1205-0245 Supporting Statement Part B- 7-29-22 Final

Supporting Statement B

Collection of Information Employing Statistical Methods

1. Describe (including a numerical estimate) the potential respondent universe and any sampling or other respondent selection methods to be used. Data on the number of entities (e.g., establishments, State and local government units, households, or persons) in the universe covered by the collection and in the corresponding sample are to be provided in tabular form for the universe as a whole and for each of the strata in the proposed sample. Indicate expected response rates for the collection as a whole. If the collection had been conducted previously, include the actual response rate achieved during the last collection.

a. Respondent Universe. The respondent universe for paid and denied claims comprises fifty-two State Workforce Agencies (SWAs), claimants, employers, and third parties. Within each SWA, the universe for paid claims is defined as all intrastate and interstate weeks paid (or offset) in the State Unemployment Insurance (UI), Unemployment Compensation for Federal Employees (UCFE), and Unemployment Compensation for Ex-servicemembers (UCX) programs. For denied claims, each SWA defines three universes of formal, documented denial decisions or determinations of ineligibility for benefits. These denial decisions are based on (a) monetary issues; (b) separation issues; and (c) nonseparation, or "continuing eligibility" issues.

2. Sampling Methodology.

BAM Paid Claims

SWAs select systematic random samples of paid UI claims each week and use the results of the BAM paid claims investigations to estimate accurately the number and dollar value of proper and improper payments (overpayments and underpayments), and their rates of occurrence. BAM paid claims also provides information that can be used for program improvement, including the type of payment error, error cause, responsible party, point of detection within the system, and the actions of claimants, employers, and agencies prior to the BAM investigation.

The Department has supplied each SWA with software that performs quality assurance edits of the sampling frames and randomly selects the BAM paid claims samples.¹ Each week a random sample is selected of both intrastate and interstate original payments (including combined wage claims) made for a week of unemployment under the State UI, UCX or UCFE programs. A minimum sample of 360 cases per year is pulled in the ten states with the smallest UI program workloads (defined as average annual UI weeks paid during the most recent five calendar years) and a minimum sample of 480 cases per year in the other states. State BAM staff audit each selected claim, examining all aspects of a claimant's eligibility to receive unemployment compensation during the sampled week. In their investigation, staff verify wages used to establish monetary entitlements, the claimant's reason for being unemployed, efforts to find work, during the week and any other factors which would have affected the claimant’s entitlement to a benefit during the sampled week or the amount of the benefit paid. Effective January 2008, paid claims selected for BAM must be matched with the National Directory of New Hires. The findings are then coded and entered into a database that is maintained on a computer located in each SWA. The Department uploads state BAM results (excluding the claimant Social Security Number and other personally identifying information) to a database maintained by the ETA Office of Unemployment Insurance. The Department publishes annual performance results and uses the data for various analytical and evaluative purposes.

BAM Denied Claims

Each week, SWAs select systematic random samples from the three separate sampling frames constructed from the universes of claims for UI for which eligibility was denied for monetary, separation, or nonseparation reasons. Samples are selected using the same sampling frame edit and sample selection software used for paid claims. The Department estimates the accuracy of decisions to deny claimants UI, based on the results of the case investigations for these samples.

Investigation of BAM denied claims follows the paid claims case investigation methodology. It evaluates denials accuracy by investigating random samples of each of the three types of denials. All states sample a minimum of 150 cases of each type of denial in each calendar year. State BAM staff review agency records and contact claimants, employers, and all other relevant parties to verify information in agency records or obtain additional information pertinent to the determination that denies eligibility. Unlike the investigation of paid claims, in which all prior determinations affecting claimant eligibility for the compensated week selected for the sample are evaluated, the investigation of denied claims is limited to the issue upon which the denial determination is based.

The Department distributes a table of random start numbers to use with the BAM paid and denied claims sample selection software. A separate random number is provided for each sample pull (paid claims, monetary denials, separation denials, nonseparation denials) for each of the 52 weekly samples.

Scope: Both paid and denied intrastate and interstate liable claims in the State UI, UCFE, and UCX programs are included in the sampling frames. Paid and denied interstate claims are included in the sampling frames of the interstate liable state. The “liable” state is the state which pays the UI benefits (that is, that state’s Unemployment Trust Fund is charged). The “agent” state is the state that processes the UI claim.

Operational Definitions of Sampling Frames: Unless otherwise stated, definitions refer to those used in ET Handbook 401, 5th edition. ETA report cell references are those used in ET Handbook 402, 5th edition.

(1) Paid Weeks

Include only paid or compensated weeks that fall into all of the following: a) regular program type (UI, UCFE, UCX, or any combination thereof), b) weeks for which the payments/offsets² are original payments (defined as the first valid payment/offset made by a state agency to a claimant for that week; offsets would normally recover overpayments established for previous weeks), c) weeks for which “total” or “part-total” payments/offsets are made, and d) weeks for which payments/offsets/intercepted payments are made to intrastate claimants, to interstate claimants by the liable state, or for combined wage claims.

Exclude weeks that all waiting weeks, weeks for which supplemental payments are made, weeks with stop payments, and all weeks paid under the Short Time Compensation (STC) [Workshare], Extended Benefits (EB), Trade Readjustment Allowance (TRA), Disaster Unemployment Assistance (DUA) programs, any temporary Federal-State supplemental compensation programs, or other special programs, such as Emergency Unemployment Compensation (including CARES Act programs).

(2) Monetary Denials

Include all initial claims that meet the definition for inclusion in the ETA 5159 Claims and Activities report on lines 101 (State UI), 102 (UCFE, No UI), and 103 (UCX only), for item 2 (new intrastate, excluding transitional), item 6 (transitional), and item 7 (interstate received as liable state) and for which eligibility was denied because of:

• Insufficient wages,

• Insufficient hours/weeks/days,

• Failure of high quarter wage test,

• Requalification wage requirement, or

• Other state monetary eligibility requirement

Exclude denied claims made under the Short Time Compensation (STC) (Workshare), Extended Benefits (EB), Trade Readjustment Allowance (TRA), Disaster Unemployment Assistance (DUA), or any temporary Federal-State supplemental compensation programs including CARES Act programs.

(3) Separation Denials

Include all separation determinations that meet the definition for inclusion in the ETA 9052 Nonmonetary Determinations Time Lapse (Detection Date) report in cells c1 (intrastate), c5 (interstate), and c193 (multi-claimant) and for which eligibility was denied based on any of the following issues:

• Voluntary quit (either personal or work connected),

• Discharge,

• Labor dispute, or

• Other separation issue reportable under definitions in ET Handbook 401

Exclude denied claims made under the STC, EB, TRA, DUA, or any temporary Federal-State supplemental compensation programs.

(4) Nonmonetary-Nonseparation Denials

Include all nonmonetary-nonseparation determinations that meet the definition for inclusion in the ETA 9052 Nonmonetary Determinations Time Lapse (Detection Date) report in cells c97 (intrastate), c101 (interstate), and c193 (multiclaimant) and for which eligibility was denied based on any of the following issues:

• Able and/or available to work,

• Actively seeking work,

• Disqualifying/unreported income,

• Refusal of suitable work or offer of job referral,

• Refusal of referral to profiling services,

• Failure to report,

• Failure to register with the employment service, or

• Other nonseparation eligibility issue (for example, alien status, athlete, school employee, seasonality, removal of disqualification, and determination of whether claimant’s activities or status constitutes service or employment).

Exclude denied claims made under the STC, EB, TRA, DUA, or any temporary Federal-State supplemental compensation programs including CARES act programs.

Frequency and Timing:

SWAs create a sampling frame file each week for all four universes. For paid claims, the survey population is selected from all weeks for which payments are made or offsets applied during a period that begins at 12:00 a.m. on Sunday and ends at 11:59 p.m. on Saturday. This interval is defined by the run time(s) of the computer programs that issue the checks or apply offsets.

The sampling frame for separation and nonseparation denied claims includes all decisions to deny UI claims issued during the period 12:00 a.m. Sunday to 11:59 p.m. Saturday. The date of the determination is the date printed on the determination notice. If no notice is issued, it is the date that the denial action was entered into the agency’s record system or that a permanent stop payment order was issued.

The sampling frame for monetary denied claims is constructed slightly differently as it is possible that a UI claim may initially be denied for insufficient wages but subsequently become monetarily eligible upon the addition of wages from out-of-state employers (combined wage claims), Federal wages (UCFE and/or UCX programs), or as a result of the application of alternate base period formulas. In order to allow time for SWAs to request and receive Federal, out of state, and recently earned wage credits, the sampling frame for monetary denials is constructed two weeks after the week ending date of the initial claim. For example, the sampling frame for batch 201810 (March 4 - 10, 2018) will consist of new initial and transitional claims filed on or before February 24 for which the most recent determination issued between February 18 and March 10 denies monetary eligibility.

c. Case Investigation. BAM paid and denied claims case investigations are conducted according to the methods and procedures documented in ET Handbook 395; case investigation procedures for both paid and denied claims are described in detail in chapter VI, except as noted in chapter VIII for denied claims investigations. The information that is collected isspecified in the data collection instruments (DCIs) for both paid and denied claims.

BAM investigators collect DCI information from SWA records, claimant questionnaires, and interviews with employers and other. The parties with information relevant to the paid or denied claim. The investigator then records this information in an automated database, which consists of individual data records for each sampled paid claim and denial.

