From Peace Corps Wiki
Current Rates: http://www.peacecorpswiki.org/ETrates
Peace Corps Manual Section 284 establishes the policies and procedures governing the circumstances under which a Peace Corps Volunteer's service may end prior to the completion of service date. An ET occurs when Volunteer "cannot or should not remain in service until his or her completion of service (COS) date." The most current report available is the FY 2009 ET Data Summary. Currently, out of 100 volunteers who begin their service, about 35 will Early Terminate sometime before their completion of service. (Source: "Early Termination in the Peace Corps", Appendix A )
There are four types of early termination:
- Resignation: A resignation is a decision made by a Volunteer that he/she no longer wishes to continue in Peace Corps service.
- Medical Separation: If a Volunteer has or develops a medical condition that Peace Corps cannot medically accommodate or resolve within forty-five (45) days, the Volunteer will be medically separated. This decision is made by the Office of Medical Services (OMS) in consultation with the Peace Corps Medical Officer and, if needed, appropriate medical consultants.
- Administrative Separation: Pursuant to the Peace Corps Act, 22 USC 2504(i), the service of a Volunteer may be terminated at any time at the pleasure of the President. The authority of the President to terminate service has been delegated to the Director of the Peace Corps. Accordingly, the Director, or anyone to whom the Director delegates such authority, may separate a Volunteer at any time purely at the discretion of the deciding official and as otherwise expressly provided herein.
- Interrupted Service: A Volunteer may be separated with interrupted service status if the Country Director determines that circumstances beyond the control of the Volunteer make it necessary for the Volunteer to leave his or her present assignment.
 Rough Estimate
A back-of-the-envelope calculation can be done as follows: Peace Corps currently accepts about 5,000 volunteers per year. If no volunteer leaves early then after 27 months a total of 11,250 volunteers should be serving right now (27 months * 5,000 per year) - from those volunteers just beginning service to those just ending. Any volunteers who ETed during that time would make the number smaller. If 10% of the volunteers have ETed then the 11,250 number would be 10% less; if 50% ETed then the 11,250 number would be half, etc. According to the Peace Corps website there are currently 7,500 volunteers serving around the world right now. This means about 3,750 volunteers out of the 11,250 total possible have Early Terminated, equaling 33%. (3,750 volunteers who ETed out of 11,250 total) This is consistent with more precise estimates.
 Determining ET Rates
Before 1981, Peace Corps used the V Year method, standing for 'Volunteer-Year'. It is the number of ET's by an average volunteer during one year of service. Since most volunteers successfully complete their service, which is longer than one year, this number would be a "fraction of an ET". If the V-Year ET Rate was 0.20 (per year) we would expect about 2 volunteers out of 10 to ET per year, or about 1 volunteer out of 10 in a six-month period.
In 1981, at the request of the General Accounting Office, the agency's Office of Special Service began to report Volunteer attrition, using the Cohort Method. This method addresses the question: 'If X number of people enter Peace Corps service during a given time period, how many actually complete their service?"
A cohort is defined as all trainees who enter on duty during a Fiscal Year (i.e. the Class of FY90), and the cohort attrition rate as the percent of Volunteers and Trainees within a cohort who do not complete service (e.g. 29% of the Class of FY90 did not complete service). Cohort attrition is also referred to as 'class of' attrition rate (where 'class of' refers to all trainees entering on duty during a Fiscal Year).
Cohort ET Rates tend to be at least double the ET reported by posts using the V Year chart method. While these latter rates (V-Year) are essential for budget calculations purposes, they provide an incomplete picture of overall attrition. Unlike the V Year chart method used by posts to calculate T and V ET Rates, the cohort method provides a complete picture of attrition because it follows people over their full tour (meaning, for most Volunteers, a 24 to 27 month period).
In the early 2000's, Peace Corps switched formulas to an Annual Method in reported ET data. The 'ET Rate' using the annual method is the number of PCVs/Ts who separated from the Peace Corps during the fiscal year divided by the total number of trainees and volunteers who served at any time during the fiscal year. Thus, an individual who served for one day and another individual who served for the full fiscal year are counted equally for the denominator of the fraction. This method addresses the question: 'Of the n total number of individuals who served at any time during a given time 12-month period, how many early terminate in that same time span?"
Because the different methods in reporting ET data use different data and calculations, care must be taken so as not to inappropriately compare the two. As the Cohort Method spans over a twenty-seven month period (or more) while the Annual method is strictly a twelve-month period, the results do not lend to comparability.
