UNEMPLOYMENT INSURANCE DETAILED INTEGRITY STATE ASSESSMENT REPORT November 2016

Transcription

1

2 ACKNOWLEDGEMENTS This report would not have been possible without the 49 states responding to the in-depth state assessment earlier this year. The Center is tremendously thankful for the states on the State Engagement Subcommittee (AL, DC, CT, FL, IA, MD, MO, NE, NJ, OK, TX, and UT) for helping shape the on-line questionnaire. The Center is extremely grateful for USDOL s input, and commitment to helping states combat UI fraud and improper payments. The Center s Steering Committee (CO, FL, MT, NE, NJ, NY, OH, SC, TX, UT, and WA) also supported the State Engagement effort and played an important role for the issuance of this report. Finally, Members of the State Engagement Team: Brian Langley Team Lead, UI SMEs: Laura Boyett, Maria Nobel, Christine Paquette, and Bill Starks, Center Project Director, Randy Gillespie, and Senior Policy Advisor, Jim Van Erden.

3 Executive Summary UNEMPLOYMENT INSURANCE DETAILED INTEGRITY STATE ASSESSMENT REPORT November 2016 The 2016 Unemployment Insurance (UI) Integrity Detailed Assessment is a snapshot in time of current state efforts to combat UI fraud and improper payments, their integrity tools, and related procedures. The responses to this assessment may not portray all of the integrity tools and practices currently being discussed or employed by the states as the assessment questions targeted specific integrity functions. This report is the second state assessment report developed under the National Association of State Workforce Agencies (NASWA) s UI Integrity Center (Center) State Engagement effort. The State Engagement effort involves states in Center activities with the goal of improving UI integrity. These activities include state assessments and reporting on current integrity efforts, convening states in regularly scheduled information sharing sessions, and creating a model integrity operational blueprint to inform and guide states integrity efforts. The information gathered through the 2016 assessment, and others, will be a catalyst for a series of white papers on integrity topics of most interest to the UI community. All of the assessment information will ultimately be loaded into a fully accessible, continuously maintained, State Portfolio Data Base (SPD). The SPD application will provide an interactive resource of state integrity practices for easier consumption and use by the states. The SPD will be available early next year through the UI Community of Practice Integrity tab on the Center s website. The SPD application in the Center s Digital Library will allow states to exchange state-specific information as it relates to UI fraud and improper payments. The 2016 state assessment is more robust than the 2015 version, with fifty main questions and over one hundred subset questions depending on how the main questions were answered. The assessment was released to states earlier this year with 49 states completing the on-line in-depth state assessment resulting in a ninety-four percent response rate. The responses show that states are making in-roads in implementing new tools and processes to help identify UI fraudulent activities, reduce UI improper payments, and strengthen trust fund solvency. This report also highlights a continuing need for future outreach to states to gather additional information and clarify existing procedures and tools. 2

4 Table of Contents Executive Summary... 2 The 2016 State Integrity Assessment... 5 Operations Section... 7 Q1) Does your state have a shareable copy of your UI agency s organizational chart?... 7 Q2) Does your state verify out-of-state work registration?... 7 Q3) Does your state have an automated case management system for the UI Benefit Payment Control (BPC) unit?... 8 Q4) Adjudication - Does your BPC Unit issue determinations?... 9 Q5) Does your state use the SIDES Earnings Verification system (E-Response)? Q6) Has your state modified its integrity messaging in the last twelve months? Protocol (IP) Address Section Q7) As part of your state s UI process, does your state capture a claimant s or employer s IP address? Q8) Does your state screen IP addresses? Q9) Some states have a repository for a "hot" IP address list for known bad or prior fraudulent IP address. Does your state compile a "hot" IP address list? Q10) Does your state have a process for verifying IP addresses that are "masked" or filed from a proxy server?. 18 Training Section Q11) Does your state UI agency offer training programs specific to addressing BPC or UI integrity issues designed to prevent, detect, and recover improper payments (both fraud and non-fraud)? Recoupment Section Q12) Does your state currently participate in TOP to collect delinquent UI employer tax debts? Q13) Does your state use any outside data sources or tools for locating (skip tracing) debtors' most current address? Q14) Does your state have an automated collections case management system (separate and apart from your automated billing system)? Q15) Does your state have any automated processes for contacting debtors to recover benefit overpayments or tax debts? Q16) Does your state allow electronic payment options on overpayments or tax debts? Q17) Does your state have a State Income Tax offset program? Q18) Does your state actively recover UI benefit overpayments through wage garnishments issued to the claimant's current employer(s)? Q19) Does your state actively process liens or levy bank accounts on overpayments or tax debts? Q20) Does your state leverage the issuance and/or renewal of any business or professional licenses to recover UI overpayment and/or delinquent employer tax debts? Q21) What other sources are actively used to recover UI overpayments and/or delinquent employer tax debts?.. 28 Q22) Does your state actively file adversary proceedings objecting to the discharge of fraudulent UI overpayments in bankruptcy? Q23) Does your state require certain employers to post a cash bond amount to ensure future tax debts are paid?. 30 Q24) Does your state refer or contract out UI debt recovery? Q25) What other collection actions does your state find motivating for employers or claimants to settle their UI debts?