All paid and denied claims investigations involve one state investigator and one claimant. The person whose claim was either paid or denied is contacted in-person, by telephone, or by mail.

BAM investigators obtain Information from employers (and their representatives) and "third parties" -- persons other than the claimant or employer, such as a doctor, school, or labor union, who possess information pertinent to the paid or denied case.

Unlike the investigation of paid claims, in which all decisions affecting claimant eligibility that precede the compensated week selected for the sample are evaluated, the investigation of denied claims is limited to the issue type upon which the denial decision was based. For example, if a continued week claim is denied because the agency determined the claimant was not available for work, then only the availability issue will be investigated. The monetary, separation and any other nonmonetary determinations which could have affected eligibility for the week claimed will not be investigated. SWAs have the flexibility to conduct the investigation of both paid denied claims for UI by in-person interview, telephone, mail or fax, as they deem appropriate.

2. Describe the procedures for the collection of information including:

a. Stratification and Sample Selection. For both paid and denied claims, each state’s sample is stratified by week (which BAM refers to as a batch). For denied claims, samples are selected from sampling frames for each of the three types of denials (monetary, separation, and nonseparation). Systematic samples are selected weekly using software and random start numbers provided by the Department. Annual estimates are weighted to reflect the sample stratification. The formulae used to produce weighted estimates for paid and denied claims accuracy rates are in Attachment B-1.

b. Estimation Procedure. See Attachment B-1 for the formulae used to estimate paid and denied claims accuracy rates and sampling variances.

c. Degree of Accuracy Needed. The Department has adopted a standard for data publication that the 95% confidence interval (roughly two times the standard error of estimate) will be estimated and displayed for each estimated accuracy rate. Attachment B-2 displays the estimated rates and sampling errors for calendar year (IPIA) 2020 BAM paid claims results for the following types of overpayments:

Overpayment Rate - The overpayment rate is defined in UIPL No. 09-13, Change 1. It is the total weighted amount of payments determined to be overpaid divided by the weighted dollar amount paid in the BAM sample population. The rate includes fraud, nonfraud recoverable, and nonfraud nonrecoverable overpayments. It excludes payments that are technically proper due to finality, warnings issued for the failure to conduct an active search for work, or due to rules other than finality. All causes and responsible parties are included in this rate.

Underpayment Rate – The underpayment rate is defined in UIPL No. 9-13 Change 1. It is the total weighted amount of payments determined to be underpaid divided by the weighted dollar amount paid in the BAM sample population. All causes and responsible parties are included in this rate. It includes errors where additional payment is made to the claimant. It excludes those errors that are technically proper due to finality rules or technically proper due to rules other than finality.

Improper Payment Rate – This rate includes UI benefits overpaid plus UI benefits underpaid divided by the total amount of UI benefits paid. Overpayments, underpayments, and total UI benefits paid are estimated from the BAM survey results of paid UI claims in the state UI, UCFE, and UCX programs. Overpayments and underpayments determined to be technically proper under state UI law for finality and other reasons are excluded from the measure.

Agency Responsibility Rate - This rate includes overpayments for which the SWA was either solely responsible or shared responsibility with claimants, employers, or third parties, such as labor unions or private employment referral agencies. The rate includes fraud, nonfraud recoverable overpayments, and nonfraud nonrecoverable overpayments. It excludes payments that are technically proper due to finality or other rules.

Fraud Rate - The definition of unemployment compensation (UC) fraud varies from state to state – there is no federal definition of fraud in the UC program. Generally, fraud involves a knowing and willful act and/or concealment of material facts to obtain or increase benefits when benefits are not due. States vary on the level of evidence required to demonstrate a knowing and willful act or the concealment of facts. An overpayment which is classified as a fraud overpayment in one state might be determined to be a nonfraud overpayment in another state. Often fraud determinations include looking at a pattern of action or the claimant’s certification of erroneous information under the penalty of perjury. Also states differ on the implementing fraud administrative penalty determinations. In some states, a fraud determination becomes effective on the date of the fraudulent act. In other states, the administrative penalty takes effect on the determination date. Since fraud determination criteria and thresholds vary throughout the SWAs, the individual state rates reflect these differences. The rate includes all causes and responsible parties.

Attachment B-3 displays the estimated rates and sampling errors for IPIA 2020 BAM denied claims results for monetary, separation, and nonseparation issues. Improper Denial Rates - BAM estimates the percentage of claimants improperly denied benefits. This rate includes three subcategories. These subcategories are monetary denials, separation denials, and nonseparation denials. The BAM program does not assign a dollar estimate to improper denial rates; however, improper denials are corrected when permitted by law.

d. Unusual problems requiring specialized sampling procedures. BAM paid and denied claims does not involve any unusual problems requiring specialized sampling procedures.

e. Use of periodic data collection to reduce burden. Less frequent data collection cycles would not be an appropriate means for reducing burden. This issue is addressed in Part A of the Justification, section A-6. To make reliable estimates of accuracy in a highly seasonal program such as UI, sampling must occur continuously. BAM paid and denied claims samples are drawn weekly. The continuous investigation of these samples, with regular data entry, also provides up-to-date information on accuracy to facilitate continuous improvement. Because the samples are weekly, they can be aggregated over various time periods for analytical purposes.

3. Describe methods to maximize response rates and to deal with issues of non-response. The accuracy and reliability of information collected must be shown to be adequate for intended uses. For collections based on sampling, a special justification must be provided for any collection that will not yield "reliable" data that can be generalized to the universe studied.

Because claimants are required to provide information concerning their continued eligibility for UI benefits, nonresponse to the BAM claimant questionnaire can affect eligibility for benefit payments. The response rate for claimant contacts (that is, the percentage of claimant questionnaires completed) for BAM paid claims is approximately14,983 or 86.95 percent of the completed cases. It is more difficult to obtain a complete questionnaire from claimants who were denied benefits. Some of these individuals have returned to work or have relocated and are unavailable for interview.

Even if claimant information cannot be obtained directly, BAM investigators can obtain sufficient information from SWA records, and other relevant parties in order to reach an informed decision concerning the accuracy of the decision to deny benefits. The BAM investigators verify all information provided by UI recipients or obtained from automated file systems and other agency records. They contact all employers for whom the claimant worked before becoming unemployed or who provided part-time work during the claims series or were contacted in job search, as well as interested third parties, such as labor unions or employment agencies. The table shows instances where improper payment were found by claimant responsiveness.

BAM Case Error Rates By Claimant Responsiveness
July 1, 2019 through March 31, 2020 (Batch Range 201927 through 202013)
		Number	Percent	# Cases	Percent
ST	Claimant	Cases in	Response	With	cases
	Response	Sample	Rates	Errors	IMPROPER*
US	Responded	14,983	86.95%	1,647	10.99%
US	Nonresponse	2,249	13.05%	468	20.81%
US	Total Cases	17,232	100.00%	2,115	12.27%
Overpayment Cause Counts and Percentages may reflect more than one payment error on a given case completed

In IPIA 2020, although the percentage of claimant questionnaires completed varied considerably by sample type, states were able to complete nearly all of their cases based on agency documentation, employer, and third party information. The following table summarizes claimant response by data collection method. Attachment B-4 displays the response rates for the IPIA 2020 BAM paid claims samples, and Attachment B-5 displays the response rates for the IPIA 2020 BAM denied claims samples.

BAM Case Completion and Claimant Interview Method -- IPIA 2020
Sample Type	Cases Sampled	Valid Cases*	Cases Completed**	Percent Complete	In-Person	Tele-Phone	Mail	No Clmnt. Inter.
Paid Claims	17,681	17,673	17,232	97.50%	2.77%	38.97%	45.20%	13.05%
Monetary	5,795	5,695	5,533	97.16%	0.45%	36.21%	18.47%	44.56%
Separation	5,733	5,701	5,594	98.12%	0.46%	36.45%	23.22%	39.73%
Nonseparation	5,782	5,728	5,594	97.66%	0.62%	40.57%	27.40%	31.28%
* Cases sampled minus cases deleted because they did not meet the definition for inclusion in the survey population and denied claims that were withdrawn by the claimant. **To meet IPIA reporting timetables, the database was frozen on 10/29/2020. The number of valid cases completed is those signed off by the BAM program’s supervisor by the close of business on 10/28/2020. (Note: As a result of COVID-19 pandemic, the sample population only includes three quarters of data).

The Department is acutely aware of the importance of claimant response to the BAM questionnaire and has established a Federal-State workgroup to examine the issue of claimant nonresponse. The Department has drafted an advisory, which is currently in Department clearance, to issue guidance to address the specific issues of adjudicating work search and reporting errors when the claimant fails to respond to the BAM audit questionnaire.

In addition, in order to reduce nonresponse error and maintain coding consistency, the Department will continue to conduct training for BAM supervisors and investigators and hold Federal-State peer reviews of completed BAM audits to ensure that coding accurately reflects state law and policy and that states are following the BAM methodology.

In order to reduce respondent burden and maximize claimant response, the number of data elements collected for DCA is significantly smaller than the amount of data collected for BAM paid claims. Because only information relevant to the monetary, separation, or nonseparation denial issue is verified, the number of data elements per case is one-third or less of the number collected for BAM paid claims, which investigates decisions at all three points in the UI claims process. In addition, SWAs follow up the initial claimant contact with a sufficient number of call-backs and re-contact attempts to demonstrate that a reasonable attempt was made to obtain the information.

