Appendix B: Impact Analysis Methodology

This section describes the methodology and data sources used to conduct the impact analysis in Chapter 2.

Calculating Total Economic Impacts

The direct economic impacts presented in the report are based on public financial data for the CSU and/or from calculations based on assumptions discussed in the following sections. The direct economic impacts included annual CSU operational expenditures, average (four-year) capital expenditures, auxiliary expenditures, and student expenditures.

The total economic impacts, including total spending, total job, and total tax impacts, were calculated using the IMPLAN economic impact software package. Within a defined study region, IMPLAN uses average expenditure data from the industries that originate the impact on supplier industries to trace and calculate the multiple rounds of secondary indirect and induced impacts that remain in the region (as opposed to “leaking out” to other areas). IMPLAN then uses this total impact to calculate total job and tax impacts.

Despite an increase in the magnitude of the CSU’s impact, the total number of supported jobs that is reported is fewer than what was reported in the 2004 analysis due to national, economy-wide increases in worker productivity. The 2004 analysis used the IMPLAN model, version 2, which relied on the most recently available data from 2001. The model used the 2001 data and industry trade flows to create industry multipliers that were used to estimate the direct job impacts associated with direct spending. The current (2010) analysis uses version 3 of the IMPLAN model, which relies on 2008 data (also the most recently available). The difference in direct employment generated by the model can be explained by changes in the underlying data that is not related to CSU specifically, but to economy-wide, increased productivity. Worker productivity increased substantially over the past decade. According to the IMPLAN model, 2003 US output per worker for the higher education sector was $58,450. In 2008, the output per worker for this sector jumped to $75,500. This change in output per worker (worker productivity) means that for each dollar invested, more worker output is produced, but by fewer employees and thus equivalent levels of direct investment generates fewer jobs (but more labor income and industry activity).

In this study, IMPLAN runs were made using California as the study region, and separately for each of the eight sub-regions defined in the report. The current (version 3) of the IMPLAN model allows for the assessment of regional interaction, and therefore can account for impact that spending in one region has on surrounding regions. Because of this enhanced capability, ICF calculated the regional impact, calibrating the model to take into account only those dollars that would be spent locally, as well as the impact on other regions and the state as a whole. All of the CSU’s direct impacts were modeled as originating in the Higher Education industry (sector #392 in IMPLAN, NAICS # 6112-3).

Operational and Capital Expenditures

This study used CSU financial statements provided by each campus and the Chancellor’s Office to estimate annual CSU expenditures systemwide as well as at the campus level. Because campus capital expenditures vary significantly from year-to-year, four years of financial statements were used to calculate an average annual capital expenditure for each campus.

Auxiliary Expenditures

Information regarding the impact of auxiliary organizations also came from internal CSU financial reports. Because the data were not broken down by the type of auxiliary enterprise; i.e., retail store, food service area, research institute, etc. The following assumptions were made regarding expenditures in each IMPLAN sector:

  • 25% Retail Trade
  • 25% Eating and Drinking Places
  • 50% Rental Housing

Student Expenditures

A significant portion of CSU student expenditures occur at auxiliary organizations (e.g., campus housing, bookstores, campus food services and parking), which are incorporated in the auxiliary organization spending as noted above.

A full accounting of student expenditures attributable to CSU operations required an estimate of off-campus student expenditures. First, it was assumed that since many students would be working in California and their region and making similar expenditures whether or not they were attending a CSU campus, the conservative assumption (i.e., an assumption that underestimates student spending impacts compared to many traditional impact calculations) was made to exclude these expenditures from the total student spending.
Only out-of-state students were counted in the statewide analysis, and only students who came from outside of the region where they attended a CSU campus were counted in the regional analysis.

