations between these indexes and the percentile ranks emerge for the 1950 and 1960 birth cohorts, who were in their thirties and forties in the Census year 1990 ( <ephtml> <math display="inline" xmlns="http://www.w3.org/1998/Math/MathML"><mi>ρ</mi><mo>=</mo><mn>0.91</mn></math> </ephtml> for the Hauser-Warren SEI and 0.82 for the Nakao-Treas prestige score). The correlations between the percentile ranks and the other occupational status measures are smallest for early birth cohorts. For example, the correlation between percentile ranks and the Hauser-Warren SEI is 0.37 for the 1790 birth cohort, only 40 percent as large as the correlation for the 1950 birth cohort. These results reaffirm our research motivation that relative occupational statuses have changed substantially over a longer-term span.</p> <p>As we discussed earlier, the Nam-Powers-Boyd occupational score is a percentile rank measure, indicating the percentage of workers in occupations that are lower in status, as measured by both education and earnings. The IPUMS Nam-Powers-Boyd score was constructed by combining median occupational education and median occupational earnings weighted by the size of each occupation. Due to data limitations, the 1950 scores were assigned to data in pre-1950 Censuses, and the scores are updated only for data after 1950.[<reflink idref="bib37" id="ref183">37</reflink>] The results show an increase in the association between the Nam-Powers-Boyd percentile scores and our occupational percentile ranks by birth cohort.</p> <hd id="AN0184233957-19">Conclusion</hd> <p>This paper consists of two major components. First, we reviewed qualitative and quantitative occupational measures developed in the sociological literature for data from preindustrial or early-industrial times to the present. We note that most of them were developed for a specific population, period, or research setting. Thus, it is difficult to compare historical measures across studies or use the same measures to study trends over time or across societies. Changes in occupational structure present major challenges for constructing occupation-based measures that would maintain comparability over a long historical period. The arrival of new technologies created many occupations while making some others obsolete along the way. Even for occupations that endured through time, the population size and composition of workers as well as their skills may have changed substantially.</p> <p>We provided a systematic review of the best-known occupational indexes in the literature. These measures consist of four broad groups: Occupational prestige scores, socioeconomic indexes, percentile scores, and job desirability indexes. Most of these measures were either developed for a single Census year or a short span of history. These indexes, however, do not work well when researchers wish to examine long-term changes over many decades or even more than a century, as the relative ordering of occupational statuses changes substantially over time due to differential educational upgrading and expansion or contraction in occupational sizes. It is time to reevaluate the comparability of occupations and their measures across time, given the rapid growth of harmonized or linked historic and modern administrative and survey data.</p> <p>Second, this paper introduced a new percentile-rank-based socioeconomic status scale constructed from U.S. Census microdata from 1850 to 2000 and ACSs from 2001 to 2018. With this new measure, researchers can measure and compare occupational statuses for workers born in different birth cohorts. It has been long accepted in sociology, either explicitly or implicitly, that there is a universal occupational status system cross-nationally and across time. Although the occupational hierarchy is remarkably stable, as suggested by the "Treiman constant," this does not mean that a worker's social status in a given occupation is invariant. In our view, social status is best viewed in relative terms when research attention is focused on social inequality, hierarchy, or status differentiation among individuals in a given population. In social status, one person's gain is another person's loss. For this reason, even when an occupation's absolute status does not change, the relative status of workers in the occupation may change as a result of such structural changes as occupational expansion or contraction. Our analysis of the past 150 years in the United States suggests that the statuses of most occupations have declined despite the stability of the relative ordering of occupations. The distances between occupations evolve as the numbers of incumbents in different occupations rise and fall over time. Our new occupation-based socioeconomic index can capture such temporal changes in occupational statuses due to changes in occupational sizes and compositions. Moreover, occupational percentile ranks based on the 1950, 1990, and 2010 Census occupational schemes developed in this paper can be easily merged with social surveys and administrative data that include occupational measures based on Census occupation codes or crosswalks.</p> <p>We present the percentile rank measure as but one potential measure that researchers use, not to the exclusion of alternative ones. We thus would caution the reader that our occupational percentile index is meant to supplement, rather than replace, traditional socioeconomic status measures such as the Duncan ([<reflink idref="bib24" id="ref184">24</reflink>]) SEI, Stevens-Featherman ([<reflink idref="bib91" id="ref185">91</reflink>]) SEI, Stevens-Cho ([<reflink idref="bib90" id="ref186">90</reflink>]) SEI, Hauser-Warren ([<reflink idref="bib52" id="ref187">52</reflink>]) SEI, Siegel ([<reflink idref="bib84" id="ref188">84</reflink>]) prestige scores, Nakao-Treas ([<reflink idref="bib73" id="ref189">73</reflink>]) prestige scores, and Nam-Powers-Boyd ([<reflink idref="bib75" id="ref190">75</reflink>]; [<reflink idref="bib76" id="ref191">76</reflink>]; [<reflink idref="bib74" id="ref192">74</reflink>]) occupational percentile scores. The earlier prestige scores and occupational indexes remain useful in research settings when the time horizon under study is relatively short. The O*NET database also offers comprehensive occupational measures for data after 2000. When multiple measures are available, choice among them should be made judiciously on substantive grounds. [<reflink idref="bib111" id="ref193">111</reflink>]) compared 15 different occupational status measures and showed that analyses of occupational stratification might be sensitive to the choice of measures. We summarize differences between these socioeconomic status measures in Appendix Table A2 in the supplemental material.