What is I-O?

Industrial-organizational (I-O) psychology is the scientific study of the workplace. Rigor and methods of psychology are applied to issues of critical relevance to business, including talent management, coaching, assessment, selection, training, organizational development, performance, and work-life balance.

>

2003 Income and Employment Survey Results for the Society for Industrial and Organizational Psychology 

Gina J. Medsker, David A. Katkowski, and Daniel Furr
Human Resources Research Organization (HumRRO)

Editors Note: This survey is labeled 2003 because the data represent salaries and employment for the calendar year of 2003, although the survey was administered in 2004. 

The surveys purpose was to collect information on income levels of industrial and organizational psychologists in SIOP and on employment and background variables that would help with interpretation of income data. Survey instructions were e-mailed on August 3, 2004 to all Members, Associate Members, and Fellows with active e-mail addresses on record (n = 3,264). The survey was electronically available until September 21; 1,116 individuals responded. This was the first SIOP income survey to be administered electronically. The 34.2% response rate for 2003 was similar to the 35.3% rate for 2000 but lower than the 72.8% rate in 1988 and 43.6% rate in 1997. Declining response rates in recent years are a problem with survey administration in general, and this may explain some of the decline for this survey. Another cause may have been the increased length of the 2000 and 2003 surveys relative to previous surveys. 

Results

Sample characteristics
Percentages of respondents by type of employer (27.1% consulting or individual practice, 33.3% academic, 8.5% public sector, and 30.9% private and nonprofit sectors) were fairly similar to those in the SIOP membership (30.0% consulting or individual practice, 34.5% academic, 5.6% public sector, and 30.0% private sector and unspecified). Table 1 compares the 2003 sample to previous survey samples on several background variables. The percentage of women has continued to increase with each survey. Percentages by type of SIOP membership on the 2003 survey were fairly similar to those for the 2000 survey, as well as to types of membership within SIOP as a whole (11% Associates, 83% Members or International Affiliates, and 6% Fellows), whereas percentages of the sample working part time and in metro New York City both dropped somewhat. In contrast to the trend on past surveys of an increasing proportion of respondents who received their doctorates 25 or more years ago, higher percentages of 2003 respondents had their doctorates 9 years or less. The percentage who received their highest degree after 1989 was also much higher for 2003 respondents (70.5%) than for the SIOP membership (55.2%). Percentages with doctorates and masters degrees were similar for 2000 and 2003 respondents. Given that 11% of SIOP members are associate members who have a masters degree, but not a doctorate, the percentage of 2003 respondents with only a masters is roughly similar to that of the SIOP membership. 

Table 1
Characteristics of Samples Across Time (Cross-Sectional) 
___________________________________________________________________________________

                                              1982          1988          1994           1997           2000          2003          2003
                                                                                                                                                          weighted
___________________________________________________________________________________

___________________________________________________________________________________
Note. n/a indicates that data are not available. Doctorate includes those with PhD, PsyD, and EdD. Statistics include both those with masters and doctorate, except for years since doctorate, and degree-doctorate, which only include those with doctorates. Masters category includes those who have nearly completed doctorates, but had not yet graduated at the time of the survey. Weighting in the final column is based on years since highest degree in the SIOP membership population. 

Income levels
Highest degree obtained. Respondents were asked to provide their 2002 and 2003 total salary or personal income, not including bonuses or other variable pay, from their primary employer. As shown in Table 2, the median incomes for respondents with doctorates and masters degrees were lower for 2002 and 2003 than for 2000. This is the only survey on which median incomes for those with doctorates have been lower than on the previous survey, although the median for those with masters degrees has shown decreases on past surveys. The higher percentages of women and those with doctorates for 9 or fewer years and the lower percentage of respondents located in metro New York in the 2003 sample are factors that are likely influencing the decline in median income. Compared to the 2000 sample, the 2003 sample is also younger (45 or older 38.8% in 2003 and 53.7% in 2000), less likely to be an owner (8.5% some type of owner in 2003 and 26.6% in 2000), and less likely to supervise others (59.9% with no subordinates in 2003 and 44.9% in 2000). (In the remainder of this report, results from analyses on income by job characteristics, employer type, or location, are only presented for 2003 income because we did not collect descriptive data on respondents jobs, employers, and locations for 2002 and cannot assume that such characteristics were the same for both 2002 and 2003.)

