An analysis of Portland-metro’s income gap

The Portland-metro region has experienced weak income growth over the past 15 years. Adjusted for inflation, Portland-metro’s per capita personal income was actually lower in 2011 than it was in 2000 ($41,302 in 2011 compared to $42,847 in 2000).

A relative decline

For example, if Portland-metro’s personal income per capita increases by 1 percent in a given year that would be an absolute gain. If other metro areas increased by 5 percent that same year, then Portland-metro would have experienced a 4 percent relative decline.

While Portland-metro’s personal income per capita has been stagnant since the late 1990s, other metro areas have grown, which explains Portland-metro’s relative decline (Figure 1). Between 1996 and 2011, Portland-metro’s per capita income went from 4 percent above the U.S. metro average to 4 percent below.

These facts raise two important questions: Why is there a downward trend in Portland-metro’s relative per capita income? And why has it settled to its current level? In other words, why did it fall and why does it remain where it is?

Earnings: The key to understanding Portland-metro’s
income gap

Personal income per capita has three main components:

  • Earnings per capita
  • Dividends, interest, and rent per capita
  • Transfer payments per capita (e.g., Social Security income)

Trend versus Level

Trend: The overall direction Portland-metro’s personal income per capita is moving over time relative to other U.S. metro areas.

Level: The position of Portland-metro’s personal income per capita relative to other U.S. metro areas at a given time.

Earnings per capita accounts for about three-quarters of Portland-metro’s downward trend relative to other metros, and about two-thirds of the difference between Portland-metro’s current level of personal income per capita and the higher metro average. Therefore, the key to understanding Portland-metro’s low personal income per capita is figuring out what happened to earnings per capita.1

Two factors could explain a metro area’s below-average earnings per capita: workers earning less per job than average, and a smaller than average share of the population in the workforce. In Portland-metro’s case, relatively low earnings are due entirely to workers earning less per job compared to the metro average, and this gap continues to grow. Portland-metro actually has a larger-than-average share of its population in the workforce, which pulls up the relative level of earnings, but this advantage is diminishing.

Earnings are strongly tied to educational attainment, so to understand both the relatively low level and downward trend in earnings, this report looks more closely at Portland-metro’s educational attainment and examines residents’ educational and occupational pursuits. In all these factors, Portland-metro differs from the metro average in ways that have significant impacts on individual and regional earnings.

The connection between educational attainment and earnings

Between 2000 and 2011, Portland-metro workers without a college degree saw a small decline in earnings relative to other metro areas (falling from 4 percent above the metro average to 2 percent above), which amounts to annual earnings of a few hundred dollars over the U.S. metro average. During the same period, the share of this group in Portland-metro’s total population declined from 40 percent to 35 percent, putting it on par with the U.S. metro average, which also declined slightly. These two factors help explain the downward trend in earnings per capita, but not the lower-than-average level.

The most significant difference in the level of Portland-metro earnings per capita compared to other metros can be found in the earnings of college graduates. Between 2000 and 2011, the average earnings for Portland-metro workers with a college degree remained a constant 10 percent below the average earnings for all U.S. metro residents with college degrees, (see Figure 2). At the same time, Portland-metro started and ended the decade with a higher share of college graduates in its population compared to the U.S. metro average. Portland-metro’s share grew from 17 percent to 19 percent, while the U.S. metro average grew from 15 percent to 17 percent. This shift has a strong negative effect on the region’s aggregate earnings because the $6,953 gap is now multiplied by a larger number of workers.

Together, these factors impact the entire region’s personal income per capita and, to a large degree, explain why Portland-metro lags behind its peers across the U.S.

Portland-metro college graduates earn less

Throughout the U.S., income is strongly tied to educational attainment. As a group, college graduates earn more than their peers without college degrees. However, as noted previously, Portland-metro’s college graduates see a smaller premium for their degrees, earning 10 percent less than college graduates elsewhere.

The relatively lower earnings among Portland-metro’s college graduates explain a substantial share of Portland-metro’s below average personal income level. In fact, if Portland-metro’s college-educated workers had earnings equal to the U.S. metro average, Portland-metro’s aggregate earnings would increase by more than $2.7 billion, and Portland-metro’s overall personal income per capita would be within 1.5 percent of the U.S. metro average.2

Such an increase could generate about $110 million for Oregon’s K-12 school operations and more than $430 million in state and local taxes and fees.

To get a better understanding of Portland-metro’s lower level of personal income per capita, this analysis looks at trends within Portland-metro’s college-educated workforce to determine if the same earnings gap exists for all college-educated workers, or if it is more pronounced for any sub-set of college graduates.