</p> <p>In his last and self-assessed "best book" ([<reflink idref="bib117" id="ref194">117</reflink>]), <emph>Notes on Social Measurement</emph>, [<reflink idref="bib26" id="ref195">26</reflink>]) pointed out the inherent limitations of measurement in sociology: "But sociology is not like physics. Nothing but physics is like physics" ([<reflink idref="bib26" id="ref196">26</reflink>]:169). According to Duncan, all social measurements suffer from certain limitations. Our percentile measure is no exception. For example, our single-number measure cannot possibly capture innumerable dimensions underlying occupations. Furthermore, normalized by a specific cohort, our measure cannot be used to study true secular changes across cohorts in social status. Finally, our measure is crude, applicable at the level of Census occupation codes, necessarily glossing over job-level, group-level, regional-level, and individual-level heterogeneity. However, despite these limitations, our percentile rank measure has the advantages of being simple, easily interpretable, widely applicable, and truly relative—reasons for their use in both sociological and economic literatures studying historical trends in social mobility ([<reflink idref="bib88" id="ref197">88</reflink>]; [<reflink idref="bib109" id="ref198">109</reflink>]). Our study introduces an additional tool with which social science researchers can analyze occupational change and status mobility. Yet, no one occupational measure satisfies all research purposes in view of the complexity of the social world. It is ultimately the task of researchers, given substantive research objectives, to select appropriate occupational constructs and ensure that research findings are not confounded by the choice of problematic metrics.</p> <hd id="AN0184233957-20">Supplemental Material</hd> <p>Graph: Supplemental material, sj-pdf-1-smr-10.1177_00491241231207914 for Occupational Percentile Rank: A New Method for Constructing a Socioeconomic Index of Occupational Status by Xi Song and Yu Xie in Sociological Methods & Research</p> <hd id="AN0184233957-21">Acknowledgments</hd> <p>The authors thank Donald Treiman and Rebecca Emigh for comments on earlier versions of the paper; Hao Dong, Joseph Ferrie, Catherine Massey, Jonathan Rothbaum, and Karen Rolf for helpful discussions; Lemeng Liang for excellent research assistance; and Jiahui Xu for helping with the website and statistical packages.</p> <ref id="AN0184233957-22"> <title> References </title> <blist> <bibl id="bib1" idref="ref10" type="bt">1</bibl> <bibtext> Anderson W. A. 1927. " The Occupational Attitudes and Choices of a Group of College Men. Part I." Social Forces. 6:278‐83.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref20" type="bt">2</bibl> <bibtext> Anderson W. A. 1928. " The Occupational Attitudes and Choice of a Group of College Men. Part II." Social Forces. 6:467‐73.</bibtext> </blist> <blist> <bibl id="bib3" idref="ref143" type="bt">3</bibl> <bibtext> Anderson W. A. 1934. " The Occupational Attitudes of College Men." The Journal of Social Psychology. 5:435‐66.</bibtext> </blist> <blist> <bibl id="bib4" idref="ref172" type="bt">4</bibl> <bibtext> Autor David. 2019. "Work of the Past, Work of the Future." National Bureau of Economic Research working paper.</bibtext> </blist> <blist> <bibl id="bib5" idref="ref1" type="bt">5</bibl> <bibtext> Blau Peter M., Duncan Otis Dudley. 1967. The American Occupational Structure. New York: Wiley.</bibtext> </blist> <blist> <bibl id="bib6" idref="ref34" type="bt">6</bibl> <bibtext> Blishen Bernard R. 1958. " The Construction and Use of An Occupational Class Scale." Canadian Journal of Economics and Political Science. 24:519‐31.</bibtext> </blist> <blist> <bibl id="bib7" idref="ref40" type="bt">7</bibl> <bibtext> Blishen Bernard R., Carroll William K.1978. " Sex Differences in a Socioeconomic Index for Occupations in Canada." Canadian Review of Sociology and Anthropology. 15:352‐71.</bibtext> </blist> <blist> <bibl id="bib8" idref="ref51" type="bt">8</bibl> <bibtext> Blishen Bernard R., McRoberts Hugh A.1976. " A Revised Socioeconomic Index for Occupations in Canada." Canadian Review of Sociology and Anthropology. 13:71‐9.</bibtext> </blist> <blist> <bibl id="bib9" idref="ref61" type="bt">9</bibl> <bibtext> Bose Christine E., Rossi Peter H. 1983. " Gender and Jobs: Prestige Standings of Occupations As Affected by Gender." American Sociological Review. 48:316‐30.</bibtext> </blist> <blist> <bibtext> Boyd Monica. 1986. " Socioeconomic Indicies and Sexual Inequality: A Tale of Scales." Canadian Review of Sociology and Anthropology. 23:457‐80.</bibtext> </blist> <blist> <bibtext> Braverman Harry. 1974. Labor and Monopoly Capital: The Degradation of Work in the Twentieth Century. New York: Monthly Review Press.</bibtext> </blist> <blist> <bibtext> Buchmann Claudia, DiPrete Thomas A., McDaniel Anne. 2008. " Gender Inequalities in Education." Annual Review of Sociology. 34:319‐37.</bibtext> </blist> <blist> <bibtext> Cain Pamela, Treiman Donald J. 1981. " The Dictionary of Occupational Titles As a Source of Occupational Data." American Sociological Review. 46:253‐78.</bibtext> </blist> <blist> <bibtext> Caplow Theodore. 1954. The Sociology of Work. Minneapolis: University of Minnesota Press.</bibtext> </blist> <blist> <bibtext> Cattell Raymond B. 1942. " The Concept of Social Status." The Journal of Social Psychology. 15:293‐308.</bibtext> </blist> <blist> <bibtext> Clark Robert E. 1948. " The Relationship of Schizophrenia to Occupational Income and Prestige." American Sociological Review. 3:325‐30.</bibtext> </blist> <blist> <bibtext> COLLABU.S. Department of Labor. 1949. Dictionary of Occupational Titles Volume 1. 2nd ed. Washington D.C.: U.S. Government Printing Office.</bibtext> </blist> <blist> <bibtext> Conk Margo A. 1978. " Occupational Classification in the United States: 1870–1940." Journal of Interdisciplinary History. 9:111‐30.</bibtext> </blist> <blist> <bibtext> Coutu Walter. 1936. " The Relative Prestige of Twenty Professions As Judged by Three Groups of Professional Students." Social Forces. 14:522‐9.</bibtext> </blist> <blist> <bibtext> Davies A. F. 1952. " Prestige of Occupations." British Journal of Sociology. 3:134‐47.</bibtext> </blist> <blist> <bibtext> Davis Jerome. 1927. " Testing the Social Attitudes of Children in the Government Schools in Russia." American Journal of Sociology. 32:947‐52.</bibtext> </blist> <blist> <bibtext> Davis Kingsley, Moore Wilbert E.1945. " Some Principles of Stratification." American Sociological Review. 10:242‐9.</bibtext> </blist> <blist> <bibtext> Davis James A., Smith Tom W.1991. General Social Surveys, 1972–1990: Cumulative Codebook. Chicago: National Opinion Research Center.</bibtext> </blist> <blist> <bibtext> Duncan Otis Dudley. 1961. "A Socioeconomic Index for All Occupations' and 'Properties and Characteristics of the Socioeconomic Index'." Pp. 109-138 and 139-161 in Occupations and Social Status, edited by Jr. Albert J. Reiss. New York: Free Press of Glencoe.</bibtext> </blist> <blist> <bibtext> Duncan Otis Dudley. 1968. "Social Stratification and Mobility: Problems in the Measurement of Trend." Pp. 675-719 in Indicators of Social Change, edited by Eleanor B. Sheldon and Wilbert E. Moore. New York: Russell Sage Foundation.</bibtext> </blist> <blist> <bibtext> Duncan Otis Dudley. 1984. Notes on Social Measurement: Historical and Critical. New York, NY: Russell Sage Foundation.