Sample weighting. Given differences in the 2003 sample relative to the 2000 sample and SIOP membership, we ran analyses with the 2003 data, as well as with 2003 data weighted to have similar percentages by year since highest degree as in the SIOP membership (using simulated replication with the weight command in SPSS). Among variables that were significantly (p < .05) and most highly correlated with 2003 primary income in the unweighted sample, years since highest degree (r = .27) is available for all in the SIOP membership database and is a variable on which the 2003 sample differs considerably from the 2000 sample and SIOP membership. Other variables with larger or similar correlations with income were academic rank (r = .67), working in an academic psychology department (r = -.45) or academic business or management department (r = .47), having academic tenure (r = .38), nonacademic job level (r = .32), years work experience in industrial and organizational psychology (r = .29), age (r = .27), and being an owner (r = .26). Years since highest degree was also highly correlated with other variables that were highly related to 2003 primary income (correlations for year of highest degree are -.87 with age, -.91 with years work experience in industrial and organizational psychology, -.64 with years with 2003 employer, -.64 with academic tenure, -.70 with academic rank, and -.48 with nonacademic job level). With the weighting, we found percentages on several sample characteristics were closer to 2000 sample characteristics (see last column in Table 1). Weighted results generally provide a better representation for the SIOP membership population; however, unweighted results are also presented for comparison. As shown under the sample sizes in the last two columns in Table 2, with the weighting, 2002 and 2003 median primary incomes were higher than in 2000.

Table 2
Demographic Comparison of Median Primary Incomes for Selected Subgroups by Year



^aThe top row contains income based on unweighted data; numbers in parentheses in the second row are sample sizes; numbers under sample sizes are based on weighting by years since highest degree in the SIOP membership. 
^bIncludes all respondents regardless of degree.
^cIncludes only respondents with a doctorate.

Gender. For unweighted data, Table 2 shows that median primary income for women was 17.4% lower than that for males in 2003 and 16.5% lower in 2002. On prior surveys, the median income for women was 18.6% lower than that for men in 1982, 19.4% lower in 1988, 22.0% lower in 1994, 21.7% lower in 1997, 17.6% lower in 1999, and 17.2% lower than in 2000. The mean primary income for women in both 2002 and 2003 ($79,822 and $86,584, respectively) was significantly (t(958) = 6.0, p < .001, two-tailed, unequal variances, and t(770) = 4.2, p < .001, two-tailed, unequal variances, respectively) lower than the mean income for men ($104,130 in 2002 and $113,271 in 2003), as it had been in 1999 and 2000. The mean salary for women was 40.4% lower in 1999, 36.8% lower in 2000, 23.3% lower in 2002, and 23.6% lower in 2003 than the mean salary for men. With weighting, medians are higher for both men and women in both 2002 and 2003. Mean incomes were also higher for both men ($124,127 in 2003 and $115,410 in 2003) and women ($93,876 in 2003 and $87,244 in 2002). However, womens median incomes were still 16.6% lower for 2002 and 2003, and their means were 32.3% lower for 2002 and 24.4% lower for 2003 based on the weighted data. Some of the discrepancy in primary income may be explained by gender differences observed in other areas. In a later section, we report results of regression analyses that allow examination of the relationship of gender to income, while controlling for other variables.

Age. As Table 2 shows, unweighted median primary income was highest in 2002 for the 55 and older group and highest for those 50 to 54 in 2003. Unweighted median incomes for respondents younger than 40 and from 45 to 49 were lower in 2002 and similar in 2003 to what they had been in 2000. For those from 40 to 44 and those 50 or older, unweighted 2002 and 2003 median incomes were both higher than in 2000. Weighted medians, which are under the sample sizes for 2002 and 2003, are mostly higher than unweighted medians; however, 2003 weighted medians for those younger than 35 and from 45 to 49 show little or no difference from the 2000 median incomes.

Status as a partner, principal, or owner. In the unweighted 2003 sample, 4.0% were sole proprietors or owners, 2.0% partners, 1.2% principals, and 1.3% primary shareholders (i.e., owners of 20% or more of a corporation). These percentages were lower than the 14.5% sole proprietor or owners, 3.1% partners, 5.1% principals, and 3.9% primary shareholders in the 2000 sample. Mean and median primary incomes for all of these groups were higher than for others in the unweighted 2003 sample: $145,233 mean and $120,000 median for sole proprietors (n = 43), $145,048 mean and $130,000 median for partners (n = 21), $128,423 mean and $125,000 median for principals (n = 13), and $271,615 mean and $155,000 median for primary shareholders (n = 13). In comparison, those who marked none of the above on ownership status had a mean income of $102,214 and median of $86,000 (n = 387), and those who had no response because they worked in universities or colleges, nonprofits, or government had a mean of $92,688 and median of $80,000 (n = 579).
With weighting, all means and medians were higher, except those for primary shareholders: For sole proprietors or owners the mean was $151,963 and median was $128,267; for partners the mean was $169,389 and median was $150,000; for principals the mean was $136,664 and median was $147,752; for primary shareholders the mean was $250,817 and median was $154,618; for those working for a for-profit organization who marked none of the above, the mean was $114,635 and median was $94,697; and for those who had no response because they worked in universities or colleges, nonprofits, or government, the mean was $100,571 and the median was $85,000.

