Deciding When to Collapse Title IV Institutions Under a Shared Administrative Unit

The second NTRN Project data set selectively collapses Title IV institutions associated with the same administrative unit. The post sharing that data set noted that collapsing is used when

• The administrative unit oversees a small number of similar institutions, and

• A substantial share of finances is reported at the administrative unit level. 

This post provides more details on how this guidance can be implemented. An accompanying Excel file provides detailed characteristics of each administrative unit associated with four-year and two-year non-profit institutions in the IPEDS data set from the 2004 through 2023 fiscal years. This table updates Table A4 from the supplementary materials in this paper, incorporating more recent data and coverage of U.S. territories.

Based on the findings in the accompanying table, the following guidelines outline when to collapse Title IV institutions under a shared administrative unit:

1. Avoid collapsing data for state systems with four-year public institutions, unless the data are already collapsed for specific years. These systems typically:

• Include a large number of institutions,

• Vary widely in scope (e.g., research intensity),

• Report only a small portion of finances at the administrative unit level.

2. Collapse data for most other administrative units—which are mostly community college districts—except in atypical cases. These units generally:

• Consist of relatively similar institutions,

• Report a substantial share of revenues and expenditures at the administrative unit level.

Exceptions to the Guidelines

While the general rules apply in most cases, a few exceptions are worth noticing:

Among systems containing four-year public institutions, only the University of Puerto Rico system presents a case where collapsing is appropriate. Between FY2004 and FY2010, the system effectively operates as one Title IV institution (i.e., all system-related observations contain the same OPEID5 Value), resulting in de facto collapsing during that period. To maintain a consistent unit of observation across the full FY2004-FY2023 period, I recommend collapsing University of Puerto Rico data for all years. If your analysis uses only post-FY2010 data, collapsing is not necessary.

For administrative units not housing four-year public institutions, I recommend collapsing data in nearly all cases—except for the Los Angeles Community College District, the Mississippi Community College Board, and the Ann G. Mendez University System. These units report only a small share of core revenues and expenditures at the administrative level. Additionally, the two community college groupings include a large number of institutions and the Ana G. Mendez system spans three different Carnegie classifications, making collapsing less appropriate.

How to Collapse Title IV Institutions in Practice

The STATA code included with the second NTRN data set selectively collapses Title IV institutions based on the guidance from this post. While the process is straightforward for those who familiar with IPEDS and STATA, it involves a few key steps:

1. Reassign OPEID5 values: Before collapsing data to the Title IV Institution level, reassign OPEID5 values so that the administrative unit observation and all associated observations share the same identifier. The values 90001-90037 are used for this purpose because they are otherwise unassigned.

2. Assign main campus status to selected observations: For each year and administrative unit, identify the associated observation with the largest enrollment and assign it main campus status. In IPEDS, this is denoted by “00” in digits 7 and 8 of the OPEID. Assign “02” to the administrative unit observation and “01” to all other associated observations.

3. Collapse the data: Sum all values for observation sharing the same OPEID5 value for each year.  Then, retain only one observation per year for each OPEID5 group, containing the summed values. 

For details on the exact implementation, refer to the STATA files merge5b.do and merge7b.do accompanying the NTRN data sets. Additionally, see this accompanying Excel file for a list of all IPEDS observations associated with each administrative unit. This table updates Table A1 from the supplementary materials associated with this paper, incorporating more recent data and coverage of U.S. territories.