</bibtext> </blist> <blist> <bibtext> Edwards Alba M. 1917. " Social-Economic Groups of the United States." American Statistical Association Quarterly. 15:643‐61.</bibtext> </blist> <blist> <bibtext> Edwards Alba M. 1933. " A Social-Economic Grouping of the Gainful Workers of the United States." Journal of the American Statistical Association. 28:377‐87.</bibtext> </blist> <blist> <bibtext> Edwards Alba M. 1938. A Social-Economic Grouping of the Gainful Workers of the United States. Washington, DC: U.S. Government Printing Office.</bibtext> </blist> <blist> <bibtext> Emigh Rebecca Jean, Riley Dylan, Ahmed Patricia. 2016. Changes in Censuses From Imperialist to Welfare States: How Societies and States Count. New York: Palgrave Macmillan.</bibtext> </blist> <blist> <bibtext> Erikson Robert, Goldthorpe John H., Portocarero Lucienne. 1979. " Intergenerational Class Mobility in Three Western European Societies: England, France and Sweden." The British Journal of Sociology. 30:415‐41.</bibtext> </blist> <blist> <bibtext> Featherman David L., Hauser Robert M. 1976. " Prestige Or Socioeconomic Scales in the Study of Occupational Achievement." Sociological Methods and Research. 4:403‐22.</bibtext> </blist> <blist> <bibtext> Featherman David L., Hauser Robert M. 1978. Opportunity and Change. New York, NY: Academic Press.</bibtext> </blist> <blist> <bibtext> Featherman David L., Sobel Michael, Dickens David. 1975. "A Manual for Coding Occupations and Industries into Detailed 1970 Categories and a Listing of 1970-basis Duncan Socioeconomic Scores." Technical report, Madison, Wisconsin: University of Wisconsin, Center for Demography and Ecology Working Paper, 75–1. (This material is reproduced in Robert M. Hauser and David L. Featherman (eds.), The Process of Stratification, New York, Academic Press, 1977.).</bibtext> </blist> <blist> <bibtext> Fox John. 2016. Applied Regression Analysis and Generalized Linear Models. Thousand Oaks, CA: Sage Publications.</bibtext> </blist> <blist> <bibtext> Fox John, Suschnigg Carole. 1989. " A Note on Gender and the Prestige of Occupations." Canadian Journal of Sociology. 14:353‐60.</bibtext> </blist> <blist> <bibtext> Ganzeboom Harry B. G., De Graaf Paul M., Treiman Donald J. 1992. " A Standard International Socio-Economic Index of Occupational Status." Social Science Research. 21:1‐56.</bibtext> </blist> <blist> <bibtext> Ganzeboom Harry B. G., Treiman Donald J. 1996. " Internationally Comparable Measures of Occupational Status for the 1988 International Standard Classification of Occupations." Social Science Research. 25:201‐39.</bibtext> </blist> <blist> <bibtext> Ganzeboom Harry B. G, Treiman Donald J. 2003. "Three Internationally Standardized Measures for Comparative Research on Occupational Status." Pp. 159-193. in Advances in Cross-Cultural Harmonisation. A European Working Book for Demographic and Socio-Economic Variables, edited by Christof Wolf and Jurgen H.P. Hoffmeyer-Zlotnik. London: Kluwer Academic Press.</bibtext> </blist> <blist> <bibtext> Glass David V (ed.). 1954. Social Mobility in Britain. London, UK: Routledge & Kegan Paul.</bibtext> </blist> <blist> <bibtext> Goldberger Arthur S. 1989. " Economic and Mechanical Models of Intergenerational Transmission." The American Economic Review. 79:504‐13.</bibtext> </blist> <blist> <bibtext> Goldthorpe John H., Hope Keith. 1974. The Social Grading of Occupations. Oxford, UK: Clarendon.</bibtext> </blist> <blist> <bibtext> Gross Edward. 1959. " The Occupational Variable As a Research Category." American Sociological Review. 24:640‐9.</bibtext> </blist> <blist> <bibtext> Hall John, Jones Caradog B. 1950. " Social Grading of Occupations." British Journal of Sociology. 1:31‐55.</bibtext> </blist> <blist> <bibtext> Haller Archibald O., Lewis David M. 1966. " The Hypothesis of Intersocietal Similarity in Occupational Prestige Hierarchies." American Journal of Sociology. 72:210‐6.</bibtext> </blist> <blist> <bibtext> Handel Michael J. 2016. " The O*NET Content Model: Strengths and Limitations." Journal for Labour Market Research. 49:157‐76.</bibtext> </blist> <blist> <bibtext> Hatt Paul K. 1950. " Occupations and Social Stratification." American Journal of Sociology. 55:533‐43.</bibtext> </blist> <blist> <bibtext> Haug Marie R. 1977. " Measurement in Social Stratification." Annual Review of Sociology. 3:51‐77.</bibtext> </blist> <blist> <bibtext> Hauser Robert M. 1980. " Some Exploratory Methods for Modeling Mobility Tables and Other Cross-Classified Data." Sociological Methodology. 11:413‐58.</bibtext> </blist> <blist> <bibtext> Hauser Robert M. 1982. " Occupational Status in the Nineteenth and Twentieth Centuries." Historical Methods: A Journal of Quantitative and Interdisciplinary History. 15:111‐26.</bibtext> </blist> <blist> <bibtext> Hauser Robert M., Featherman David L. 1977. The Process of Stratification: Trends and Analyses. New York: Academic Press.</bibtext> </blist> <blist> <bibtext> Hauser Robert M., Warren John Robert. 1997. " Socioeconomic Indexes for Occupations: A Review, Update, and Critique." Sociological Methodology. 27:177‐298.</bibtext> </blist> <blist> <bibtext> Hauser Robert M., Warren John Robert, Huang Min-Hsiung, Carter Wendy Y. 2000. "Occupational Status, Education, and Social Mobility in the Meritocracy." Pp. 179–229 in Meritocracy and Economic Inequality, edited by Arrow K., S. Bowles, and S. Durlauf. Princeton, NJ: Princeton University Press.</bibtext> </blist> <blist> <bibtext> Hodge Robert W. 1961. Occupational Composition and Status Crystallization. Master's thesis, University of Chicago.</bibtext> </blist> <blist> <bibtext> Hodge Robert W. 1981. " The Measurement of Occupational Status." Social Science Research. 10:396‐415.</bibtext> </blist> <blist> <bibtext> Hodge Robert W., Siegel Paul M., Rossi Peter H. 1964. " Occupational Prestige in the United States, 1925–63." American Journal of Sociology. 70:286‐302.</bibtext> </blist> <blist> <bibtext> Hodge Robert W., Treiman Donald J., Rossi Peter H. 1966. "A Comparative Study of Occupational Prestige." Pp. 309-321 in Pp, edited by R. Bendix and S. Lipset. New York, NY: Free Press.</bibtext> </blist> <blist> <bibtext> Hollingshead August B. 1949. Elmtown's Youth. New York: John Wiley.</bibtext> </blist> <blist> <bibtext> Hollingshead August B. 1957. "Two Factor Index of Social Position." in New Haven. CT: Privately Printed.</bibtext> </blist> <blist> <bibtext> Hollingshead August B. 1975. "Four Factor Index of Social Status." Unpublished working paper, Department of Sociology, Yale University.</bibtext> </blist> <blist> <bibtext> Hollingshead August B., Redlich F. C. 1958. Social Class and Mental Illness. New York: John Wiley.</bibtext> </blist> <blist> <bibtext> Hout Michael, DiPrete Thomas A. 2006. " What We Have Learned: RC28's Contributions to Knowledge About Social Stratification." Research in Social Stratification and Mobility. 24:1‐20.</bibtext> </blist> <blist> <bibtext> Hout Michael, Smith Tom W., Marsden Peter V. 2015. "Prestige and Socioeconomic Scores for the 2010 Census Codes." in GSS Methodological Report, No. 124. Chicago: NORC.