SWAs administering the BAM program are encouraged to:

• Use all available data collection methods -- in-person, telephone, mail, e-mail, and fax -- to complete their investigations;

• Be as flexible as feasible in accommodating the schedules of claimants, employers, and other relevant parties;

• Develop clear and concise questionnaires and scripts which clearly explain the purpose of the data collection effort and minimize the time commitment of the respondent. To this end the Department shares examples and prototype case investigation materials in order to disseminate best practices as widely as possible;

• Clearly inform the respondents that the privacy of the information they provide will be strictly maintained and that any information that can identify an individual, such as a claimant’s social security number, will not be shared with the Department’s or any other state’s record systems; and

• Emphasize to respondents that the major objective of the BAM program is the improvement of the UI system, and that their cooperation will contribute to insuring that individuals who are in fact eligible for UI benefits receive them.

4. Describe any tests of procedures or methods to be undertaken. Testing is encouraged as an effective means of refining collections of information to minimize burden and improve utility. Tests must be approved if they call for answers to identical questions from 10 or more respondents. A proposed test or set of test may be submitted for approval separately or in combination with the main collection of information.

Paid Claims

In 1991 the Department of Labor completed a pilot test of the feasibility and cost-effectiveness of telephone contacts in lieu of in-person interviews with claimants, employers, and third parties. Four states participated in the pilot test, giving a wide range of economic, social and geographical environments. The pilot showed that the telephone was reasonably effective in detecting overpayment and underpayment errors: the patterns of erroneous payments by type and cause were basically the same as detected by the in-person control investigations. Although the rate of dollars overpaid discovered by the two methods in one state was virtually identical, in the other three the telephone estimate was only 60% of the in-person estimate. The pilot showed that the telephone methodology was very effective for certain aspects of BAM investigations, but less so for others. It also showed that BAM investigations could be done considerably less expensively by telephone--at about half the cost, confirming the estimate from a similar pilot project conducted in Idaho in the late 1980s.

Denied Claims

In 1987 the Department completed a five-state pilot test of using the BAM field-check methodology for determining the accuracy of benefit denial decisions. Three different sampling designs were evaluated in the 1986-87 pilot: (1) separate sampling frames for monetary, separation, and nonseparation (continuing eligibility) denials and a single sampling frame for all paid claims; (2) separate sampling frames for denials and decisions to affirm eligibility at the monetary, separation, and nonseparation points of determination in the UI claims process; and (3) a longitudinal approach, in which claimants were sampled at the time that the initial claim was filed, and eligibility determinations (either to deny or affirm eligibility) were investigated as they occurred during the claims process. The 1997-98 DCA pilot was based on model 1, which was the simplest design and preserved the design used for BAM paid claims. As noted in Part A, the Department has relied on results of the 1997-98 DCA pilot to estimate case-completion times and burden hours for national implementation of DCA.

5. Provide the name and telephone number of individuals consulted on statistical aspects of the design and the name of the agency unit, contractor(s), grantee(s), or other person(s) who will actually collect and/or analyze the information for the agency.

The following individual consulted on statistical aspects of the design.

Andy Spisak

571 481-0450

The following individual collect and analyze the paid and denied claims data and may be contacted for further information:

Ross Miller, Employment and Training Administration, Office of Unemployment Insurance

Phone: 202-693-3178, E-mail: [email protected]

Estimation Procedure for Benefit Accuracy Measurement

BAM Paid Claims

1. Ratio Estimate of Overpayment Rate

The parameter to be estimated, R_o, is the ratio of Unemployment Insurance (UI) benefits overpaid to total UI benefits paid: R_o = Y/X, where Y = Total dollars overpaid in the population and X = Total UI benefits paid in the population.

R_o is estimated by the sample ratio:

r_o =

where:

H = Number of batches (weekly samples) in the period for which the estimate is being made.

N_h = Total number of UI payments in the population for batch h. (Note: This value is available from state automated record systems and does not have to be estimated.)

m_h = Number of completed sample cases in batch h.

x_hi = Amount of UI benefits paid/offset for the i^th case in batch h.

y_hi = Dollars overpaid for the i^th case in batch h.

Nonresponse is assumed to be random.

2. Sampling Variance of Ratio Estimate of Overpayment Rate

The following formula will be used to estimate the sampling variance of the ratio estimate of the BAM paid claims overpayment rate.

(Note: Because the sampling fractions, f_h=m_h/N_h, are negligible, the term (1-f_h) has been omitted from the equations.)

estVar(r_o)=

=

where:

is the sample variance of the dollars overpaid;

is the sample variance of the dollars paid/offset; and

is the sample covariance of the dollars overpaid and the dollars paid/offset.

X = Total population dollars paid/offset for the H batches.

(Note: This value is available from state automated record systems and does not have to be estimated.)

3. Ratio Estimate of Overpayment Rate for Subgroups

The parameter to be estimated, R_ok, is the ratio of Unemployment Insurance (UI) benefits overpaid to total UI benefits paid for population subgroup k: R_ok = Y_k/X_k, where Y_k=Total dollars overpaid in the population for the k^th subgroup and X_k=Total UI benefits paid in the population for the k^th subgroup.

R_ok is estimated by the sample ratio:

r_ok =

where:

x_hik = Amount of UI benefits paid/offset for the i^th case in the k^th subgroup in batch h.

x_hik = x_hi, for hi in the k^th subgroup, and

x_hik = 0, for hi not in the k^th subgroup

y_hik = Dollars overpaid for the i^th case in the k^th subgroup in batch h.

y_hik = y_hi, for hi in the k^th subgroup, and

y_hik = 0, for hi not in the k^th subgroup

Nonresponse is assumed to be random.

4. Sampling Variance of Ratio Estimate of Overpayment

Rate for Subgroups

The following formula will be used to estimate the sampling variances of the ratio estimate of the overpayment rate for subgroups.

(Note: Because the sampling fractions, f_h=m_h/N_h, are negligible, the term (1-f_h) has been omitted from the equations.)

estVar(r_ok)=

where:

is the sample variance of the dollars overpaid in the k^th subgroup;

is the sample variance of the dollars paid/offset in the k^th subgroup; and

is the sample covariance of the dollars overpaid and the dollars paid/offset.

is the estimated total dollars paid/offset for the H batches.

In the preceding formulas,

x_hik = x_hi, for hi in the kth subgroup, and

x_hik = 0, for hi not in the kth subgroup;

y_hik = y_hi, for hi in the kth subgroup, and

y_hik = 0, for hi not in the kth subgroup

x_hk = Amount of UI benefits paid/offset in the k^th subgroup in the sample in batch h.

5. Ratio Estimate of Proper Payment Rate

The parameter to be estimated, R_p, is the ratio of Unemployment Insurance (UI) benefits properly paid to total UI benefits paid: R_p = Z/X, where Z = Total dollars properly paid in the population and X = Total UI benefits paid in the population.

R_p is estimated by the sample ratio:

r_p =

where H, N_h, m_h, and x_hi are defined as in 1., above, and

z_hi = Dollars properly paid (dollars paid - dollars overpaid) for the i^th case in batch h.

6. Sampling Variance of Ratio Estimate of Proper Payment Rate

The following formula will be used to estimate the sampling variance of the ratio estimate of the BAM paid claims proper payment rate.

(Note: Because the sampling fractions, f_h=m_h/N_h, are negligible, the term (1-f_h) has been omitted from the equations.)

estVar(r_p) =

where H, N_h, m_h, X, and s²_xh are defined as in 1. and 2., above;

s²_zh is the sample variance of the dollars properly paid; and

s_zxh is the sample covariance of the dollars properly paid and dollars paid.

7. Ratio Estimate of Proper Payment Rate for Subgroups

The parameter to be estimated, R_pk, is the ratio of Unemployment Insurance (UI) benefits properly paid to total UI benefits paid for population subgroup k: R_pk = Z_k/X_k, where Z_k=Total dollars properly paid in the population for the k^th subgroup and X_k=Total UI benefits paid in the population for the k^th subgroup.

R_pk is estimated by the sample ratio r_pk which is defined as the estimator r_ok in section 3, above, except that:

z_hik = Dollars properly paid (dollars paid - dollars overpaid) for the i^th case in the k^th subgroup in batch h.

z_hik = z_hi, for hi in the k^th subgroup, and

z_hik = 0, for hi not in the k^th subgroup

8. Sampling Variance of Ratio Estimate of Proper Payment

Rate for Subgroups

The following formula will be used to estimate the sampling variances of the ratio estimate of the proper payment rate for subgroups.

(Note: Because the sampling fractions, f_h=m_h/N_h, are negligible, the term (1-f_h) has been omitted from the equations.)

estVar(r_pk)=

where H, N_h, m_h, , and s²_xh(k) are defined as in 1.and 4., above;

s²_zh(k) is the sample variance of the dollars properly paid in the k^th subgroup; and

s_zxh(k) is the sample covariance of the dollars properly paid and dollars paid in the k^th subgroup.

9. Ratio Estimate of Underpayment Rate

The parameter to be estimated, R_u is the ratio of Unemployment Insurance (UI) benefits underpaid to total UI benefits paid: R_u = U/X, where U = Total dollars underpaid in the population and X = Total UI benefits paid in the population.

R_u is estimated by the sample ratio:

r_u=

where H, N_h, m_h, and x_hi are defined as in 1., above, and

u_hi = Dollars underpaid for the i^th case in batch h.

10. Sampling Variance of Ratio Estimate of Underpayment Rate

The following formula will be used to estimate the sampling variance of the ratio estimate of the BAM paid claims underpayment rate.