The biggest income gap: White, college-educated men

Figure 3 shows the income gaps for Portland-metro workers and their subgroups. By far, the largest income gap is associated with college-educated workers, and then with white, college-educated workers. Within white, college-educated graduates, males have a bigger gap than their female counterparts. In fact, “prime-age,” white college-educated males are the biggest contributor to the earnings gap, accounting for $1.25 billion of the $2.83 billion total earnings gap.

Because they are the largest contributor to the regional income gap, the remainder of this report focuses on factors affecting the earnings of Portland-metro’s white, college-educated males (although, clearly, the region must focus on correcting the income gaps associated with all categories of workers). Another reason to look at this subgroup in greater depth is that the characteristics of the white, college-educated male subgroup remain relatively consistent among U.S. metro areas. As a result, there is a consistent basis for comparing Portland-metro’s performance to other metro areas.

In labor economics, prime-age for workers is defined as 25-54 years old. This analysis breaks down Portland-metro’s white, college-educated males further into two age groups: the younger group, aged from 25 to 39, who have graduated from college and are available for full-time employment; and the older group, aged from 40 to 54, who have valuable experience but are not yet scaling back hours, as workers over 55 tend to do. This division identifies differences across age subgroups that reveal important trends.

Fewer hours worked

Portland-metro’s white, college-educated males report working fewer hours per week than their peers in other metro areas (see Figure 4). In 2010, Portland-metro’s 25- to 39-year-old white, college-educated males reported that they usually work 42 hours per week, compared to an average of 45 hours per week in other metro areas. Portland-metro’s 40- to 54-year-old subgroup reported that they usually work 45 hours per week, compared to 46 hours in other metro areas.

The gap in hours worked per week has grown over time. In 1980 and 1990, Portland-metro’s white, college-educated males were on par with their peers in other metro areas. However, by 2000, Portland-metro’s white, college-educated males reported working fewer hours per week. This gap grew over the past decade for the younger subgroup but narrowed slightly for the older subgroup.

Portland-metro’s younger and older white, college-educated males are also less likely to work a full year than their peers elsewhere. Across all metro areas in 2010, 85 percent of the younger subgroup worked between 50 and 52 weeks per year, compared to 79 percent in the Portland-metro area. The difference is smaller for the older subgroup, with 89 percent in all metro areas working between 50 and 52 weeks per year, compared to 86 percent in Portland-metro. These differences have remained relatively consistent since 1990.

Combined, fewer hours per week and fewer weeks per year mean that Portland-metro’s white, college-educated males work fewer total hours per year. As shown in Figure 5, Portland-metro’s younger white, college-educated males averaged 1,916 hours in 2010—nearly 200 hours less than the 2,100 averaged by their peers in other metro areas.3 Portland-metro’s younger white, college-educated males also worked fewer hours per year than their peers in the other regions Portland-metro was compared to in previous Value of Jobs reports (Seattle, Denver, Minneapolis, Cincinnati, St. Louis and Sacramento). In fact, Portland-metro’s 25- to 39-year-old white, college-educated males rank near the bottom for hours worked—270 out of the 284 U.S. metro areas.

Portland-metro’s older group worked more hours than their younger counterparts, averaging 2,124 hours per year. However, they still worked fewer hours than five of the six comparison regions (Sacramento was slightly lower and Seattle is only slightly higher). Portland-metro’s older white, college-educated males also ranked near the bottom in total hours worked—235 out of 284 U.S. metro areas.

These findings indicate that on average prime-age, white, college-educated males in Portland-metro work fewer hours than their peers elsewhere, but they do not explain why. A number of factors can play a role in how many hours a person works, including personal lifestyle choices and external economic factors such as job availability. Whatever the reasons, the consequences for Portland-metro’s economy are the same – less income for workers and fewer tax dollars for public services.

Lower earnings per hour

The other factor in the earnings equation is hourly wages. Again, Portland-metro’s prime-age, white, college-educated males are at a disadvantage compared to their peers elsewhere. In 2010, the average hourly wage for Portland-metro’s younger subgroup was approximately $2.50 less than in other metro areas ($31.32 versus $33.82), and the average hourly wage for Portland-metro’s older subgroup was $3.65 less ($46.44 versus $50.09).

While these differences have fluctuated somewhat over time, average hourly earnings for Portland-metro’s prime-age, white, college-educated males have been consistently below the U.S. metro average since 1980, as shown in Figure 6.