</bibtext> </blist> <blist> <bibtext> Hunt William C. 1897. " Workers At Gainful Occupations At the Federal Censuses of 1870, 1880, and 1890." Bulletin of the Department of Labor, No. 11:393‐433.</bibtext> </blist> <blist> <bibtext> Inkeles Alex, Rossi Peter H. 1956. " National Comparisons of Occupational Prestige." American Journal of Sociology. 61:329‐39.</bibtext> </blist> <blist> <bibtext> Jarvis Benjamin F., Song Xi. 2017. " Rising Intragenerational Occupational Mobility in the United States, 1969 to 2011." American Sociological Review. 82:568‐99.</bibtext> </blist> <blist> <bibtext> Jencks Christopher, Perman Lauri, Rainwater Lee. 1988. " What Is a Good Job? A New Measure of Labor-Market Success." American Journal of Sociology. 93:1322‐57.</bibtext> </blist> <blist> <bibtext> Jonsson Jan O., Grusky David B., Di Carlo Matthew, Pollak Reinhard, Brinton Mary C. 2009. " Microclass Mobility: Social Reproduction in Four Countries." American Journal of Sociology. 114:977‐1036.</bibtext> </blist> <blist> <bibtext> Kahl Joseph A. 1957. The American Class Structure. New York: Rinehart & Co.</bibtext> </blist> <blist> <bibtext> Kraus Vered, Schild Erling O., Hodge Robert W. 1978. " Occupational Prestige in the Collective Conscience." Social Forces. 56:900‐18.</bibtext> </blist> <blist> <bibtext> Lehman Harvey C., Witty Paul A. 1931. " Further Study of the Social Status of Occupations." The Journal of Educational Sociology. 5:101‐12.</bibtext> </blist> <blist> <bibtext> Lipset Seymour M., Bendix Reinhard. 1959. Social Mobility in Industrial Society. Berkeley, CA: University of California Press.</bibtext> </blist> <blist> <bibtext> Nakao Keiko, Treas Judith. 1994. " Updating Occupational Prestige and Socioeconomic Scores: How the New Measures Measure Up." Sociological Methodology. 1:1‐72.</bibtext> </blist> <blist> <bibtext> Nam Charles B., Boyd Monica. 2004. " Occupational Status in 2000: Over a Century of Census-Based Measurement." Population Research and Policy Review. 23:327‐58.</bibtext> </blist> <blist> <bibtext> Nam Charles B., Powers Mary G. 1968. " Changes in the Relative Status Level of Workers in the United States, 1950–1960." Social Forces. 47:158‐70.</bibtext> </blist> <blist> <bibtext> Nam Charles B., Powers Mary G. 1983. The Socioeconomic Approach to Status Measurement (With a Guide to Occupational and Socioeconomic Status Scores). Houston: Cap and Gown Press.</bibtext> </blist> <blist> <bibtext> National Center for O* NET Development. 2000. Transitioning O* NET to the Standard Occupational Classification. Raleigh, North Carolina: National Center for O* NET Development (https://<ulink href="http://www.onetcenter.org/reports/TRreport˙xwk.html">www.onetcenter.org/reports/TRreport˙xwk.html</ulink>).</bibtext> </blist> <blist> <bibtext> North Cecil, Hatt Paul K. 1947. " Jobs and Occupations: A Popular Evaluation." Opinion News. 9:3‐13.</bibtext> </blist> <blist> <bibtext> Osgood C. E., Stagner R., 1941. " Analysis of a Prestige Frame of Reference by a Gradient Technique." Journal of Applied Psychology. 25:275‐90.</bibtext> </blist> <blist> <bibtext> Powers Mary G. 1982. "Measures of Socioeconomic Status: An Introduction." Pp. 1-28 in Measures of Socioeconomic Status: Current Issues, edited by M. G. Powers, Boulder, CO: Westview Press for the American Association for the Advancement of Science.</bibtext> </blist> <blist> <bibtext> Reiss Jr. Albert J (ed.). 1961. Occupations and Social Status. New York: The Free Press of Glencoe.</bibtext> </blist> <blist> <bibtext> Rossiter W. S. 1909. A Century of Population Growth: From the First Census of the United States to the Twelfth 1790–1900. Washington, DC: U.S. Government Printing Office.</bibtext> </blist> <blist> <bibtext> Ruggles Steven. 2014. " Big Microdata for Population Research." Demography. 51:287‐97.</bibtext> </blist> <blist> <bibtext> Siegel Paul S. 1971. Prestige in the American Occupational Structure. Ph.D. thesis, Department of Sociology, University of Chicago. March 1971.</bibtext> </blist> <blist> <bibtext> Smith Mapheus. 1943. " An Empirical Scale of Prestige Status of Occupations." American Sociological Review. 8:185‐92.</bibtext> </blist> <blist> <bibtext> Smith Tom W., Son Jaesok. 2014. "Measuring Occupational Prestige on the 2012 General Social Survey." in GSS Methodological Report, No. 122. NORC.</bibtext> </blist> <blist> <bibtext> Song Xi, Campbell Cameron D. 2017. " Genealogical Microdata and Their Significance for Social Science." Annual Review of Sociology. 43:75‐99.</bibtext> </blist> <blist> <bibtext> Song Xi, Massey Catherine G., Rolf Karen A., Ferrie Joseph P., Rothbaum Jonathan L., Xie Yu. 2020. " Long-Term Decline in Intergenerational Mobility in the United States Since the 1850s." Proceedings of the National Academy of Sciences. 117:251‐8.</bibtext> </blist> <blist> <bibtext> Sorokin Pitrim A. 1959[1927]. Social and Cultural Mobility. Glencoe, IL: Free Press.</bibtext> </blist> <blist> <bibtext> Stevens Gillian, Cho Joo Hyun. 1985. " Socioeconomic Indexes and the New 1980 Census Occupational Classification Scheme." Social Science Research. 14:142‐68.</bibtext> </blist> <blist> <bibtext> Stevens Gillian, Featherman David L.1981. " A Revised Socioeconomic Index of Occupational Status." Social Science Research. 10:364‐95.</bibtext> </blist> <blist> <bibtext> Stevens Gillian, Hoisington Elizabeth. 1987. " Occupational Prestige and the 1980 U.S. Labor Force." Social Science Research. 16:74‐105.</bibtext> </blist> <blist> <bibtext> Svalastoga Kaare. 1959. Prestige, Class, and Mobility. London, UK: Heinemann.</bibtext> </blist> <blist> <bibtext> Treas Judith, Tyree Andrea. 1979. " Prestige Versus Socioeconomic Status in the Attainment Processes of American Men and Women." Social Science Research. 8:201‐21.</bibtext> </blist> <blist> <bibtext> Treiman Donald J. 1975. " Problems of Concept and Measurement in the Comparative Study of Occupational Mobility." Social Science Research. 4:183‐230.</bibtext> </blist> <blist> <bibtext> Treiman Donald J. 1976. " A Standard Occupational Prestige Scale for Use with Historical Data." The Journal of Interdisciplinary History. 7:283‐304.</bibtext> </blist> <blist> <bibtext> Treiman Donald J. 1977. Occupational Prestige in Comparative Perspective. New York: Academic Press.</bibtext> </blist> <blist> <bibtext> U.S. Bureau of the Census. 1910. Thirteenth Census of the United States: Classified Index to Occupations. Population Division. Washington D.C.: U.S. Government Printing Office.</bibtext> </blist> <blist> <bibtext> U.S. Bureau of the Census. 1915. Index to Occupations: Alphabetical and Classified. Washington D.C.: U.S. Government Printing Office.</bibtext> </blist> <blist> <bibtext> U.S. Department of Labor. 1965. Dictionary of Occupational Titles Volume I. 3rd ed. Washington D.C.: U.S. Government Printing Office.</bibtext> </blist> <blist> <bibtext> U.S. Department of Labor. 1977. Dictionary of Occupational Titles. 4th ed. Washington D.C.: U.S. Government Printing Office.</bibtext> </blist> <blist> <bibtext> U.S. Department of Labor. 