(Note: Because the sampling fractions, f_h=m_h/N_h, are negligible, the term (1-f_h) has been omitted from the equations.)

estVar(r_u) =

where H, N_h, m_h, X, and s²_xh are defined as in 1. and 2., above;

s²_uh is the sample variance of the dollars underpaid; and

s_uxh is the sample covariance of the dollars underpaid and dollars paid.

11. Ratio Estimate of Underpayment Rate for Subgroups

The parameter to be estimated, R_uk, is the ratio of Unemployment Insurance (UI) benefits underpaid to total UI benefits paid for population subgroup k: R_uk = U_k/X_k, where U_k=Total dollars underpaid in the population for the k^th subgroup and X_k=Total UI benefits paid in the population for the k^th subgroup.

R_uk is estimated by the sample ratio r_uk which is defined as the estimator r_ok in section 3, above, except that:

u_hik = Dollars underpaid for the i^th case in the k^th subgroup in batch h.

u_hik = u_hi, for hi in the k^th subgroup, and

u_hik = 0, for hi not in the k^th subgroup

12. Sampling Variance of Ratio Estimate of Underpayment

Rate for Subgroups

The following formula will be used to estimate the sampling variances of the ratio estimate of the underpayment rate for subgroups.

(Note: Because the sampling fractions, f_h=m_h/N_h, are negligible, the term (1-f_h) has been omitted from the equations.)

estVar(r_uk)=

where H, N_h, m_h, , and s²_xh(k) are defined as in 1.and 4., above;

s²_uh(k) is the sample variance of the dollars underpaid in the k^th subgroup; and

s_uxh(k) is the sample covariance of the dollars underpaid and dollars paid in the k^th subgroup.

Confidence Intervals

The 95% confidence interval for any estimated ratio r_θ (1, 3, 5, 7, 9, or 11, above) is:

Coefficient of Variation

The coefficient of variation (cv) of an estimate r_θ is:

BAM Denied Claims

Equations for Case Error Estimates

The following notation will be used:

H = the number of weeks (batches) in the period for which the estimate is

being made.

N_h = the number of denied claims in week h.

X_h = the number of claims in week h which were erroneously denied.

P_h = X_h/N_h = the proportion of claims in week h which were erroneously

denied.

N_ = = total number of denied claims in the period.

X_ = = total number of claims erroneously denied in the period.

The parameter to be estimated, P, is the proportion of claims erroneously denied during the period. Estimates will be made for each of the three denial universes -- monetary, separation, and nonseparation. We wish to estimate:

P = X_/N_ =

Now let

m_h = the number of completed sample claims for week h.

m_ = = total number of completed sample claims in the period.

x_h = the number of claims in week h which were erroneously denied.

= proportion of sample claims in week h which were erroneously denied.

If it is assumed that non-response is "at random", then .

It follows that is unbiased for P. Furthermore, as sampling is independent within each week (stratum), it follows that

where f_h = m_h/N_h. The usual estimator for is

.

If f_h is negligible then

can be used for variance estimation.

Proportions for Subgroups

The proportion of denial actions which were incorrectly decided may be estimated for population subgroups, for example UI program (State UI, UCFE, UCX), filing method (in-person, telephone, mail), or demographic classifications.

Building on the notation above, for the k^th subgroup and the h^th week let

N_hk = the number of denied claims.

X_hk = the number of claims were erroneously denied.

P_hk = X_hk/N_hk = the proportion of claims which were erroneously denied.

Then for the k^th subgroup we have

N_k = = total number of denied claims in the period.

X_k = = total number of claims erroneously denied in the period.

The parameter to be estimated, P_k, is the proportion of claims erroneously denied during the period for subgroup k. Analogous to previous work, we can write

P_k = X_k/N_k = .

Note that neither X_k nor N_k is known. For the k^th subgroup, h^th week, let

m_hk = the number of completed sample claims for week h.

x_hk = the number of claims in week h which were erroneously denied.

Assuming nonresponse is "at random", is unbiased for X_k and is unbiased for N_k. The ratio estimator is approximately unbiased for P_k, and

where f_hk = m_hk/N_hk and θ_hk = N_hk/N_h. Assuming that f_hk is negligible, an estimate for the variance is given by

where

and

{  .

Confidence Intervals

The 95% confidence interval for any estimate (u) is:

Coefficient of Variation

The coefficient of variation (cv) of an estimate u is:

UI Benefit Accuracy Measurement Rates
Batch Range 201927 through 202026
ST	Sample	Total Amount of benefits compensated in population sampled	Overpay-ment (OP) Rate (a)	OP Rate 95% CI +/-	Underpay-ment (UP) Rate (c)	UP Rate 95% CI +/-	Improper Payment Rate (OP+UP) [(a)+(c)]	Fraud Rate	Fraud Rate 95% CI +/-	Agency Responsible Rate	AGY Resp Rate 95% CI +/-
US	17,232	$20,448,076,675	8.722%	0.705%	0.445%	0.098%	9.167%	3.167%	0.520%	0.905%	0.214%
AK	361	$59,986,181	8.158%	2.885%	0.954%	0.569%	9.112%	1.487%	1.413%	0.553%	0.845%
AL	362	$106,304,296	10.420%	3.261%	0.256%	0.219%	10.676%	2.565%	1.750%	1.464%	1.300%
AR	358	$76,782,664	4.630%	2.547%	0.640%	0.500%	5.270%	1.782%	1.411%	1.526%	2.026%
AZ	357	$136,251,651	10.808%	3.389%	0.007%	0.010%	10.815%	3.314%	1.894%	1.265%	1.152%
CA	569	$3,679,853,734	9.921%	2.631%	0.492%	0.341%	10.413%	6.931%	2.310%	1.157%	0.921%
CO	316	$265,180,370	7.271%	3.180%	0.455%	0.366%	7.726%	0.358%	0.598%	0.354%	0.642%
CT	362	$454,965,736	11.075%	3.671%	0.213%	0.211%	11.289%	1.124%	0.718%	3.350%	2.102%
DC	359	$83,349,445	8.384%	2.765%	0.288%	0.335%	8.672%	1.622%	1.342%	0.863%	0.988%
DE	270	$50,937,009	5.423%	2.860%	0.099%	0.125%	5.522%	0.157%	0.311%	2.325%	2.110%
FL	356	$230,034,550	9.224%	3.028%	0.126%	0.216%	9.350%	1.368%	1.254%	5.460%	2.506%
GA	360	$246,377,147	4.412%	2.730%	0.000%	0.000%	4.412%	0.302%	0.596%	1.118%	1.072%
HI	263	$100,835,590	4.660%	2.867%	0.160%	0.168%	4.820%	1.244%	1.845%	0.393%	0.775%
IA	360	$309,734,838	5.231%	2.037%	0.131%	0.142%	5.362%	0.895%	0.899%	1.118%	1.131%
ID	273	$67,106,593	4.481%	2.515%	0.123%	0.190%	4.604%	1.617%	1.429%	0.996%	1.507%
IL	360	$1,272,294,270	8.739%	3.104%	1.212%	0.688%	9.951%	2.661%	1.878%	0.462%	0.516%
IN	333	$177,663,185	8.193%	3.090%	0.000%	0.000%	8.193%	0.000%	0.000%	3.538%	2.079%
KS	353	$98,043,883	12.919%	3.629%	0.010%	0.019%	12.929%	1.455%	1.506%	0.692%	0.824%
KY	316	$228,426,788	7.412%	2.646%	0.119%	0.144%	7.531%	2.473%	1.644%	0.608%	0.846%
LA	349	$94,064,159	9.500%	3.557%	0.302%	0.482%	9.802%	4.447%	2.730%	2.481%	2.342%
MA	348	$962,100,388	6.853%	2.381%	0.780%	0.657%	7.633%	2.966%	1.535%	1.093%	0.826%
MD	360	$309,285,680	5.914%	2.394%	0.041%	0.057%	5.955%	1.507%	1.024%	0.643%	0.810%
ME	276	$63,421,941	3.889%	3.068%	0.333%	0.301%	4.222%	0.814%	0.892%	1.308%	2.286%
MI	348	$569,696,101	30.371%	5.032%	0.126%	0.117%	30.497%	1.209%	1.085%	0.442%	0.638%
MN	313	$570,444,852	6.937%	2.910%	0.055%	0.077%	6.992%	2.304%	1.595%	0.000%	0.000%