 Comparison Between Methods
|V Year||Cohort Method||Annual Method|
|before 1981||1981 - early 2000s||used currently|
|roughly half of Cohort value||"complete picture of attrition"||roughly half to a third of Cohort Value|
|n/a||FY 1990: 29% ET Rate||FY 1990: 13% ET Rate|
|n/a|| Meaning: 29% of all volunteers who|
entered service in FY 1990 ET'ed
sometime before their COS date
| Meaning: 13% of all volunteers who|
served any portion of the fiscal year
of 1990 ET'ed within that year
 Converting Between Methods
Since an exact conversion can only be made with raw data, the following will be an approximate conversion between the two methods based on the following assumptions:
- Every year had the same number of volunteers who started service in that year - in other words, every Cohort group was the same size.
- Within a given year volunteers entered Peace Corps service randomly between January 1st and December 31st .
- Full service was exactly 27 months.
- Those who Early Terminated did so randomly, anywhere from one day of training to right before their anticipated COS date.
These assumptions make every cohort identical and every year identical. While it may be an idealized model, for its simplicity it gives a reasonably accurate conversion from one method to the other.
Roughly, for every 2% away from 10% the Annual Method is the Cohort method will be 5% away from 30%. For more information on the derivation of the conversation chart presented here, see the report Annual and Cohort Early Termination Rates
 Current ET Rates
Latest report: FY 2009 ET Data Summary
- PCVs/Ts : all Peace Corps Volunteers on Duty in FY2009
- ET: Early Terminations (RS+MD+AD+IS)
- RS: Resignations
- MD: Medical Separations
- AD: Administrative Separations
- IS: Interrupted Service
- RS/ET % : Percent of Early Terminations that were resignations
- Annual ET%: ET rate currently used by Peace Corps
- Example: 10.0% of all volunteers worldwide who served any portion of the fiscal year of 2009 ET'ed within that same year.
- Cohort ET%: Estimated Cohort Rate ("Out of 100 volunteers") 
- Example: The worldwide annual ET rate (10.0%) is roughly equivalent to 29.2% of volunteers ET'ing sometime before their COS date.
 Historical ET Rates
Historical ET Rates depends on which formula you used in calculating them. Prior to 2000 Peace Corps used the Cohort Method in calculating ET Rates, and Historical ET Rates were reported using the same method. Since then, Peace Corps has switched to the Annual Method and recently published Historical ET Rates have presented data to correspond to the current method.
|The Historical ET Rate using the Cohort Method. A cohort is defined as all trainees who enter on duty during a Fiscal Year (i.e. the Class of FY90) and the cohort attrition rate as the percent of Volunteers and Trainees within a cohort who do not complete service.|
Example: 29% of all volunteers who entered service in FY 1990 ET'ed sometime before their COS date.
|The Historical ET Rate using the Annual Method is the number of PCVs/Ts who separated from the Peace Corps during the fiscal year divided by the total number of trainees and volunteers who served at any time during the fiscal year.|
Example: 13% of all volunteers who served any portion of FY 1990 ET'ed within that year
For a more complete table of volunteer numbers from 1961 to 1970, using the statistics from the table above, see here. Estimates of current numbers are given below.
|Year|| Number of volunteers|
arriving that year (est.)
| Number of volunteers|
serving at any moment (est.)
| ET rate|
Because the different methods in reporting ET data use different data and calculations, care must be taken so as not to inappropriately compare the two. As the Cohort Method spans over a twenty-seven month period (or more) while the Annual method is strictly a twelve-month period, the results do not lend to comparability. Care must be taken in comparing ET Rates reported from documents using different formulas for measuring ET Rates.
Spreadsheet data using the Cohort Method
Reports using the Annual Method
- 2009: FY 2009 ET Data Summary
- 2008: FY 2008 ET Data Summary
- 2007: FY 2007 Quantitative Early Termination Report
- 2006: FY 2006 Quantitative Early Termination Report
- 2005: FY 2005 Quantitative Early Termination Report
- 2005: FY 2005 Quantitative Early Termination Report (Appendices)
- 2005: Resignation Reason Report (January 2005 - December 2005)
Reports using both the Annual Method and Cohort Method
Reports using the Cohort Method
- 1999: Report on Early Terminations May 1999
- 1997: Early Termination Work Group Progress Report, March 1997
- 1996: Peace Corps Report on Early Termination FY90-FY96
- 1995: Reasons for Early Termination of Peace Corps Service; A Pilot Study to Improve Data Collection (Dec 1995)
- 1992: FY1992 Report of Early Termination Statistics
(Note: All documents were obtained from Peace Corps, by request, using the Freedom of Information Act and are in public domain)