1991. Dictionary of Occupational Titles Volume 1. 4th Edition, Revised. Washington D.C.: U.S. Government Printing Office.</bibtext> </blist> <blist> <bibtext> U.S. Department of Labor. 1998. O*NET 98: Data Dictionary. Washington D.C.: U.S. Government Printing Office.</bibtext> </blist> <blist> <bibtext> U.S. Department of Labor. 1939a. Dictionary of Occupational Titles Part I. Washington D.C.: U.S. Government Printing Office.</bibtext> </blist> <blist> <bibtext> U.S. Department of Labor. 1939b. Dictionary of Occupational Titles, Part II: Group Arrangement of Occupational Titles and Codes. Washington D.C.: U.S. Government Printing Office.</bibtext> </blist> <blist> <bibtext> van Leeuwen Marco H. D., Maas Ineke. 2011. HISCLASS: A Historical International Social Class Scheme. Leuven, Belgium: Leuven University Press.</bibtext> </blist> <blist> <bibtext> van Leeuwen Marco H. D., Maas Ineke, Miles Andrew. 2002. HISCO: Historical International Standard Classification of Occupations. Leuven, Belgium: Leuven University Press.</bibtext> </blist> <blist> <bibtext> Wallace Michael, Kalleberg Arne L. 1982. " "Industrial Transformation and the Decline of Craft: The Decomposition of Skill in the Printing Industry 1931–1978." American Sociological Review. 47:307‐24.</bibtext> </blist> <blist> <bibtext> Ward Zachary. forthcoming. "Intergenerational Mobility in American History: Accounting for Race and Measurement Error." American Economic Review https://<ulink href="http://www.nber.org/papers/w29256">www.nber.org/papers/w29256</ulink>.</bibtext> </blist> <blist> <bibtext> Warner W. Lloyd, Meeker Marchia, Kenneth Eells. 1949. Social Class in America: A Manual of Procedure for the Measurement of Social Status. Chicago: Science Research Associates.</bibtext> </blist> <blist> <bibtext> Warren John Robert, Sheridan Jennifer T., Hauser Robert M. 1998. " Choosing a Measure of Occupational Standing: How Useful Are Composite Measures in Analyses of Gender Inequality in Occupational Attainment? " Sociological Methods and Research. 27:3‐76.</bibtext> </blist> <blist> <bibtext> Weeden Kim A., Grusky David B. 2005. " The Case for a New Class Map." American Journal of Sociology. 111:141‐212.</bibtext> </blist> <blist> <bibtext> Weeden Kim A., Grusky David B. 2012. " The Three Worlds of Inequality." American Journal of Sociology. 117:1723‐85.</bibtext> </blist> <blist> <bibtext> Wegener Bernd. 1992. " Concepts and Measurement of Prestige." Annual Review of Sociology. 18:253‐80.</bibtext> </blist> <blist> <bibtext> Welch M. K. 1950. " The Ranking of Occupations on the Basis of Social Status." Occupations; the vocational guidance journal. 27:237‐41.</bibtext> </blist> <blist> <bibtext> Wright Erik Olin. 1997. Class Counts: Comparative Studies in Class Analysis. New York: Cambridge University Press.</bibtext> </blist> <blist> <bibtext> Xie Yu. 2007. " Otis Dudley Duncan's Legacy: The Demographic Approach to Quantitative Reasoning in Social Science." Research in Social Stratification and Mobility. 25:141‐56.</bibtext> </blist> </ref> <ref id="AN0184233957-23"> <title> Footnotes </title> <blist> <bibtext> Occupational codes and percentile rank files presented in this paper can be downloaded from the project website https://occrank.shinyapps.io/occrank/ or the OSF website https://osf.io/x7rnw/.</bibtext> </blist> <blist> <bibtext> The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.</bibtext> </blist> <blist> <bibtext> The authors received no financial support for the research, authorship and/or publication of this article.</bibtext> </blist> <blist> <bibtext> Xi Song https://orcid.org/0000-0003-3463-045X Yu Xie https://orcid.org/0000-0002-3240-8620</bibtext> </blist> <blist> <bibtext> Supplemental material and Appendix for this article are available online.</bibtext> </blist> <blist> <bibtext> Structural changes can lead to significant differences between occupational classifications in historical and contemporary societies. Van Leeuwen, Maas, and Miles ([107]) created the HISCO, which provides a systematic basis of occupational titles and definitions for a variety of countries in the nineteenth and twentieth centuries. Specifically, they rely on the 1,506 occupational categories in the 1968 version of the ISCO scheme to derive the 1000 most frequent occupational titles from eight countries from the seventeenth to the twentieth centuries.</bibtext> </blist> <blist> <bibtext> In 1860, the occupational list included 584 categories, and both males and females aged 15 and above were asked to fill out a questionnaire (see a review in [64]). In 1870, 1880, and 1890, the number of occupations in the Census list was reduced to 338, 265, and 218, respectively.https://journals.sagepub.com/doi/suppl/10.1177/00491241231207914 includes a timeline that shows changes in the Census Bureau's occupational classifications.</bibtext> </blist> <blist> <bibtext> The DOT provides detailed descriptions of occupations with respect to the complexity of work performed, education and training time expected, aptitude, temperament, interests appropriate for the occupation, physical demands, and working conditions.</bibtext> </blist> <blist> <bibtext> The occupational groups include professionals; proprietors, managers, and officials (farmers [owners and tenants], wholesale and retail dealers, and other proprietors, managers, and officials); clerks and kindred workers; skilled workers and foremen; semi-skilled workers (semi-skilled workers in manufacturing, and other semi-skilled workers); and unskilled workers (farm laborers, factory and building construction laborers, other laborers, and servant classes).</bibtext> </blist> <blist> <bibtext> These classifications include (1) professional and high administrative, (2) managerial and executive, (3) inspectional, supervisory, and other nonmanual higher grade, (4) inspectional, supervisory, and other non-manual lower grade, (5) skilled manual and routine grades of nonmanual, (6) semi-skilled manual, and (7) unskilled manual.</bibtext> </blist> <blist> <bibtext> [37]) and [68]) have provided detailed discussions on the history of the categorical versus continuous approach to socioeconomic status measures.</bibtext> </blist> <blist> <bibtext> A few follow-up studies published between 1927 and 1942 have adopted this approach by asking high-school or college students and workers to rate different sets of occupations ([1], [2], [3]; [15]; [19]; [71]; [79]; [85]).</bibtext> </blist> <blist> <bibtext> Some studies also argued that the similarity in prestige scores between societies might be overestimated, as only a small and biased sample of occupational titles is translatable across societies ([45]).</bibtext> </blist> <blist> <bibtext> [59]) also proposed the idea of measuring occupational status using a composite index. The Hollingshead Index of Social Position (ISP) ranks occupations and education from 1 (high) to 7 (low) and creates a composite by weighting occupational scores by a factor of 7 and educational scores by a factor of 4. The resulting index in the range of 11 to –77 is then divided into five classes, with the highest class scoring from 11 to 17, and the lowest class scoring from 61 to 77.