UI Benefit Accuracy Measurement Rates
Batch Range 201927 through 202026
ST	Sample	Total Amount of benefits compensated in population sampled	Overpay-ment (OP) Rate (a)	OP Rate 95% CI +/-	Underpay-ment (UP) Rate (c)	UP Rate 95% CI +/-	Improper Payment Rate (OP+UP) [(a)+(c)]	Fraud Rate	Fraud Rate 95% CI +/-	Agency Responsible Rate	AGY Resp Rate 95% CI +/-
MO	360	$175,221,994	5.337%	2.320%	0.000%	0.000%	5.337%	2.011%	1.389%	0.000%	0.000%
MS	366	$46,170,446	7.614%	2.999%	0.141%	0.242%	7.755%	4.443%	2.111%	0.620%	0.915%
MT	270	$82,272,140	4.459%	1.993%	0.084%	0.090%	4.544%	0.788%	0.778%	0.726%	1.068%
NC	390	$191,574,760	15.145%	2.295%	0.420%	0.316%	15.565%	2.429%	0.950%	0.248%	0.235%
ND	232	$61,907,303	8.401%	4.082%	0.079%	0.124%	8.480%	2.766%	3.346%	0.475%	0.637%
NE	269	$47,614,113	6.559%	3.158%	0.222%	0.418%	6.780%	1.017%	1.330%	2.397%	2.175%
NH	273	$30,113,746	3.767%	2.384%	0.563%	0.518%	4.330%	0.470%	0.950%	0.000%	0.000%
NJ	340	$1,360,072,672	5.784%	2.134%	1.432%	0.502%	7.216%	0.401%	0.575%	0.351%	0.392%
NM	336	$99,978,433	5.637%	2.376%	0.128%	0.141%	5.765%	3.739%	2.031%	1.018%	1.058%
NV	244	$143,774,810	6.462%	3.258%	0.083%	0.135%	6.545%	4.122%	2.690%	0.949%	1.290%
NY	363	$1,755,610,315	8.863%	2.946%	0.328%	0.248%	9.191%	4.614%	2.134%	0.129%	0.219%
OH	340	$608,897,999	6.298%	2.267%	0.160%	0.161%	6.459%	0.450%	0.605%	1.086%	1.138%
OK	368	$206,233,171	3.820%	2.119%	0.111%	0.173%	3.931%	2.115%	1.626%	1.045%	0.960%
OR	350	$371,494,915	8.703%	2.970%	0.873%	0.725%	9.576%	3.665%	1.893%	1.315%	1.297%
PA	362	$1,396,896,112	8.981%	2.870%	0.242%	0.383%	9.223%	4.319%	2.070%	0.822%	0.933%
PR	325	$96,002,481	5.459%	2.339%	0.588%	0.356%	6.047%	2.405%	1.613%	5.217%	2.332%
RI	340	$101,788,595	15.717%	4.210%	0.307%	0.263%	16.024%	2.638%	1.448%	0.972%	0.659%
SC	352	$114,088,024	11.804%	3.375%	0.075%	0.106%	11.879%	6.813%	2.590%	0.360%	0.534%
SD	267	$20,658,822	1.617%	1.102%	0.026%	0.031%	1.643%	0.753%	0.861%	0.187%	0.360%
TN	361	$138,001,157	11.259%	3.279%	0.039%	0.058%	11.299%	2.063%	1.434%	2.886%	1.823%
TX	369	$1,664,577,771	4.258%	2.049%	0.256%	0.291%	4.514%	1.625%	1.378%	0.286%	0.563%
UT	362	$107,318,873	4.283%	2.225%	0.126%	0.218%	4.408%	1.027%	0.915%	0.233%	0.477%
VA	350	$206,867,859	9.944%	3.146%	0.076%	0.098%	10.021%	0.340%	0.399%	2.632%	1.652%
VT	58	$17,223,626	6.163%	5.767%	0.290%	0.412%	6.453%	2.481%	4.018%	1.934%	2.790%

UI Benefit Accuracy Measurement Rates
Batch Range 201927 through 202026
ST	Sample	Total Amount of benefits compensated in population sampled	Overpay-ment (OP) Rate (a)	OP Rate 95% CI +/-	Underpay-ment (UP) Rate (c)	UP Rate 95% CI +/-	Improper Payment Rate (OP+UP) [(a)+(c)]	Fraud Rate	Fraud Rate 95% CI +/-	Agency Responsible Rate	AGY Resp Rate 95% CI +/-
WA	343	$716,963,885	9.661%	3.906%	0.118%	0.166%	9.779%	2.600%	2.062%	0.984%	1.234%
WI	334	$275,545,974	18.564%	5.228%	0.296%	0.510%	18.860%	4.140%	2.769%	0.175%	0.290%
WV	356	$155,279,070	4.246%	2.214%	0.676%	0.616%	4.922%	1.060%	1.377%	0.802%	0.826%
WY	270	$42,786,558	5.135%	2.961%	0.806%	0.916%	5.941%	1.500%	1.780%	0.054%	0.107%
These data are based on a completion rate of 97.50% and are subject to change upon completion of the remaining cases.
Prepared by: ETA Office of Unemployment Insurance on 30 Nov 20
Note: 95% C.I. is the 95 percent confidence interval for the estimated rate. The interval is the range between the rate minus the value in the 95% C.I. column and the rate plus the value in the 95% C.I. column. For example, the interval for 10.0% +/- 2.5 is 7.5% to 12.5%. The true rate is expected to lie within 95 percent of the intervals constructed from repeated samples of the same size and selected in the same manner as the BAM PCA sample.

BENEFIT ACCURACY MEASUREMENT
DENIED CLAIMS ACCURACY
IMPROPER DENIAL RATES REPORT
Batch Range: 201927~202026
		Type		Adjusted
		Population	Cases	Improper	95% C.I	Improper	95% C.I
State:	Denial Type	of Denials	Completed*	Denial	(+/-)	Denial**	(+/-)
United States	Monetary	547,920	5,271	17.68%	1.20%	14.19%	1.13%
United States	Separation	972,218	5,585	10.25%	0.75%	7.05%	0.61%
United States	Nonseparation	1,796,096	5,589	13.11%	0.80%	9.36%	0.70%
Alaska	Monetary	1,954	99	4.50%	4.42%	2.98%	3.31%
Alaska	Separation	6,304	107	5.16%	4.26%	4.40%	4.00%
Alaska	Nonseparation	18,428	105	11.08%	6.10%	11.08%	6.10%
Alabama	Monetary	9,363	107	2.70%	3.06%	1.90%	2.63%
Alabama	Separation	15,364	110	2.46%	2.79%	2.46%	2.79%
Alabama	Nonseparation	14,794	112	0.84%	1.64%	0.84%	1.64%
Arkansas	Monetary	1,079	103	10.49%	5.94%	8.73%	5.61%
Arkansas	Separation	12,517	112	0.89%	1.75%	0.89%	1.75%
Arkansas	Nonseparation	8,358	112	1.65%	2.27%	1.65%	2.27%
Arizona	Monetary	37,572	105	0.88%	1.73%	0.88%	1.73%
Arizona	Separation	15,191	105	2.61%	2.98%	1.91%	2.65%
Arizona	Nonseparation	16,121	105	4.13%	3.20%	4.13%	3.20%
California	Monetary	85,010	148	40.98%	8.85%	31.01%	8.45%
California	Separation	120,841	159	13.96%	5.28%	13.96%	5.28%
California	Nonseparation	185,786	151	27.56%	7.55%	24.67%	7.38%
Colorado	Monetary	1,108	76	32.58%	10.27%	27.34%	8.91%
Colorado	Separation	29,591	102	6.87%	5.10%	6.07%	4.85%
Colorado	Nonseparation	16,849	102	7.64%	5.99%	5.67%	5.33%
Connecticut	Monetary	3,193	116	7.82%	5.64%	3.73%	3.85%
Connecticut	Separation	9,616	116	11.54%	5.96%	10.30%	5.87%
Connecticut	Nonseparation	11,940	116	9.74%	5.14%	3.34%	3.30%
District of Columbia	Monetary	1,448	93	18.00%	8.02%	13.95%	7.26%
District of Columbia	Separation	2,254	111	5.50%	4.31%	3.62%	3.47%
District of Columbia	Nonseparation	7,954	110	10.70%	5.30%	6.63%	4.48%
Delaware	Monetary	1,426	113	5.50%	4.25%	3.77%	3.58%
Delaware	Separation	1,462	115	5.67%	4.43%	1.66%	2.21%
Delaware	Nonseparation	3,235	116	0.98%	1.89%	0.98%	1.89%
Florida	Monetary	19,204	114	5.74%	4.28%	5.74%	4.28%
Florida	Separation	20,472	116	11.13%	6.49%	7.15%	5.35%
Florida	Nonseparation	33,629	115	8.33%	5.39%	8.33%	5.39%
Georgia	Monetary	12,281	112	44.15%	11.02%	37.80%	10.79%
Georgia	Separation	30,917	117	5.88%	4.44%	3.38%	3.40%
Georgia	Nonseparation	21,260	117	2.23%	2.56%	2.23%	2.56%