In order to create these estimates, the following calculations were made:

  1. The CSU maintains a data set called Residence of Total Enrollment by Campus. This data set contains the number of students by campus and by county of residence, with a separate accounting of out-of-state students. The number of out-of-region students for each region was calculated, as well as the number of out-of-state students for California as a whole.
  2. The CSU’s Cost of Attendance 08-09 was then used to estimate, by campus, how much a student typically spends, excluding items from on-campus and auxiliary organizations, such as food, housing, and books. By multiplying this average off-campus spending by the number out-of-state/out-of-region students, the total spending (excluding food, housing, and books) by out-of-state/out-of-region students was determined.
  3. It was assumed that all expenditures for books would occur at auxiliaries, and therefore, were excluded from the additional student spending estimate because this category of expenditure was accounted for as part of the auxiliary organization spending.
  4. Some students live in on-campus housing and some live off-campus. It was assumed that for students staying in on-campus housing, all food and housing expenditures would occur at auxiliaries. For students not staying in on-campus housing, it was assumed that none of their food and housing budget was spent at auxiliaries.
    The CSU’s Housing Occupancy database was used to determine the percentage of students living in on-campus housing.
  5. On every campus, the number of out-of-region students exceeded the number of students living in on-campus housing. Therefore, except for San Francisco and Los Angeles who specified numbers of in-region residents on campus, it was assumed that 100 percent of the on-campus housing was occupied by out-of-region students. The “left-over” out-of-region students were assumed to reside in off-campus housing. Data relating to financial aid was used to estimate housing and food expenditures, which were then added to the total calculated above. This sum of spending described above became the total direct impact of student expenditures and was provided as input into IMPLAN like the other direct spending impacts of the CSU. As previously noted, the assumptions used in this analysis to generate the additional student spending were intentionally conservative; that is, they are believed to significantly understate the total additional student spending impact.

Alumni Impacts

Alumni impacts are treated differently than the other spending impacts in IMPLAN because they are not expenditures by the CSU but by CSU graduates. Thus, instead of treating the direct impact as originating from the Higher Education sector, these expenditures were assumed to originate from the Household Expenditure sector.
The method used to assess the direct impact of alumni consisted of the following steps:

  1. Data on CSU degrees granted by campus were collected, dating back to 1970-1971. It was assumed that CSU graduates from that year and later years who were still residents of the state would still be in the labor force.
  2. In order to determine the percentage of CSU graduates still residing in California and having an impact on the California economy, an average annual “out-migration rate” was calculated using Census data for two periods: 1985-1990 and 1995-2000. The out-migration rate was relatively close in magnitude for both time periods. It was assumed that CSU graduates were as likely to move out of the state as other Californians. This rate was compounded to estimate the cumulative probability of having left California to determine the percentage of each year’s CSU graduating class that remains in the state.
  3. It was assumed that CSU graduates were 25 years old upon graduation. This was assumed in order to determine the age of each year’s graduating class and the number of CSU graduates who remained in California in each of several age cohorts: 25-34, 35-44, 45-54, and 55-64.
  4. Migration Analysis