</bibtext> </blist> <blist> <bibtext> [24]) matched the occupational titles in the NORC occupational prestige survey with occupational titles in the 1950 Census classification. Many of the NORC titles were too specific, and only 45 out of the 90 NORC titles were reasonably equivalent to Census titles. Duncan also picked 16 occupations that were poorly matched to NORC titles and compared the NORC prestige ratings for these 16 titles with their predicted socioeconomic indexes. The results show a high correlation between these two measures with and without these additional occupations, suggesting good predictive power of the method using the original 45 occupations.</bibtext> </blist> <blist> <bibtext> As [24]) argued, given that the original NORC prestige score was created by "an arbitrary weighted summation procedure," he decided to explore an alternative measure that relies only on the raw data.</bibtext> </blist> <blist> <bibtext> These respondents include those who reported their level of education in the Census as "high school 4," "college 1 to 3," and "college 4 or more."</bibtext> </blist> <blist> <bibtext> Fox ([35]:287) shows that Duncan's original SEI estimates are likely to be sensitive to the presence of just two outlier observations, railroad conductors and ministers, in the 45 occupations included in the prestige regression analysis. Had these outliers been deleted, the coefficient of occupational income would become much larger than that of occupational education in Duncan's original SEI regression model.</bibtext> </blist> <blist> <bibtext> Occupational prestige is derived from subjective evaluations of the social standing of occupations, whereas occupational socioeconomic status reflects the objective socioeconomic compositions of workers in a particular occupation.</bibtext> </blist> <blist> <bibtext> The Hollingshead Two-Factor Index is a weighted sum of two ordinal scales from years of education (codes 1–7) and occupational status (codes 1–7) of household heads. The Three-Factor Index further includes the quality of the neighborhood (codes 1–6). The Four-Factor Index includes occupation, education, sex, and marital status based on information from both spouses in a household. The total score is either the individual score for unmarried persons or the average of both spouses for married couples.</bibtext> </blist> <blist> <bibtext> [76]) derived their occupational status scores for only civilian male workers 14 years of age and older, as comparable data for females were not available in published data tabulations.</bibtext> </blist> <blist> <bibtext> [75]) compared percentile scores of occupations in the 1950 and 1960 Censuses and found a high correlation coefficient of 0.96. They concluded that changing proportions of different occupations only lead to a slight decline in percentile scores. Yet they have never examined long-term change in the percentile score.</bibtext> </blist> <blist> <bibtext> We thank a reviewer for bringing the connection between IJD and O*NET to our attention.</bibtext> </blist> <blist> <bibtext> The year 1890 was missing because the original Census records were destroyed.</bibtext> </blist> <blist> <bibtext> For the year 1970, six 1-percent samples were drawn independently from the population data and the two 1-percent samples included in our analysis are known as Form 1 and Form 2.</bibtext> </blist> <blist> <bibtext> The documentation of these classifications can be found in "Integrated Occupation and Industry Codes and Occupational Standing Variables in the IPUMS" (https://usa.ipums.org/usa/chapter4/chapter4.shtml) and the "Alphabetic Index of Occupations and Industries: 1950" (https://usa.ipums.org/usa/resources/volii/Occupations1950.pdf). We acknowledge that using the 1950 codes for workers after 1970 in our index poses a limitation due to the significant changes in the occupation structure since 1950. However, we still calculate and release the index with the 1950 codes for later years, recognizing its potential usefulness in certain research settings.</bibtext> </blist> <blist> <bibtext> Occupations with codes above 970 are excluded from our analysis. These occupations include 979, not yet classified; 980, keeps house/housekeeping at home/housewife; 981, imputed keeping house (1850–1900); 982, helping at home/helps parents/housework; 983, at school/student; 984, retired; 985, unemployed/without occupation; 986, invalid/disabled with no occupation reported; 987, inmate; 990, new worker; 991, gentleman/lady/at leisure; 995, other nonoccupational; 997, occupation missing/unknown; and 999, N/A (blank).</bibtext> </blist> <blist> <bibtext> For example, OCC1950 does not contain detailed categories for Computer and Information Systems Managers, Computer Scientists and Systems Analysts/Network Systems Analysts/Web Developers, Computer Programmers that exist in OCC2010. All these 2010 categories were merged into categories for Professional, Technical and Kindred Workers (n.e.c.) in 1950.</bibtext> </blist> <blist> <bibtext> The selection criterion is based on the sample line weighted (SLWT) observations. According to IPUMS, SLWT should be used in analyses that rely on the "sample line" variables (e.g., education) from the 1940 and 1950 Censuses. The value of SLWT is zero for non-sample-line persons.</bibtext> </blist> <blist> <bibtext> [10]) showed that socioeconomic indexes derived from both men and women provides better results than those derived from male workers alone as the former better captures female disadvantages in the labor force.</bibtext> </blist> <blist> <bibtext> The results of Treiman's ranks of occupations for birth cohorts 1790–1980 are not included in the paper but are available upon request.</bibtext> </blist> <blist> <bibtext> The results of the relative sizes of occupations by birth cohort are not included in the paper but are available upon request.</bibtext> </blist> <blist> <bibtext> The original percentile data that produce these graphs will be posted on this project's website.</bibtext> </blist> <blist> <bibtext> The Duncan SEI is constructed by assigning the original Duncan SEI scores to each occupation using the IPUMS OCC1950 variable. The Siegel prestige scores are constructed by assigning the original Siegel prestige scores to each occupation using the IPUMS OCC1950 variable. More information on the construction of OCC1950 and occupational standing measures can be found in "Integrated Occupation and Industry Codes and Occupational Standing Variables in the IPUMS" (https://usa.ipums.org/usa/chapter4/chapter4.shtml).