BENEFIT ACCURACY MEASUREMENT
DENIED CLAIMS ACCURACY
IMPROPER DENIAL RATES REPORT
Batch Range: 201927~202026
		Type		Adjusted
		Population	Cases	Improper	95% C.I	Improper	95% C.I
State:	Denial Type	of Denials	Completed*	Denial	(+/-)	Denial**	(+/-)
Hawaii	Monetary	554	85	11.60%	7.12%	11.60%	7.12%
Hawaii	Separation	2,561	83	16.49%	8.15%	10.18%	5.29%
Hawaii	Nonseparation	8,848	88	16.49%	6.93%	16.49%	6.93%
Iowa	Monetary	6,292	89	27.52%	12.68%	25.42%	12.68%
Iowa	Separation	11,558	102	11.38%	5.72%	9.33%	5.38%
Iowa	Nonseparation	14,124	98	18.40%	8.61%	13.51%	7.34%
Idaho	Monetary	1,620	103	10.37%	6.22%	9.82%	6.13%
Idaho	Separation	4,117	105	5.35%	4.70%	5.35%	4.70%
Idaho	Nonseparation	16,177	105	15.02%	7.18%	10.65%	6.21%
Illinois	Monetary	7,107	91	21.96%	8.25%	20.02%	7.79%
Illinois	Separation	40,008	106	17.48%	7.55%	9.12%	5.03%
Illinois	Nonseparation	64,851	105	17.46%	7.69%	15.49%	7.48%
Indiana	Monetary	12,309	104	7.17%	5.38%	7.17%	5.38%
Indiana	Separation	15,002	104	9.45%	6.02%	8.86%	5.91%
Indiana	Nonseparation	60,739	104	10.40%	6.45%	7.99%	5.46%
Kansas	Monetary	4,009	109	6.76%	4.52%	6.76%	4.52%
Kansas	Separation	11,947	115	8.93%	4.96%	6.54%	4.59%
Kansas	Nonseparation	15,364	112	8.23%	5.92%	8.23%	5.92%
Kentucky	Monetary	7,923	109	10.77%	6.26%	9.90%	6.02%
Kentucky	Separation	12,633	111	4.45%	4.02%	2.14%	2.98%
Kentucky	Nonseparation	12,062	143	2.78%	3.31%	2.30%	3.17%
Louisiana	Monetary	9,386	110	8.32%	5.03%	7.53%	4.79%
Louisiana	Separation	10,383	111	8.34%	5.53%	3.35%	3.40%
Louisiana	Nonseparation	29,529	111	7.06%	5.03%	3.89%	3.87%
Massachusetts	Monetary	13,888	88	36.86%	12.39%	33.37%	12.03%
Massachusetts	Separation	15,968	91	18.18%	9.55%	13.24%	8.79%
Massachusetts	Nonseparation	47,845	88	16.05%	7.43%	15.14%	7.21%
Maryland	Monetary	6,228	118	14.07%	6.45%	10.47%	5.74%
Maryland	Separation	26,384	116	11.29%	6.05%	9.60%	5.92%
Maryland	Nonseparation	26,738	115	5.70%	4.29%	5.24%	4.20%
Maine	Monetary	1,272	109	25.16%	9.28%	21.25%	8.68%
Maine	Separation	2,787	111	9.87%	4.01%	0.51%	0.97%
Maine	Nonseparation	5,035	115	5.67%	4.00%	3.85%	3.33%
Michigan	Monetary	19,051	102	15.10%	6.71%	14.06%	6.61%
Michigan	Separation	38,787	99	14.14%	7.23%	9.44%	5.54%
Michigan	Nonseparation	128,628	98	10.22%	5.63%	2.88%	3.33%

BENEFIT ACCURACY MEASUREMENT
DENIED CLAIMS ACCURACY
IMPROPER DENIAL RATES REPORT
Batch Range: 201927~202026
		Type		Adjusted
		Population	Cases	Improper	95% C.I	Improper	95% C.I
State:	Denial Type	of Denials	Completed*	Denial	(+/-)	Denial**	(+/-)
Minnesota	Monetary	3,818	92	21.12%	9.29%	16.23%	8.80%
Minnesota	Separation	13,323	95	22.11%	8.89%	8.81%	6.58%
Minnesota	Nonseparation	51,614	93	6.23%	4.65%	3.82%	3.21%
Missouri	Monetary	11,553	113	2.68%	2.95%	1.90%	2.61%
Missouri	Separation	24,880	113	14.00%	7.23%	1.73%	2.49%
Missouri	Nonseparation	42,463	112	44.37%	10.47%	10.31%	5.66%
Mississippi	Monetary	2,925	102	8.80%	6.13%	7.15%	5.21%
Mississippi	Separation	10,664	103	3.90%	4.03%	3.34%	3.88%
Mississippi	Nonseparation	25,926	103	8.83%	5.05%	6.36%	4.28%
Montana	Monetary	1,377	94	5.24%	4.31%	3.01%	2.88%
Montana	Separation	3,832	109	3.09%	2.72%	1.47%	2.01%
Montana	Nonseparation	7,981	110	10.82%	5.63%	9.65%	5.14%
North Carolina	Monetary	10,894	111	17.28%	7.72%	11.99%	6.41%
North Carolina	Separation	18,842	117	8.61%	5.64%	3.45%	3.21%
North Carolina	Nonseparation	27,773	117	27.00%	7.42%	22.92%	6.82%
North Dakota	Monetary	1,743	94	13.75%	7.79%	8.26%	5.53%
North Dakota	Separation	2,692	96	10.22%	5.99%	4.97%	4.36%
North Dakota	Nonseparation	9,342	95	4.63%	4.60%	3.62%	4.16%
Nebraska	Monetary	1,426	104	12.10%	6.08%	8.68%	4.89%
Nebraska	Separation	14,705	107	9.27%	5.49%	6.50%	4.78%
Nebraska	Nonseparation	14,580	106	18.84%	7.86%	13.68%	6.81%
New Hampshire	Monetary	592	114	19.30%	7.22%	13.06%	5.79%
New Hampshire	Separation	1,604	116	9.83%	5.70%	1.70%	2.28%
New Hampshire	Nonseparation	6,119	116	6.97%	4.88%	5.22%	4.21%
New Jersey	Monetary	19,650	103	13.85%	6.18%	8.75%	5.06%
New Jersey	Separation	34,197	111	10.01%	5.68%	5.51%	4.45%
New Jersey	Nonseparation	36,051	110	9.65%	6.06%	6.21%	4.60%
New Mexico	Monetary	2,111	100	45.01%	11.74%	41.34%	11.66%
New Mexico	Separation	4,911	108	14.69%	7.05%	7.64%	5.11%
New Mexico	Nonseparation	14,178	108	4.70%	3.83%	3.60%	3.52%
Nevada	Monetary	3,130	80	28.64%	10.12%	21.35%	9.27%
Nevada	Separation	14,965	90	4.25%	4.34%	0.79%	1.55%
Nevada	Nonseparation	16,796	89	7.41%	4.97%	6.43%	4.58%
New York	Monetary	51,200	102	28.43%	38.06%	25.99%	38.02%
New York	Separation	47,612	110	11.44%	5.65%	7.57%	5.28%
New York	Nonseparation	108,440	111	11.15%	6.42%	4.74%	4.29%

BENEFIT ACCURACY MEASUREMENT
DENIED CLAIMS ACCURACY
IMPROPER DENIAL RATES REPORT
Batch Range: 201927~202026
		Type		Adjusted
		Population	Cases	Improper	95% C.I	Improper	95% C.I
State:	Denial Type	of Denials	Completed*	Denial	(+/-)	Denial**	(+/-)
Ohio	Monetary	18,162	105	16.93%	7.45%	11.09%	5.76%
Ohio	Separation	23,851	105	9.46%	5.92%	5.91%	4.79%
Ohio	Nonseparation	70,729	105	12.42%	5.78%	7.89%	4.77%
Oklahoma	Monetary	6,699	115	5.76%	4.31%	4.81%	4.31%
Oklahoma	Separation	13,819	121	5.78%	4.19%	0.00%	0.00%
Oklahoma	Nonseparation	40,110	119	1.07%	2.09%	1.07%	2.09%
Oregon	Monetary	5,921	107	24.40%	7.83%	15.73%	6.80%
Oregon	Separation	14,017	115	8.84%	4.95%	6.55%	4.17%
Oregon	Nonseparation	21,736	114	10.87%	5.47%	9.63%	5.28%
Pennsylvania	Monetary	43,542	111	6.01%	4.08%	3.40%	2.80%
Pennsylvania	Separation	44,229	113	16.44%	6.31%	10.44%	5.67%
Pennsylvania	Nonseparation	97,037	116	14.47%	6.29%	11.00%	6.02%
Puerto Rico	Monetary	1,998	78	28.51%	9.86%	17.80%	7.33%
Puerto Rico	Separation	4,541	98	0.00%	0.00%	0.00%	0.00%
Puerto Rico	Nonseparation	8,712	88	2.80%	4.01%	2.80%	4.01%
Rhode Island	Monetary	1,640	106	7.05%	4.95%	1.67%	2.07%
Rhode Island	Separation	3,320	106	11.80%	7.13%	8.69%	6.21%
Rhode Island	Nonseparation	4,749	106	4.71%	4.34%	1.50%	2.91%
South Carolina	Monetary	13,697	113	4.51%	3.95%	3.66%	3.58%
South Carolina	Separation	24,163	117	0.87%	1.70%	0.87%	1.70%
South Carolina	Nonseparation	51,502	117	10.37%	4.04%	1.37%	1.95%
South Dakota	Monetary	466	102	4.36%	3.52%	2.22%	2.23%
South Dakota	Separation	1,477	104	5.57%	4.38%	2.93%	2.52%
South Dakota	Nonseparation	2,112	105	2.92%	3.25%	2.11%	2.87%
Tennessee	Monetary	6,756	101	9.75%	6.19%	5.77%	5.16%
Tennessee	Separation	11,189	108	12.70%	6.34%	5.05%	4.11%
Tennessee	Nonseparation	21,762	106	16.37%	7.87%	10.08%	6.06%
Texas	Monetary	47,346	112	5.48%	4.37%	5.48%	4.37%
Texas	Separation	111,939	115	7.64%	4.39%	4.40%	3.55%
Texas	Nonseparation	160,979	115	2.52%	2.88%	1.69%	2.38%
Utah	Monetary	1,925	106	6.14%	4.46%	6.14%	4.46%
Utah	Separation	6,201	109	3.91%	3.52%	3.31%	3.31%
Utah	Nonseparation	27,614	111	3.06%	3.01%	3.06%	3.01%
Virginia	Monetary	5,036	92	17.28%	8.21%	4.06%	3.96%
Virginia	Separation	18,715	111	10.26%	5.83%	10.26%	5.83%
Virginia	Nonseparation	11,316	111	8.94%	5.57%	8.94%	5.57%