Years since doctorate. Figure 1 displays unweighted 2003 incomes from the primary employer for respondents with doctorates by the number of years since they received their degree. Respondents who received doctorates from 20 to 24 years ago had the highest median income, and those who received doctorates 25 years ago or more had the highest mean income. Because the highest degree is the doctorate for 87% of the sample, this variable is the same as the variable used to weight the data (years since highest degree) for most of the sample, so results are fairly similar for weighted and unweighted data and are not shown for the unweighted data.

Maximum 75th Percentile Median 25th Percentile Minimum

____________________________________________________________________________________
                               < 2               24             59            1014             1519           2024               25+
____________________________________________________________________________________

____________________________________________________________________________________
Note. Extreme values are not presented in the figure. Doctoral respondents only. 

Figure 1. Descriptive statistics representing 2003 primary income as a function of years since obtaining the doctorate based on weighted data.

Geographic location of employment. Specific metro areas listed on the survey were chosen because they are typically the highest paid in the U.S. With unweighted data, San Francisco/San Jose had the highest 2003 median income ($120,000), followed by Manhattan ($118,500), Other New York Metro ($110,000), and Boston ($102,000). With weighting (Figure 2), medians for most areas went up, except for San Francisco/San Jose, Manhattan, and Other New York Metro, which went down slightly. Based on either unweighted or weighted data, the four cities with the top medians in 2003 were also among those with the top five medians in 2000 results: Boston, San Francisco/San Jose, Other New York Metro, and Manhattan. 

Note. Doctoral respondents only. Sample sizes by location are in parentheses. 

Figure 2. 2003 median primary income for doctorates as a function of location based 
on weighted data.

Type of principal employment. Of respondents with doctorates, over half indicated that their principal employer was a university or college (38.0%, n = 354) or private-sector consulting organization (20.7%, n = 193). With weighting, the two biggest employer categories were still universities and colleges, with 40.2%, and private-sector consulting organizations, with 20.4%, but median income for private sector health care, with its small sample size (n = 6 unweighted), dropped from the highest median to eighth highest (see Figure 3). Based on weighted data, pharmaceuticals had the highest median, followed by telecommunications, self-employed consulting, and technology, computers, and software. All of these had been among the employer types with the top five highest medians in the unweighted data, as well. 

Note. Doctoral respondents only. Sample sizes by location are in parentheses. 

Figure 3. 2003 median primary income for doctorates by type of 
primary employer based on weighted data.

Type of academic employment. For those working in universities or colleges, the unweighted mean income differed by highest degree a department offered (bachelors $58,606, n = 29; masters $77,971, n = 116; doctorate $91,548, n = 194; F(2,336) = 11.06, p < .001). The unweighted mean income also differed by accreditation status (accredited $86,796, n = 303; not accredited $66,881, n = 26; F(1,327) = 5.87, p = .016). In addition, the unweighted mean income in business or management departments ($107,409, n = 137) was significantly higher (F(1,313) = 108.75, p < .001) than the unweighted mean in psychology departments ($66,090, n = 178). The unweighted mean income did not differ significantly (F(1,343) = .20, p = .66) for private ($82,449, n = 87) and public institutions ($84,665, n = 258). Based on weighted data, those in psychology departments with the bachelors as highest degree had a mean of $53,835 and median of $51,886, with the masters as highest degree a mean of $61,239 and median of $57,861, and with the doctorate as highest degree a mean of $81,032 and median of $73,020. For weighted data for business or management, those in departments with the bachelors as highest degree had a mean of $64,352 and median of $65,940, with the masters as highest degree had a mean of $111,517 and median of $95,000, and with the doctorate as highest degree had a mean of $120,358 and median of $110,000.

Academic titles by department type. Figure 4 shows unweighted 2003 income for psychology and business/management departments for the five academic titles that had adequate sample sizes. Five deans from business or management responded and had a mean primary income of $142,400 and a median of $130,000. Distinguished or chaired professors had the highest primary incomes compared to other ranks in the same type of department. Based on unweighted data, there were significant differences between incomes in psychology and business/management departments for assistant professors (F(1,118) = 111.90, p < .001), associate professors (F(1,76) = 62.15, p < .001), and full professors (F(1,59) = 19.66, p < .001); a marginally significant difference for department chairs (F(1,10) = 3.66, p = .09); and no significant difference for distinguished and chaired professors (F(1,19) = .53, p = .48). Figure 4 tables present both weighted and unweighted results.

