Skip to contents

Program Expense Ratio

Source: R/get-program-expenses-ratio.R
get_program_expenses_ratio.Rd

Share of total expenses devoted to program services.

Formula:

per = program_expenses / total_expenses

Definitional Range

Bounded zero to one. Zero means no spending on programs (all overhead); one means all spending is on programs (no administration or fundraising). In practice, values above 0.95 are unusual and may reflect accounting misallocation. The empirical range for most operating nonprofits is approximately 0.50 to 0.95.

Benchmarks and rules of thumb

  • >= 75%: Charity Navigator threshold for a favorable efficiency score.

  • >= 65%: BBB Wise Giving Alliance minimum standard.

  • Below 50%: Commonly flagged as a concern, though legitimate for startup-phase organizations.

  • Caution: high ratios achieved by underinvesting in administration (the "starvation cycle") can signal long-term organizational weakness (Lecy & Searing 2015).

Calculated For: 990 filers only.

Usage

get_program_expenses_ratio( df,
  program_expenses = "F9_09_EXP_TOT_PROG",
  total_expenses   = c( "F9_09_EXP_TOT_TOT", "F9_01_EXP_TOT_CY" ),
  winsorize = 0.98 ,
  range     = "zo",
  sanitize  = TRUE,
  summarize = FALSE )

Arguments

df

A data.frame containing the fields required for computing the metric.

program_expenses

Program service expenses.

total_expenses

Total functional expenses. Accepts one or two column names; if two are provided they are coalesced with the 990 value taking priority over 990EZ.

winsorize

The winsorization value (between 0 and 1), defaults to 0.98, which winsorizes at the 1st and 99th percentiles.

range

Character string specifying the theoretical range of the ratio, used to set winsorization bounds. Default "zo". Options: "np" (negative to positive), "zp" (zero to positive), "zo" (zero to one), "nz" (negative to zero), or a custom "lo;hi" pair (e.g. "0;10").

sanitize

Logical (default TRUE). If TRUE, NA values in the financial input columns are imputed to zero before the ratio is computed, respecting form scope: Part X and VIII/IX fields (990 only) are imputed only for 990 filers; Part I summary fields (990 + 990EZ) are imputed for all filers. The returned dataframe always contains the original unmodified input columns.

summarize

Logical. If TRUE, prints a summary() of the results and plots density curves for all four output columns (raw, winsorized, z-score, percentile). Defaults to FALSE.

Value

Object of class data.frame: the original dataframe appended with four new columns:

- `prog_exp`   - program expense ratio (raw)
- `prog_exp_w` - winsorized version
- `prog_exp_z` - standardized z-score (based on winsorized values)
- `prog_exp_p` - percentile rank (1-100)

Details

Primary uses and key insights

The program expense ratio is arguably the most widely cited single metric in nonprofit financial analysis. It measures what fraction of total expenses is devoted directly to mission-related program services, as opposed to management, administration, and fundraising. It is the primary basis for charity rating systems (Charity Navigator, Better Business Bureau Wise Giving Alliance) and is commonly used by donors, funders, and board members as a shorthand efficiency measure.

However, it is also one of the most criticized metrics in the literature because it can be gamed (by misallocating shared costs to programs), because overhead is necessary for organizational effectiveness, and because the "ideal" program ratio varies substantially by mission type and operating model.

Formula variations and their sources

Program service expenses / total functional expenses (Part IX line 25B / line 25A). For 990EZ filers, Part I total expenses is used as the denominator fallback; however, the 990EZ does not separately report program expenses, so this ratio is primarily meaningful for full-990 filers.

An alternative formulation uses total revenue rather than total expenses in the denominator, but the expense-to-expense version is standard and avoids distortions from surplus or deficit years.

Canonical citations

  • Tuckman, H.P. & Chang, C.F. (1991). A methodology for measuring the financial vulnerability of charitable nonprofit organizations. Nonprofit and Voluntary Sector Quarterly, 20(4), 445-460.

  • Charity Navigator. Financial Health Methodology. charitynavigator.org.

    • Uses program expense ratio as a primary rating component (target: 75 percent or more).

  • Lecy, J.D. & Searing, E.A. (2015). Anatomy of the nonprofit starvation cycle. Nonprofit and Voluntary Sector Quarterly, 44(3), 539-563. - Critical analysis of program expense ratio benchmarks and their perverse incentives.

  • Overhead Myth Campaign (GuideStar, BBB Wise Giving, Charity Navigator, 2013).

    • Industry statement cautioning against overreliance on overhead ratios.

Variables used:

  • F9_09_EXP_TOT_PROG: Program service expenses (program_expenses, 990)

  • F9_09_EXP_TOT_TOT: Total functional expenses (total_expenses, 990)

  • F9_01_EXP_TOT_CY: Total expenses from Part I (total_expenses, 990EZ fallback)

Examples

library( fiscal )
data( dat10k )

d <- get_program_expenses_ratio( df = dat10k )
#>    :: Total expenses equal to zero :: 143 case(s) replaced with NaN
head( d[ , c( "prog_exp", "prog_exp_w", "prog_exp_z", "prog_exp_p" ) ] )
#>     prog_exp prog_exp_w   prog_exp_z prog_exp_p
#>        <num>      <num>        <num>      <int>
#> 1:        NA         NA           NA         NA
#> 2: 0.8313573  0.8313573  -0.06024897         49
#> 3: 0.8111404  0.8111404  -0.19228323         45
#> 4:        NA         NA           NA         NA
#> 5: 0.0000000  0.0000000 -10.31395412          1
#> 6:        NA         NA           NA         NA