    Location

    Bachelor’s Degrees Granted

    Master’s Degrees Granted

    Bachelor’s Graduates Remaining

    Master’s Graduates Remaining

    Bachelor’s Graduates Lost

    Master’s Graduates Lost

    Bakersfield

    27,394

    7,220

    22,587

    5,943

    4,807

    1,277

    Channel Islands

    2,873

    139

    2,768

    136

    105

    3

    Chico

    103,323

    10,484

    80,412

    8,241

    22,911

    2,243

    Dominguez Hills

    49,200

    21,803

    39,649

    18,055

    9,551

    3,748

    Fresno

    109,865

    19,875

    85,506

    15,767

    24,359

    4,108

    Fullerton

    155,122

    32,493

    124,071

    25,903

    31,051

    6,590

    Hayward

    76,894

    25,033

    60,347

    20,538

    16,547

    4,495

    Humboldt

    47,045

    5,360

    36,591

    4,167

    10,454

    1,193

    Long Beach

    179,333

    38,438

    139,765

    30,448

    39,568

    7,990

    Los Angeles

    100,610

    34,227

    76,627

    26,278

    23,983

    7,949

    Maritime Academy

    1,369

    0

    1,265

    0

    104

    0

    Monterey Bay

    5,360

    376

    4,983

    356

    377

    20

    Northridge

    157,187

    32,518

    124,251

    25,914

    32,936

    6,604

    Pomona

    98,041

    10,888

    77,698

    8,718

    20,343

    2,170

    Sacramento

    142,416

    30,674

    112,335

    24,307

    30,081

    6,367

    San Bernardino

    55,262

    16,664

    46,018

    14,045

    9,244

    2,619

    San Diego

    198,174

    50,790

    154,793

    40,520

    43,381

    10,270

    San Francisco

    146,683

    47,391

    115,719

    37,100

    30,964

    10,291

    San José

    155,312

    51,768

    119,612

    41,758

    35,700

    10,010

    San Luis Obispo

    114,920

    10,742

    90,582

    8,449

    24,338

    2,293

    San Marcos

    17,420

    1,816

    15,850

    1,672

    1,570

    144

    Sonoma

    46,655

    6,913

    36,895

    5,400

    9,760

    1,513

    Stanislaus

    34,175

    4,331

    27,833

    3,563

    6,342

    768

    Entire CSU

    2,024,633

    459,943

    1,596,156

    367,280

    428,477

    92,663

  5. Data from the U.S. Census Bureau’s Current Population Survey (CPS) were used to estimate the weighted average salary of CSU bachelor’s and master’s degree-level alumni based on their age.1,2 The CPS data provide average salary, by age and sex, for Californians of different levels of educational attainment, including high school graduates, individuals with some college but not holding degrees, bachelor’s degree recipients, and master’s degree recipients. The weighted average salaries for each age cohort are presented below.
  6. Weighted Average Salaries

    Age Cohort

    Educational Attainment

    High School

    Some College, No Degree

    Bachelor's Degree

    Master's Degree

    25-34

    $28,224

    $31,956

    $48,445

    $55,635

    35-44

    $36,917

    $42,558

    $70,054

    $84,491

    45-54

    $35,338

    $41,742

    $64,810

    $72,603

    55-64

    $27,533

    $32,218

    $43,378

    $45,805

  7. For each graduation year, the total earnings of CSU alumni were calculated by multiplying the number of bachelor’s degree recipients remaining in California by the weighted average bachelor’s degree salary for that year. The calculation was repeated for master’s degree holders, and the two totals were summed. This total, summed for every year back to 1970-1971, provides an estimate of the total annual earnings of CSU alumni still living in California.
  8. The amount of total earnings that is attributable to the alumni’s CSU degree is the difference between the weighted average salary associated with their final educational level minus the weighted average salary associated with their previous educational level. For individuals with master’s degree, for example, the amount of earnings that is attributable to the alumni’s CSU master’s degree is the weighted average master’s salary minus the weighted average bachelor’s salary. For bachelor’s degree holders, the amount of earnings attributable to the alumni’s CSU degree is the average bachelor’s salary minus the average salary for either a high school graduate or transfer student who already had some college credit.
  9. Some students come to the CSU with a high school diploma only; others transfer after completing some college. The salary differences between bachelor’s degree recipients and high school graduates were calculated as well as the salary difference between bachelor’s degree recipients and transfer students with some college credits. These two differences were weighted based on historical data for the split between the two sources of students to the CSU (first-time freshmen with a high school diploma and transfer students). Lastly, a total earnings differential attributable to the CSU degrees was calculated.


1   Source: U.S. Census Bureau, Current Population Survey, Annual Social and Economic Supplements, Table P-28, Educational Attainment—Workers 18 Years Old and Over by Mean Earnings, Age, and Sex: 1991 to 2008

2   The weights for the weighted average earnings calculations were the number of male and female workers in each age cohort earning sex-specific mean earnings.