_______________________________________________________________________________________________
                                                                                                                                                                                                        Distinguished                                                       Assistant                     Associate                                                      Department                  or Chaired 

Psychology (Unweighted Data)

Business or Management (Unweighted Data)

Psychology (Weighted Data)

Business or Management (Weighted Data)

___________________________________________________________________________________
Note. The figure shows unweighted data. Extreme values are not presented in the figure. Doctoral respondents only. 
^aNot enough cases to report.

Figure 4. 2003 primary income by type of university or college department and academic title based on unweighted and weighted data.

Nonacademic job titles. Figure 5 shows unweighted 2003 primary income by level of job for those in private sector, nonprofit, and government organizations. Tables with the figure show weighted and unweighted results. The president/chief executive officer level has a lower median but a higher mean and much broader range than the vice-president level in either weighted or unweighted data. 

_______________________________________________________________________________________________
                                                                                                                                                                                                       
                                                      Consultant,                                                      Manager/        Manager/ 
                                    Entry        Researcher,         Senior          First-Line        Director           Director           Vice-             President 
                                    Level        Practitioner          Level          Supervisor       HR/OB            Non-HR        President          or CEO
_______________________________________________________________________________________________

Unweighted Data

Weighted Data

___________________________________________________________________________________
Note. The figure shows unweighted data. Extreme values are not presented in the figure. Doctoral respondents only. 
^aNot enough cases to report.

Figure 5. 2003 primary income in private sector, nonprofit, and government organizations by job level based on unweighted and weighted data.

Starting salaries. With unweighted data from those who had hired new doctoral graduates in 2003 (n = 73), the mean starting salary was $65,362 and median was $60,000. Based on unweighted data from those who hired new masters degree graduates (n = 59), the mean was $47,808 and median was $43,000. For 23 respondents who self-reported that they had a doctorate and worked in their current position for their employer one year or less, the 2003 unweighted mean primary income was $63,652 and median was $65,000, so these self-report figures are similar to the figures from those who said they had hired new doctoral graduates. 

Retirement, bonus, and raise information
Retirement plans. For 2003, 76.8% (n = 819) of respondents indicated that their employer contributes to a defined contribution plan for them, and 35.6% (n = 380) indicated that their employer provides a defined benefit plan. For 566 respondents who reported the percentage of income that their employer contributed to a defined contribution plan in 2003, the unweighted mean was 6.9% and median was 6.0%; the weighted mean was 7.1% and median was 6.0%. For 104 respondents who reported the percentage of income that their employer will provide after they retire through a defined benefit plan, the unweighted mean was 43.7% and median was 50.0%; the weighted mean was 44.3% and median was 50.0%.

Bonuses and stock options. For 2003, respondents in the private sector were most likely to report a bonus (71.8%), followed by government and military (40.7%; though only 10.5% in state and local government reported a bonus), nonprofit (39.6%), university or college (13.5%), and self-employed (3.9%). The most prevalent types of bonuses were for individual (30.8% of respondents), organizational (28.2%), and group, department, or unit performance (13.8%). Less than 5.0% received bonuses for special projects (4.3%), other reasons (3.2%), signing on or recruiting (1.8%), retention (1.3%), or exercising stock options (0.9%). To examine size of bonus by type, data from 206 respondents reporting an amount for only a single type of bonus were used. In unweighted data, the mean was 12.0% and median was 10.0% for sign-on or recruiting bonuses (n = 9); the mean was 14.6% and median was 4.7% for individual performance bonuses (n = 90); the mean was 12.6% and median was 11.8% for group, department or unit performance bonuses (n = 11); the mean was 9.7% and median was 5.9% for organizational performance bonuses (n = 75); and the mean was 10.8% and median was 8.5% for special project bonuses (n = 21). In weighted data, the mean was 15.2% and median was 16.1% for sign-on or recruiting bonuses (n = 5); the mean was 15.8% and median was 4.3% for individual performance bonuses (n = 88); the mean was 12.3% and median was 13.0% for group, department, or unit performance bonuses (n = 9); the mean was 12.7% and median was 7.7% for organizational performance bonuses (n = 72); and the mean was 9.0% and median was 7.0% for special project bonuses (n = 19).