BENEFIT ACCURACY MEASUREMENT
DENIED CLAIMS ACCURACY
IMPROPER DENIAL RATES REPORT
Batch Range: 201927~202026
		Type		Adjusted
		Population	Cases	Improper	95% C.I	Improper	95% C.I
State:	Denial Type	of Denials	Completed*	Denial	(+/-)	Denial**	(+/-)
Vermont	Monetary	237	25	0.00%	0.00%	0.00%	0.00%
Vermont	Separation	914	42	9.08%	10.94%	4.32%	5.90%
Vermont	Nonseparation	951	37	2.15%	4.13%	2.15%	4.13%
Washington	Monetary	13,287	95	9.75%	6.11%	8.51%	5.61%
Washington	Separation	20,616	104	13.46%	7.36%	12.12%	6.88%
Washington	Nonseparation	72,502	104	21.90%	7.69%	16.65%	7.81%
Wisconsin	Monetary	5,244	100	9.59%	6.42%	9.59%	6.42%
Wisconsin	Separation	15,889	103	9.69%	5.20%	8.24%	5.92%
Wisconsin	Nonseparation	61,402	107	20.83%	8.98%	20.83%	8.98%
West Virginia	Monetary	939	98	18.33%	7.77%	15.31%	7.35%
West Virginia	Separation	7,158	112	1.93%	2.68%	1.93%	2.68%
West Virginia	Nonseparation	5,163	112	5.10%	4.57%	3.22%	3.66%
Wyoming	Monetary	1,329	93	16.28%	8.54%	9.46%	6.72%
Wyoming	Separation	1,289	103	9.04%	5.99%	6.92%	5.52%
Wyoming	Nonseparation	8,163	103	7.04%	4.46%	7.04%	4.46%
Note: 95% C.I. is the 95 percent confidence interval for the estimated rate. The interval is the range between the rate minus the value in the 95% C.I. column and the rate plus the value in the 95% C.I. column. For example, the interval for 10.0% +/- 2.5 is 7.5% to 12.5%. The true rate is expected to lie within 95 percent of the intervals constructed from repeated samples of the same size and selected in the same manner as the BAM DCA sample.
*Excludes cases not meeting DCA definition for inclusion in population, withdrawn claims, and claims for which monetary eligibility was established upon receipt of CWC, UCFE, and/or UCX wage credits.
**Adjusted rate excludes erroneous denials that were corrected by agency or reversed on appeal prior to DCA case completion.

PAID CLAIMS ACCURACY
CASE COMPLETION AND TIME LAPSE REPORT
Batch Range: 201927 ~ 202026
	Cases	Cases	Percent	60 Day	90 Day
State	Sampled	Completed	Completed	Time Lapse	Time Lapse
AK	361	361	100.00	77.01	82.83  +
AL	362	362	100.00	92.27	99.72
AR	358	358	100.00	70.67	80.17  +
AZ	357	357	100.00	63.31  *	91.32  +
CA	569	569	100.00	98.95	100.00
CO	336	316	94.05	69.35  *	88.99  +
CT	362	362	100.00	78.73	85.64  +
DC	359	359	100.00	86.91	97.21
DE	270	270	100.00	100.00	100.00
FL	357	356	99.72	99.44	99.72
GA	360	360	100.00	55.56  *	83.06  +
HI	350	350	100.00	62.86  *	65.71  +
IA	360	360	100.00	59.72  *	76.11  +
ID	273	273	100.00	94.51	98.53
IL	360	360	100.00	72.22	81.11  +
IN	333	333	100.00	71.17	73.57  +
KS	353	353	100.00	71.39	84.14  +
KY	374	373	99.73	16.04  *	24.06  +
LA	349	349	100.00	95.13	97.71
MA	348	348	100.00	60.92  *	72.99  +
MD	360	360	100.00	86.94	96.94
ME	276	276	100.00	82.61	94.57  +
MI	348	348	100.00	74.43	79.02  +
MN	313	313	100.00	97.76	99.36
MO	360	360	100.00	78.06	80.56  +
MS	369	369	100.00	69.38  *	76.96  +
MT	270	270	100.00	84.07	90.37  +
NC	390	390	100.00	63.85  *	92.56  +
ND	271	271	100.00	82.29	82.66  +
NE	269	269	100.00	93.31	99.63
NH	280	280	100.00	91.43	96.07
NJ	343	340	99.13	76.38	83.09  +
NM	336	336	100.00	81.85	85.71  +
NV	350	276	78.86	60.86  *	66.00  +

PAID CLAIMS ACCURACY
CASE COMPLETION AND TIME LAPSE REPORT
Batch Range: 201927 ~ 202026
	Cases	Cases	Percent	60 Day	90 Day
State	Sampled	Completed	Completed	Time Lapse	Time Lapse
NY	440	367	83.41	74.32	83.41  +
OH	340	340	100.00	58.24  *	67.65  +
OK	368	368	100.00	98.91	100.00
OR	350	350	100.00	89.43	96.86
PA	362	362	100.00	96.41	100.00
PR	366	338	92.35	15.57  *	31.97  +
RI	348	348	100.00	77.87	88.79  +
SC	352	352	100.00	75.28	78.98  +
SD	267	267	100.00	52.81  *	76.78  +
TN	361	361	100.00	88.09	100.00
TX	407	407	100.00	90.91	100.00
UT	362	362	100.00	92.54	96.69
VA	350	350	100.00	99.14	100.00
VT	146	91	62.33	21.23  *	28.08  +
WA	343	343	100.00	44.90  *	62.39  +
WI	334	334	100.00	77.54	93.11  +
WV	356	356	100.00	97.19	99.44
WY	270	270	100.00	64.07  *	72.96  +
Note:	Time lapse has been adjusted for cases reopened with code '3'.
*	Failed to meet 60 day time lapse standard of 70% complete.
+	Failed to meet 90 day time lapse standard of 95% complete.
	The Benefit Accuracy Measurement program was suspended due to Covid-19 epidemic excluding data for batch range 202014 through 202026. As a result case completion timeliness was negatively impacted

DENIED CLAIMS ACCURACY
CASE COMPLETION AND TIME LAPSE REPORT - DCA
Batch Range: 201927 ~ 202026
	Denial	Cases	Cases	Percent	60 Day	90 Day
State	Type	Sampled	Completed	Completed	Time Lapse	Time Lapse
AK	Monetary	99	99	100.00%	86.87%	89.90%
AK	Separation	107	107	100.00%	87.85%	88.79%
AK	Nonseparation	105	105	100.00%	87.62%	87.62%
AL	Monetary	110	110	100.00%	93.64%	100.00%
AL	Separation	110	110	100.00%	97.27%	100.00%
AL	Nonseparation	112	112	100.00%	95.54%	100.00%
AR	Monetary	112	112	100.00%	73.21%	77.68% +
AR	Separation	112	112	100.00%	79.46%	89.29%
AR	Nonseparation	112	112	100.00%	81.25%	87.50%
AZ	Monetary	105	105	100.00%	89.52%	100.00%
AZ	Separation	105	105	100.00%	81.90%	92.38%
AZ	Nonseparation	105	105	100.00%	83.81%	96.19%
CA	Monetary	158	158	100.00%	100.00%	100.00%
CA	Separation	159	159	100.00%	100.00%	100.00%
CA	Nonseparation	151	151	100.00%	100.00%	100.00%
CO	Monetary	108	98	90.74%	75.93%	88.89%
CO	Separation	110	102	92.73%	80.91%	90.00%
CO	Nonseparation	107	102	95.33%	84.11%	94.39%
CT	Monetary	116	116	100.00%	86.21%	88.79%
CT	Separation	116	116	100.00%	87.07%	89.66%
CT	Nonseparation	116	116	100.00%	84.48%	87.93%
DC	Monetary	108	108	100.00%	92.59%	98.15%
DC	Separation	111	111	100.00%	91.89%	97.30%
DC	Nonseparation	110	110	100.00%	95.45%	99.09%
DE	Monetary	127	127	100.00%	99.21%	100.00%
DE	Separation	115	115	100.00%	100.00%	100.00%
DE	Nonseparation	116	116	100.00%	100.00%	100.00%
FL	Monetary	114	114	100.00%	100.00%	100.00%
FL	Separation	116	116	100.00%	100.00%	100.00%
FL	Nonseparation	116	115	99.14%	99.14%	99.14%
GA	Monetary	117	117	100.00%	59.83% *	82.05% +
GA	Separation	117	117	100.00%	65.81%	87.18%
GA	Nonseparation	117	117	100.00%	65.81%	87.18%
HI	Monetary	107	107	100.00%	73.83%	76.64% +
HI	Separation	108	108	100.00%	68.52%	75.00% +
HI	Nonseparation	107	107	100.00%	73.83%	76.64% +
IA	Monetary	105	102	97.14%	73.33%	80.95% +
IA	Separation	105	105	100.00%	79.05%	85.71%
IA	Nonseparation	105	101	96.19%	77.14%	83.81% +