</bibtext> </blist> <blist> <bibtext> The original Siegel prestige scores were based on the 1960 Census occupations, which contained more occupational categories than those in the 1950 Census. The IPUMS project team aggregated the 1960 Census scheme to harmonize them with the 1950 Census scheme. When a 1960 occupation corresponds to several occupations in the 1950 classification scheme, the IPUMS data assigned the same 1960 prestige scores to their corresponding 1950 occupational categories. Conversely, when a 1950 occupation corresponds to several occupations in the 1960 classification scheme, the IPUMS data calculated the 1950 prestige score using the weighted average of the 1960 occupational categories based on the prestige score and number of observations within each 1960 category.</bibtext> </blist> <blist> <bibtext> The Hauser-Warren SEI is constructed by assigning the original Hauser-Warren SEI scores to each occupation using the IPUMS OCC1990 variable. The Nakao-Treas prestige scores are constructed by assigning the original Nakao-Treas prestige scores to each occupation using the IPUMS OCC1990 variable. The original Nakao-Treas scores were based on the 1980 Census occupations, which are similar to the 1990 Census occupations. The IPUMS project team aggregated several 1990 occupational categories to harmonize them with the 1980 Census scheme. When the 1990 scheme is more detailed than the 1980 scheme used by Nakao and Treas (1994), the IPUMS data assigned the same 1980 prestige scores to all of the 1990 occupational categories. Conversely, when the 1980 scheme is more detailed than the 1990 scheme, the IPUMS data calculated the 1990 score using the weighted average of the 1980 occupational categories based on the prestige score and number of observations within each 1980 category.</bibtext> </blist> <blist> <bibtext> The IPUMS Nam-Powers-Boyd occupational status score was constructed by combining median education and median earnings for each occupation in OCC1950, weighted by the size of each occupation. The score was calculated using year-specific earnings and education data after the 1950 Census. Years before 1950 were assigned the 1950 values.</bibtext> </blist> </ref> <aug> <p>By Xi Song and Yu Xie</p> <p>Reported by Author; Author</p> <p></p> <p>Xi Song is an associate professor of sociology and demography at the University of Pennsylvania. Her research interests include social mobility, occupations and work, population studies, and quantitative methodology. Her ongoing projects examine economic mobility through lifecycle and intergenerational processes, the evolution of occupational structure, new occupational ranking and autocoding methods, kinship network, and the link between intra– and intergenerational mobility.</p> <p>Yu Xie is Bert G. Kerstetter '66 University Professor of Sociology and has a faculty appointment at the Princeton Institute of International and Regional Studies, Princeton University. He is also a Visiting Chair Professor of the Center for Social Research, Peking University. His main areas of interest are social stratification, demography, statistical methods, Chinese studies, and sociology of science. His recently published works include: Marriage and Cohabitation (University of Chicago Press 2007) with Arland Thornton and William Axinn, Statistical Methods for Categorical Data Analysis with Daniel Powers (Emerald 2008, second edition), and Is American Science in Decline? (Harvard University Press, 2012) with Alexandra Killewald.</p> </aug> <nolink nlid="nl1" bibid="bib14" firstref="ref2"></nolink> <nolink nlid="nl2" bibid="bib33" firstref="ref3"></nolink> <nolink nlid="nl3" bibid="bib110" firstref="ref4"></nolink> <nolink nlid="nl4" bibid="bib43" firstref="ref5"></nolink> <nolink nlid="nl5" bibid="bib69" firstref="ref6"></nolink> <nolink nlid="nl6" bibid="bib81" firstref="ref7"></nolink> <nolink nlid="nl7" bibid="bib41" firstref="ref8"></nolink> <nolink nlid="nl8" bibid="bib53" firstref="ref9"></nolink> <nolink nlid="nl9" bibid="bib29" firstref="ref11"></nolink> <nolink nlid="nl10" bibid="bib31" firstref="ref12"></nolink> <nolink nlid="nl11" bibid="bib40" firstref="ref13"></nolink> <nolink nlid="nl12" bibid="bib49" firstref="ref14"></nolink> <nolink nlid="nl13" bibid="bib68" firstref="ref15"></nolink> <nolink nlid="nl14" bibid="bib72" firstref="ref16"></nolink> <nolink nlid="nl15" bibid="bib89" firstref="ref17"></nolink> <nolink nlid="nl16" bibid="bib112" firstref="ref18"></nolink> <nolink nlid="nl17" bibid="bib116" firstref="ref19"></nolink> <nolink nlid="nl18" bibid="bib21" firstref="ref21"></nolink> <nolink nlid="nl19" bibid="bib24" firstref="ref22"></nolink> <nolink nlid="nl20" bibid="bib37" firstref="ref23"></nolink> <nolink nlid="nl21" bibid="bib38" firstref="ref24"></nolink> <nolink nlid="nl22" bibid="bib51" firstref="ref25"></nolink> <nolink nlid="nl23" bibid="bib52" firstref="ref26"></nolink> <nolink nlid="nl24" bibid="bib56" firstref="ref27"></nolink> <nolink nlid="nl25" bibid="bib73" firstref="ref28"></nolink> <nolink nlid="nl26" bibid="bib76" firstref="ref29"></nolink> <nolink nlid="nl27" bibid="bib93" firstref="ref30"></nolink> <nolink nlid="nl28" bibid="bib97" firstref="ref31"></nolink> <nolink nlid="nl29" bibid="bib96" firstref="ref33"></nolink> <nolink nlid="nl30" bibid="bib106" firstref="ref35"></nolink> <nolink nlid="nl31" bibid="bib107" firstref="ref36"></nolink> <nolink nlid="nl32" bibid="bib18" firstref="ref38"></nolink> <nolink nlid="nl33" bibid="bib82" firstref="ref39"></nolink> <nolink nlid="nl34" bibid="bib98" firstref="ref41"></nolink> <nolink nlid="nl35" bibid="bib99" firstref="ref42"></nolink> <nolink nlid="nl36" bibid="bib104" firstref="ref43"></nolink> <nolink nlid="nl37" bibid="bib105" firstref="ref44"></nolink> <nolink nlid="nl38" bibid="bib17" firstref="ref45"></nolink> <nolink nlid="nl39" bibid="bib100" firstref="ref46"></nolink> <nolink nlid="nl40" bibid="bib101" firstref="ref47"></nolink> <nolink nlid="nl41" bibid="bib102" firstref="ref48"></nolink> <nolink nlid="nl42" bibid="bib13" firstref="ref49"></nolink> <nolink nlid="nl43" bibid="bib34" firstref="ref50"></nolink> <nolink nlid="nl44" bibid="bib103" firstref="ref52"></nolink> <nolink nlid="nl45" bibid="bib77" firstref="ref53"></nolink> <nolink nlid="nl46" bibid="bib39" firstref="ref55"></nolink> <nolink nlid="nl47" bibid="bib95" firstref="ref56"></nolink> <nolink nlid="nl48" bibid="bib27" firstref="ref58"></nolink> <nolink nlid="nl49" bibid="bib28" firstref="ref59"></nolink> <nolink nlid="nl50" bibid="bib30" firstref="ref60"></nolink> <nolink nlid="nl51" bibid="bib44" firstref="ref64"></nolink> <nolink nlid="nl52" bibid="bib10" firstref="ref65"></nolink> <nolink nlid="nl53" bibid="bib113" firstref="ref69"></nolink> <nolink nlid="nl54" bibid="bib22" firstref="ref71"></nolink> <nolink nlid="nl55" bibid="bib11" firstref="ref72"></nolink> <nolink nlid="nl56" bibid="bib12" firstref="ref74"></nolink> <nolink nlid="nl57" bibid="bib78" firstref="ref75"></nolink> <nolink nlid="nl58" bibid="bib86" firstref="ref76"></nolink> <nolink nlid="nl59" bibid="bib114" firstref="ref77"></nolink> <nolink nlid="nl60" bibid="bib84" firstref="ref78"></nolink> <nolink nlid="nl61" bibid="bib23" firstref="ref80"></nolink> <nolink nlid="nl62" bibid="bib63" firstref="ref81"></nolink> <nolink nlid="nl63" bibid="bib92" firstref="ref83"></nolink> <nolink nlid="nl64" bibid="bib15" firstref="ref86"></nolink> <nolink nlid="nl65" bibid="bib16" firstref="ref87"></nolink> <nolink