Pay raises. A majority of respondents (70.8%) reported receiving a pay raise. The largest average raises, as a percent of base salary before the raise, were for a higher level job at a new employer (n = 13, 27.7% mean and 20.0% median unweighted; 26.6% mean and 23.3% median weighted), a job with similar responsibility at a new employer (n = 14, 13.8% mean and 13.5% median unweighted; 13.5% mean and 11.9% median weighted), a promotion with the same employer (n = 47, 12.9% mean and 9.5% median unweighted; 11.0% mean and 9.0% median weighted), or an increase in responsibility with the same employer (n = 40, 10.6% mean and 10.0% median unweighted; 10.8% mean and 10.0% median weighted). Raises tended to be smaller for a job transfer at the same employer (n = 7, 8.1% mean and 5.0% median unweighted; 8.8% mean and 4.8% weighted) or for the same job at the same employer (n = 537, 5.0% mean and 4.0% median unweighted; 4.6% mean and 4.0% median weighted).

Regression analyses
We analyzed the relationships of personal and employment characteristics with income from the primary employer using unweighted data and listwise deletion in separate regression equations for respondents who worked in universities or colleges and for those working for nonacademic employers because we had collected data on several different variables for the two groups (e.g, type of academic department, different job titles, ownership status for nonacademics). The equation for the academic sample accounted for more variance in 2003 income from the primary employer (R² = .74, R²_adj = .70, F(38,246) = 18.16, p < .001) than the equation for the nonacademic sample (R² = .41, R²_adj = .34, F(49,461) = 6.40, p < .001). 

For the academic sample, being female, working outside the U.S. or Canada (compared to areas not listed on the survey that are in the U.S.), working in departments whose highest offered degree was a bachelors or masters (compared to those that offer a doctorate), and rank as an associate professor (compared to rank as an assistant professor) had significant, negative coefficients (p < .05), but status as an APA fellow, average hours worked per week for the primary employer, working in a business/management or other type department (compared to a psychology department), working in an accredited department or school, having tenure, and being a distinguished or chaired professor (compared to an assistant professor) had significant, positive coefficients (p < .05). The negative coefficient for associate professors, relative to assistant professors, suggests a wage compression problem for the academic sample. In the equation for nonacademics, years experience in industrial and organizational psychology, average hours worked per week for the primary employer, working in Manhattan (compared to working in unlisted areas in the U.S.), being a vice-president or president/chief executive officer (compared to a senior consultant, researcher, or practitioner), and working in technology, computer or software organizations (compared to consulting organizations) had significant, positive coefficients (p < .05).

Discussion

The 2003 survey was the first SIOP income and employment survey to be administered via e-mail and the Internet. Although the 2003 response rate was very similar to that for the 2000 survey, the 2003 sample was considerably different from the 2000 sample in terms of age and percentages of males and females and different from the 2000 sample and SIOP membership in terms of years since receiving ones highest degree. Therefore, a weighted sample was used to better reflect the SIOP membership. When the 2003 sample was weighted to have the same percentages of year since highest degree as in the SIOP membership population, mean and median incomes showed an increase compared to 2000. Results based on weighted data were more consistent with trends from past surveys than results based on unweighted data and likely better represented 2003 income levels for the SIOP membership. Weighting may be needed again on future SIOP surveys if there are significant differences between the characteristics of the respondents and those of the SIOP population. Offering both a paper-based and electronic version of the survey might help determine whether the switch to an electronic survey influences the average age and experience of the respondents.

Separate regression equations for those employed in academia and for those employed in the private sector, nonprofit sector, and government were analyzed for 2003 data, whereas all sectors had been combined in regression analyses for past surveys. Although such separate regression analyses lower the statistical power to detect significant relationships, results suggest that factors influencing income may differ by the economic sector in which one is employed. 

Authors Notes

The Human Resources Research Organization (HumRRO) conducted the 2003 Income and Employment Survey of the membership of the Society for Industrial and Organizational Psychology (SIOP) as a service to SIOP. We would like to acknowledge the support of Kenexa, who programmed the survey for Internet administration. We would also like to acknowledge the involvement of Lee Hakel in the SIOP Administrative Office and Mark Schmit, Doug Reynolds, Julia McElreath, Gary Greguras, Pauline Velez, Fritz Drasgow, Lois Tetrick, Irene Sasaki, Angelo DeNisi, Michael Burke, Liberty Munson, Steven Rogelberg, Leaetta Hough, and Jose Cortina, who reviewed a survey draft. A more detailed version of this report is available at www.siop.org. Please address correspondence to the first author at HumRRO, 66 Canal Center Plaza, Suite 400, Alexandria, VA 22314 or at gmedsker@humrro.org.

July 2005 Table of Contents | TIP Home | SIOP Home