These differences describe an important source of Portland-metro’s lower earnings, but they do not fully explain why Portland-metro’s younger and older white, college-educated males earn less per hour. There are undoubtedly a number of factors, potentially including the type of jobs available and the educational background workers bring to their careers.

Fewer business majors, more humanities degrees

Portland-metro’s population has a relatively large share of prime-age, white, college-educated males compared to other metros, especially for the younger sub-group (see Figure 7). Their educational attainment also differs from other metro areas; Portland-metro’s prime-age, white, college-educated workers are more likely to hold only a Bachelor’s degree, while workers in other metros are more likely to hold an advanced degree. This results in less overall earning power for the region. While an important finding, it still does not explain all of the differences in income.

A more significant factor could be the kind of degrees earned by Portland-metro’s prime-age, white, college-educated males as shown in Figure 8. Portland-metro’s cohort is more likely to have chosen a major that leads to lower-paying jobs, which has a lifelong impact on earnings. For example, Portland-metro has more humanities majors, who earn an average of $54,840 per year, and social science majors, who earn an average of $70,902 per year. At the same time, Portland-metro has fewer business majors, who tend to be higher earners, averaging $89,065 per year.

Further amplifying the earnings gap, Portland-metro’s humanities majors earn about $18,247 less than their counterparts elsewhere (a 33 percent gap), and Portland-metro’s social science majors earn $26,424 less than social science majors elsewhere (a 37 percent gap). In fact, in Portland-metro, every college major earns less than the U.S. metro average. Undoubtedly, a number of factors contribute to this outcome, including educational qualifications, hours worked, hourly pay and the availability of jobs that pay higher salaries.

Fewer higher-paying occupations more lower-paying occupations,

A similar pattern emerges with occupational choices. Figure 9 shows occupations in which the share of prime-age, white, college-educated males differs from the U.S. metro average by more than 1 percent. While Portland-metro has more than the average share of relatively higher-paying fields such as architecture and engineering, it also has a larger share of workers in the lower-paying fields of arts, design, entertainment and media.

On the other hand, Portland-metro has a smaller than average share of three of the highest paying occupations: management, legal services, and business operations and finance. Furthermore, for these three occupations, Portland-metro has among the largest earnings gaps compared to other metros. The legal services occupations in Portland-metro pay an average of $42,706 less per year than the U.S. metro average, business operations/finance occupations pay $19,226 less, and management occupations pay $11,241 less.

Finally, Portland-metro’s prime-age, white, college-educated males are more likely to have moved here from other states, despite the lower earning potential. Only 32 percent of Portland-metro’s college graduates were born in Oregon, while nationally an average of 46 percent of college graduates are native to their respective metro areas. Given the lower-than-average wages they face, Portland-metro’s college graduates may have non-financial reasons for choosing to locate here.

Ultimately, differences in majors, occupations and education beyond college can explain about 15 percent of the income differences between Portland-metro and other regions. Portland-metro’s prime-age, white, male college graduates work substantially fewer hours and earn lower hourly wages for nearly all college majors and occupational categories. Combined, these two factors of Portland-metro’s prime-age, white, college-educated male workforce account for an earnings level that is $1.25 billion per year less than it would be if Portland-metro performed at the U.S. metro average.

1 In 1997, Portland-metro earnings per capita were $1,457 (in 2011 dollars) more than the U.S. metro average; dividend, interest, and rent per capita were $937 more than the U.S. metro average, and transfers per capita were $723 below the U.S. metro average. In 2011, Portland-metro’s earnings per capita were $1,251 (in 2011 dollars) below the U.S. metro average; dividend, interest, and rent per capita were $22 more than the U.S. metro average, and transfers per capita were $638 below the U.S. metro average. Below-average transfer payments account for the remaining difference in the level of per capita personal income, and declining dividends, interest, and rent account for the remaining difference in the trend.

2 This calculation assumes everything else remains constant. It is possible, though, that other values would change with the average wages of college graduates. It is possible that rising incomes would increase the amount the region earned in non-wage benefits (which count toward the Bureau of Economic Analysis’ measures of earnings and personal income).

3 The American Community Survey reports a range for weeks worked. As such, the report authors only calculated a range for annual hours worked for all prime-age, white, college-educated males. In this section, the report uses the lower bound estimate for hours worked; however, the results and conclusions would not change if the report used the upper bound estimate. Detailed labor studies have shown that workers tend to overestimate the number of hours per week they work, but they do so consistently. There is no evidence that Portland-metro’s workers are more or less accurate in their reporting than others, so the trends and the differences between groups we observe are accurate even if the actual number of hours per week may be overstated.