Grants & Funding
Recovery Act Funds
Office of Justice Programs and Office on Violence Against Women Request To Use Statistical Methodology for Reporting Recovery Act Job Creation and Retention 1
FREQUENTLY ASKED QUESTIONS
All prime recipients of American Recovery and Reinvestment Act of 2009 (Recovery Act) funds received from the Office of Justice Programs (OJP) and Office on Violence Against Women (OVW) must comprehensively count and report the total number and types of jobs created and retained as a result of that funding to www.federalreporting.gov on a quarterly basis. In very limited circumstances, OJP or OVW and the Office of Management and Budget (OMB) may approve the use of statistical methods to estimate job creation and retention for some prime recipients that prove that comprehensively counting created and retained jobs will be overly costly or burdensome. The following Frequently Asked Questions (FAQ) describe when statistical methods may be approved to estimate created and retained jobs and how to request approval to use these methods.
- What reported data may be statistically estimated?
All data reported by Recovery Act prime recipients and subrecipients to OJP, OVW, OMB, and www.federalreporting.gov must be actual counts except, in limited cases and with OJP or OVW and OMB approval, the number and types of jobs created and retained. Additional guidance regarding the calculation and estimation of jobs created and retained is outlined in OMB M-09-21, sections 5.2 through 5.4.2
- Under what circumstances are prime recipients permitted to use statistical estimates rather than direct comprehensive job counts?
OJP and OVW expect recipients to collect job creation and retention counts directly; however, OJP or OVW may, in coordination with OMB, approve requests to statistically estimate job counts in cases where recipients demonstrate that comprehensive job counts will create an unreasonable burden and disrupt the recipient's ability to effectively implement the underlying mission of the program. However, because it is important to have accurate counts of the number of jobs created and retained by Recovery Act programs, OJP, OVW, and OMB will only approve the use of statistical estimates for reporting job counts in exceptional cases. As recipients must collect and report all other required data for the Recovery Act directly, job counts are not expected to burden recipients excessively.
- How should a prime recipient request to use a statistical methodology?
Recipients should request to statistically estimate the number of jobs created/retained for subrecipients and vendors through the Grant Adjustment Notices (GAN) module of the Grant Management System (GMS). Recipients should select "Program Office Approvals" as the GAN type, select "other" as the Approval Type, and enter the written request.
Recipients must submit their request to use a statistical methodology at least 30 days before the end of the first reporting period for which they would like to use statistical methodology and must submit a separate request for each award. Only one approval is required, per award, for the duration of the grant unless there are changes to the methodology or personnel carrying out the methodology.
- What should be included in the request to use a statistical methodology?
Prime recipients that request permission to estimate jobs created and retained using a statistical methodology must include the following in their GAN request:
- A detailed justification explaining why directly collecting information from subrecipients and vendors is overly costly or burdensome.
The justification should include detailed estimates of the costs and burdens associated with direct collection of job creation and retention counts from sub-recipients and vendors, including the number of subrecipients and vendors, the number of subrecipients and vendors creating and retaining jobs, a thorough description of the practical difficulties or obstacles to collecting job creation and retention counts from subrecipients and vendors, and any other information deemed relevant to OJP's or OVW's consideration of the request.
- A technical description of the recipient's proposed statistical methodology with sufficient detail and supporting documentation of the sample design and estimation procedures so that the validity of the methodology can be assessed.
The technical description must include a detailed plan for developing survey-based estimates of jobs created and retained. It must explain how the plan will result in reliable estimates of the number and types of jobs created and retained; examine the internal and external validity of data on which jobs estimates will be based; and include a description of how the recipient will address nonresponse, selection bias, and estimate robustness to alternative specifications.
(1) The proposed methodology must adhere to standards and guidelines published by OMB and, in particular, must clearly demonstrate that accepted theory and methods, and OMB "best practices" will be employed in the job creation and retention count estimate development.3 In particular, the recipient's methodology must describe:
- How weights appropriate for the sample design will be employed to calculate estimates. The recipient may employ an alternative method (e.g., ratio estimators) to calculate estimates if it has evaluated (and documented) the alternative method and determined that it leads to acceptable results.
- The use of auxiliary data to improve precision and/or reduce the error associated with survey-based estimates.
- How variance estimates will be calculated through a method appropriate to the proposed survey's sample design taking into account probabilities of selection, stratification, clustering, and the effects of nonresponse, post-stratification, and ranking. It is expected that the estimates produced will reflect any design effect resulting from a complex design.
(2) A commitment to subject all models used for developing job creation and retention count estimates to the following:
- Sensitivity analysis to determine if changes in key model data inputs cause key model outputs to respond in a sensible fashion;
- Model validation to analyze its performance by comparing the results to available independent information sources; and
- Demonstration of reproducibility to show that, given the same inputs, the model produces similar results.
(3) An agreement to fully document and disclose in reports to OJP (see below) all methods and models used to generate job creation and retention count estimates, as well as all data utilized as input for the model(s).
- A statement of the qualifications of the persons who will be carrying out the statistical methodology.
In its proposal, the recipient must demonstrate that the proposed methodology will be implemented by or under the supervision of a qualified statistician or highly experienced social science researcher.
- A description of the plan to monitor or verify the data used as part of the statistical methodology.
- Initial Report
In addition to the job creation and retention count estimates, the initial progress report submitted to OJP or OVW must include full documentation of all estimation methods employed as well as all data utilized by the models. This will help OJP or OVW ensure objectivity, utility, transparency, and reproducibility of the estimates and projections reported by the recipient. OJP and OVW will archive data and models so that the estimates can be independently reproduced and verified.