nlid="nl66" bibid="bib115" firstref="ref88"></nolink> <nolink nlid="nl67" bibid="bib50" firstref="ref91"></nolink> <nolink nlid="nl68" bibid="bib65" firstref="ref92"></nolink> <nolink nlid="nl69" bibid="bib57" firstref="ref93"></nolink> <nolink nlid="nl70" bibid="bib62" firstref="ref96"></nolink> <nolink nlid="nl71" bibid="bib70" firstref="ref98"></nolink> <nolink nlid="nl72" bibid="bib20" firstref="ref99"></nolink> <nolink nlid="nl73" bibid="bib47" firstref="ref100"></nolink> <nolink nlid="nl74" bibid="bib91" firstref="ref109"></nolink> <nolink nlid="nl75" bibid="bib90" firstref="ref110"></nolink> <nolink nlid="nl76" bibid="bib32" firstref="ref114"></nolink> <nolink nlid="nl77" bibid="bib55" firstref="ref115"></nolink> <nolink nlid="nl78" bibid="bib94" firstref="ref116"></nolink> <nolink nlid="nl79" bibid="bib19" firstref="ref117"></nolink> <nolink nlid="nl80" bibid="bib58" firstref="ref121"></nolink> <nolink nlid="nl81" bibid="bib59" firstref="ref122"></nolink> <nolink nlid="nl82" bibid="bib60" firstref="ref123"></nolink> <nolink nlid="nl83" bibid="bib61" firstref="ref124"></nolink> <nolink nlid="nl84" bibid="bib75" firstref="ref126"></nolink> <nolink nlid="nl85" bibid="bib74" firstref="ref128"></nolink> <nolink nlid="nl86" bibid="bib25" firstref="ref132"></nolink> <nolink nlid="nl87" bibid="bib48" firstref="ref134"></nolink> <nolink nlid="nl88" bibid="bib80" firstref="ref135"></nolink> <nolink nlid="nl89" bibid="bib67" firstref="ref137"></nolink> <nolink nlid="nl90" bibid="bib42" firstref="ref138"></nolink> <nolink nlid="nl91" bibid="bib46" firstref="ref140"></nolink> <nolink nlid="nl92" bibid="bib54" firstref="ref146"></nolink> <nolink nlid="nl93" bibid="bib83" firstref="ref148"></nolink> <nolink nlid="nl94" bibid="bib87" firstref="ref149"></nolink> <nolink nlid="nl95" bibid="bib26" firstref="ref153"></nolink> <nolink nlid="nl96" bibid="bib111" firstref="ref161"></nolink> <nolink nlid="nl97" bibid="bib36" firstref="ref163"></nolink> <nolink nlid="nl98" bibid="bib66" firstref="ref173"></nolink> <nolink nlid="nl99" bibid="bib108" firstref="ref176"></nolink> <nolink nlid="nl100" bibid="bib35" firstref="ref181"></nolink> <nolink nlid="nl101" bibid="bib117" firstref="ref194"></nolink> <nolink nlid="nl102" bibid="bib88" firstref="ref197"></nolink> <nolink nlid="nl103" bibid="bib109" firstref="ref198"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1473582
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Occupational Percentile Rank: A New Method for Constructing a Socioeconomic Index of Occupational Status
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Xi+Song%22">Xi Song</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0003-3463-045X">0000-0003-3463-045X</externalLink>)<br /><searchLink fieldCode="AR" term="%22Yu+Xie%22">Yu Xie</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-3240-8620">0000-0002-3240-8620</externalLink>)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Sociological+Methods+%26+Research%22"><i>Sociological Methods & Research</i></searchLink>. 2025 54(2):355-396.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: SAGE Publications. 2455 Teller Road, Thousand Oaks, CA 91320. Tel: 800-818-7243; Tel: 805-499-9774; Fax: 800-583-2665; e-mail: journals@sagepub.com; Web site: https://sagepub.com
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 42
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2025
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Socioeconomic+Status%22">Socioeconomic Status</searchLink><br /><searchLink fieldCode="DE" term="%22Occupations%22">Occupations</searchLink><br /><searchLink fieldCode="DE" term="%22National+Surveys%22">National Surveys</searchLink><br /><searchLink fieldCode="DE" term="%22Statistical+Data%22">Statistical Data</searchLink><br /><searchLink fieldCode="DE" term="%22Social+Science+Research%22">Social Science Research</searchLink><br /><searchLink fieldCode="DE" term="%22Sociology%22">Sociology</searchLink>
– Name: SubjectThesaurus
  Label: Assessment and Survey Identifiers
  Group: Su
  Data: <searchLink fieldCode="SU" term="%22American+Community+Survey%22">American Community Survey</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1177/00491241231207914
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0049-1241<br />1552-8294
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: In this paper, we propose a method for constructing an occupation-based socioeconomic index that can easily incorporate changes in occupational structure. The resulting index is the occupational percentile rank for a given cohort, based on contemporaneous information pertaining to educational composition and the number of workers at the occupation level. An occupation may experience an increase or decrease in its occupational rank due to changes in relative sizes and educational compositions across occupations. The method is flexible in dealing with changes in occupational and educational measurements over time. Applying the method to U.S. history from the mid-nineteenth century to the present day, we derive the index using IPUMS U.S. Census microdata from 1850 to 2000 and the American Community Surveys (ACSs) from 2001 to 2018. Compared to previous occupational measures, this new measure takes into account occupational status evolvement caused by long-term secular changes in occupational size and educational composition. The resulting percentile rank measure can be easily merged with social surveys and administrative data that include occupational measures based on the U.S. Census occupation codes and crosswalks.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2025
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1473582
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1473582
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1177/00491241231207914
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 42
        StartPage: 355
    Subjects:
      – SubjectFull: Socioeconomic Status
        Type: general
      – SubjectFull: Occupations
        Type: general
      – SubjectFull: National Surveys
        Type: general
      – SubjectFull: Statistical Data
        Type: general
      – SubjectFull: Social Science Research
        Type: general
      – SubjectFull: Sociology
        Type: general
      – SubjectFull: American Community Survey
        Type: general
    Titles:
      – TitleFull: Occupational Percentile Rank: A New Method for Constructing a Socioeconomic Index of Occupational Status
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Xi Song
      – PersonEntity:
          Name:
            NameFull: Yu Xie
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 05
              Type: published
              Y: 2025
          Identifiers:
            – Type: issn-print
              Value: 0049-1241
            – Type: issn-electronic
              Value: 1552-8294
          Numbering:
            – Type: volume
              Value: 54
            – Type: issue
              Value: 2
          Titles:
            – TitleFull: Sociological Methods & Research
              Type: main
ResultId 1