DENIED CLAIMS ACCURACY
CASE COMPLETION AND TIME LAPSE REPORT - DCA
Batch Range: 201927 ~ 202026
	Denial	Cases	Cases	Percent	60 Day	90 Day
State	Type	Sampled	Completed	Completed	Time Lapse	Time Lapse
ID	Monetary	104	104	100.00%	93.27%	97.12%
ID	Separation	105	105	100.00%	98.10%	100.00%
ID	Nonseparation	105	105	100.00%	97.14%	99.05%
IL	Monetary	106	106	100.00%	83.02%	87.74%
IL	Separation	106	106	100.00%	81.13%	91.51%
IL	Nonseparation	106	106	100.00%	79.25%	88.68%
IN	Monetary	104	104	100.00%	81.73%	83.65% +
IN	Separation	104	104	100.00%	79.81%	82.69% +
IN	Nonseparation	104	104	100.00%	81.73%	83.65% +
KS	Monetary	113	113	100.00%	85.84%	89.38%
KS	Separation	115	115	100.00%	84.35%	92.17%
KS	Nonseparation	112	112	100.00%	83.04%	91.96%
KY	Monetary	115	114	99.13%	26.09% *	35.65% +
KY	Separation	115	115	100.00%	33.04% *	60.87% +
KY	Nonseparation	149	149	100.00%	48.99% *	83.22% +
LA	Monetary	111	111	100.00%	93.69%	98.20%
LA	Separation	111	111	100.00%	98.20%	99.10%
LA	Nonseparation	111	111	100.00%	89.19%	93.69%
MA	Monetary	105	105	100.00%	60.00%	71.43% +
MA	Separation	105	105	100.00%	58.10% *	74.29% +
MA	Nonseparation	105	105	100.00%	69.52%	76.19% +
MD	Monetary	119	119	100.00%	89.08%	95.80%
MD	Separation	116	116	100.00%	89.66%	98.28%
MD	Nonseparation	115	115	100.00%	84.35%	93.91%
ME	Monetary	114	114	100.00%	86.84%	98.25%
ME	Separation	111	111	100.00%	86.49%	96.40%
ME	Nonseparation	115	115	100.00%	91.30%	98.26%
MI	Monetary	102	102	100.00%	79.41%	84.31% +
MI	Separation	101	101	100.00%	81.19%	82.18% +
MI	Nonseparation	101	101	100.00%	86.14%	87.13%
MN	Monetary	94	94	100.00%	100.00%	100.00%
MN	Separation	95	95	100.00%	100.00%	100.00%
MN	Nonseparation	93	93	100.00%	100.00%	100.00%
MO	Monetary	113	113	100.00%	80.53%	80.53 % +
MO	Separation	113	113	100.00%	77.88%	81.42 % +
MO	Nonseparation	112	112	100.00%	83.04%	83.93 % +
MS	Monetary	107	107	100.00%	78.50%	83.18 % +
MS	Separation	107	107	100.00%	77.57%	83.18 % +
MS	Nonseparation	107	107	100.00%	80.37%	84.11 % +
DENIED CLAIMS ACCURACY
CASE COMPLETION AND TIME LAPSE REPORT - DCA
Batch Range: 201927 ~ 202026
	Denial	Cases	Cases	Percent	60 Day	90 Day
State	Type	Sampled	Completed	Completed	Time Lapse	Time Lapse
MT	Monetary	106	106	100.00%	92.45%	95.28%
MT	Separation	109	109	100.00%	89.91%	93.58%
MT	Nonseparation	110	110	100.00%	85.45%	91.82%
NC	Monetary	115	115	100.00%	73.91%	95.65%
NC	Separation	117	117	100.00%	88.03%	97.44%
NC	Nonseparation	117	117	100.00%	82.05%	96.58%
ND	Monetary	104	104	100.00%	88.46%	92.31%
ND	Separation	104	104	100.00%	92.31%	92.31%
ND	Nonseparation	104	104	100.00%	91.35%	91.35%
NE	Monetary	104	104	100.00%	96.15%	100.00%
NE	Separation	107	107	100.00%	96.26%	100.00%
NE	Nonseparation	106	106	100.00%	94.34%	100.00%
NH	Monetary	114	114	100.00%	92.98%	97.37%
NH	Separation	116	116	100.00%	99.14%	100.00%
NH	Nonseparation	116	116	100.00%	99.14%	100.00%
NJ	Monetary	109	109	100.00%	77.98%	83.49 % +
NJ	Separation	111	111	100.00%	80.18%	84.68 % +
NJ	Nonseparation	111	110	99.10%	88.29%	90.09%
NM	Monetary	107	107	100.00%	82.24%	87.85%
NM	Separation	108	108	100.00%	86.11%	89.81%
NM	Nonseparation	108	108	100.00%	87.96%	90.74%
NV	Monetary	111	90	81.08%	63.96%	66.67% +
NV	Separation	110	98	89.09%	73.64%	74.55% +
NV	Nonseparation	110	96	87.27%	70.91%	71.82% +
NY	Monetary	113	113	100.00%	92.04%	100.00%
NY	Separation	114	113	99.12%	92.11%	99.12%
NY	Nonseparation	114	114	100.00%	91.23%	100.00%
OH	Monetary	105	105	100.00%	63.81%	71.43 % +
OH	Separation	105	105	100.00%	66.67%	73.33 % +
OH	Nonseparation	105	105	100.00%	65.71%	73.33 % +
OK	Monetary	120	120	100.00%	98.33%	100.00%
OK	Separation	121	121	100.00%	100.00%	100.00%
OK	Nonseparation	119	119	100.00%	98.32%	100.00%
OR	Monetary	115	115	100.00%	99.13%	100.00%
OR	Separation	115	115	100.00%	98.26%	100.00%
OR	Nonseparation	114	114	100.00%	100.00%	100.00%
PA	Monetary	111	111	100.00%	100.00%	100.00%
PA	Separation	113	113	100.00%	98.23%	99.12%
PA	Nonseparation	116	116	100.00%	99.14%	100.00%

DENIED CLAIMS ACCURACY
CASE COMPLETION AND TIME LAPSE REPORT - DCA
Batch Range: 201927 ~ 202026
	Denial	Cases	Cases	Percent	60 Day	90 Day
State	Type	Sampled	Completed	Completed	Time Lapse	Time Lapse
PR	Monetary	123	100	81.30%	26.02% *	47.15 % +
PR	Separation	106	106	100.00%	37.74% *	56.60 % +
PR	Nonseparation	103	93	90.29%	28.16% *	46.60 % +
RI	Monetary	106	106	100.00%	62.26%	76.42 % +
RI	Separation	106	106	100.00%	70.75%	79.25 % +
RI	Nonseparation	106	106	100.00%	66.04%	77.36 % +
SC	Monetary	115	115	100.00%	76.52%	80.87 % +
SC	Separation	117	117	100.00%	77.78%	79.49 % +
SC	Nonseparation	117	117	100.00%	78.63%	79.49 % +
SD	Monetary	105	105	100.00%	52.38% *	73.33 % +
SD	Separation	104	104	100.00%	54.81% *	72.12 % +
SD	Nonseparation	105	105	100.00%	59.05% *	73.33 % +
TN	Monetary	107	107	100.00%	89.72%	100.00%
TN	Separation	108	108	100.00%	87.04%	100.00%
TN	Nonseparation	106	106	100.00%	94.34%	100.00%
TX	Monetary	115	115	100.00%	96.52%	100.00%
TX	Separation	115	115	100.00%	93.04%	100.00%
TX	Nonseparation	115	115	100.00%	94.78%	100.00%
UT	Monetary	111	111	100.00%	93.69%	98.20%
UT	Separation	109	109	100.00%	93.58%	96.33%
UT	Nonseparation	111	111	100.00%	91.89%	98.20%
VA	Monetary	111	111	100.00%	100.00%	100.00%
VA	Separation	111	111	100.00%	100.00%	100.00%
VA	Nonseparation	111	111	100.00%	100.00%	100.00%
VT	Monetary	53	28	52.83%	43.40% *	52.83 % +
VT	Separation	58	42	72.41%	44.83% *	70.69 % +
VT	Nonseparation	62	37	59.68%	37.10% *	56.45 % +
WA	Monetary	107	107	100.00%	65.42%	81.31 % +
WA	Separation	104	104	100.00%	56.73% *	73.08 % +
WA	Nonseparation	104	104	100.00%	63.46%	78.85 % +
WI	Monetary	102	102	100.00%	77.45%	95.10%
WI	Separation	103	103	100.00%	83.50%	93.20%
WI	Nonseparation	107	107	100.00%	85.05%	94.39%
WV	Monetary	111	111	100.00%	98.20%	100.00%
WV	Separation	112	112	100.00%	100.00%	100.00%
WV	Nonseparation	112	112	100.00%	100.00%	100.00%
WY	Monetary	102	102	100.00%	71.57%	82.35 % +
WY	Separation	103	103	100.00%	75.73%	84.47 % +
WY	Nonseparation	104	103	99.04%	75.00%	82.69 % +

	Note:		Time lapse has been adjusted for cases reopened with code '3'.
	*		Failed to meet 60 day time lapse standard of 60% complete.
	+		Failed to meet 90 day time lapse standard of 85% complete.
		The Benefit Accuracy Measurement program was suspended due to Covid-19 epidemic excluding data for batch range 202014 through 202026. As a result case completion timeliness was negatively impacted

1 Some SWAs have updated the COBOL software edit and selection process as software programs and claimstaking processes have changed. ETA has published further guidance for states’ updating these systems and monitored system changes. See UIPL No. 25-20; web published https://wdr.doleta.gov/directives/corr_doc.cfm?DOCN=6819; June 15, 2020.

2 Offset of benefits are considered as being constructively received by the claimant.

B-9