5 Prosecution Section Q26) Does your state prosecute for criminal charges on UI fraud overpayment cases? Q27) Does your state prosecute for criminal charges on employer tax cases? Q29) Does your state use law enforcement officers to support the criminal prosecution of UI fraud cases? Data Analytics and Predictive Modeling Section Q30) Does your state use predictive modeling for fraud prevention and detection? Q31) Does your state use data analytics for fraud prevention and detection? Q32) Does your state have any Standard Operating Procedures or other documentation of work processes used in conjunction with your data analytics and predictive modeling applications that could be shared with the Center? 42 Cross Match Section Q33) Does your UI agency cross match with your state s State Directory of New Hires (SDNH)? Q34) Which UI claims does your state run against the National Directory of New Hires (NDNH)? Q35) What is the frequency of your cross match with the NDNH? Q36) Does your state UI agency cross match against incarceration data? Q37) Does your state have an agreement in place with the U.S. Social Security Administration to access their Prisoner Update Processing System (PUPS) information? Q38) Does your state cross match Continued Claims against death records or the U.S. Social Security Administration's Death Master File (DMF)? Q39) Does your state run a cross match on death records or the Death Master File against claimants with existing overpayments? Q40) Does your state conduct fictitious employer detection cross matches? Q41) Does your state's quarterly wage cross match run more than once a quarter? Q42) As it relates to cross matches being run, does your state have a process (automated or not) for prioritizing hits to investigators as workload assignments? Q43) Has your state implemented a new cross match or revised an existing cross match in the last 12 months? ID Verification Section Q44) Does your state have a verification process for Identity Theft? Q45) Does your state cross match against driver license or state identification card information? Integrity Data Hub Section Q46) What specific national database cross matches, if made available by the IDH, would provide the most value to your state? Q47) Does your state currently collect bad (suspicious) actor information? Q48) Would your state be willing to submit claimant and employer UI data to the IDH? Q49) Do states prefer that the Data Hub pushes information to states or that states pull information from the Data Hub Q50) Has your state changed, or initiated, a new or innovative process in the last twelve months that relates to UI Integrity in any of the following areas that was not answered in any of the prior questions? Conclusion Future Efforts ATTACHMENT A The UI Integrity In-depth State Assessment questionnaire sections

6 The 2016 State Integrity Assessment The Center is a joint Federal and State initiative through NASWA, funded by the U.S. Department of Labor, to assist states in their efforts to improve integrity in their unemployment insurance programs. The Center s mission is to be a source for integrity strategies, information, and resources, focusing particularly on the prevention, detection, and recovery of improper payments, fraud and delinquent employer contributions. The completion of this assessment and accompanying report falls under the Center s State Engagement effort, a cornerstone of the Center s mission. This report facilitates that effort by gathering and sharing integrity-related information among states. The 2016 Unemployment Insurance Integrity Detailed Assessment Report (IDAR) describes the inventory of specific UI integrity practices, tools, and strategies currently used in state operations. This report is the second in a continuing series of state assessment reports with the first completed and published in October This IDAR is divided into nine main sections including: Operations; Internet Protocol (IP) Usage; Training; Benefit Recoupment; Fraud Prosecutions; Predictive Modeling and Data Analytics; Data Cross matching; Identity Verification; and Integrity Data Hub. The information gathered through this assessment, and other state polling, will serve as a catalyst for upcoming white papers on integrity topics of most interest to the UI community. All of the assessment information will ultimately be loaded into a fully accessible, continuously maintained SPD application of state integrity practices for easier consumption and use by the states. The information will be available early next year through the UI Community of Practice Integrity tab on the Center s website. The 2016 IDAR contains fifty primary questions and more than one hundred subset questions that were addressed depending on how the primary questions were answered. While the assessment was comprehensive in nature, the 5

7 responses from states may not portray all of the integrity tools and practices currently being discussed or employed. The assessment was released to states at the end of February 2016 and responses were complete in April Forty-nine states completed the on-line assessment resulting in a ninety-four percent response rate. For an overview of the assessment questions, please see Appendix A. For more information on the Center and all of our initiatives, please visit: and click the Integrity tab or 6

8 Operations Section Q1) Does your state have a shareable copy of your UI agency s organizational chart? The Center gathered individual state organizational information to better inform its work as we continue to map out our engagement strategy. Developing a better understanding of various state structures throughout the country, and the roles and responsibilities of each, will be helpful when states assess their own restructuring needs and/or look for examples of how effective operations are structured. Thirty-six states shared a copy of their organizational chart. The information gathered is not listed in this report but will be made available online through the SPD application. Insights gained from a review of the organizational charts include the following: Twenty-five states have their Benefit Payment Control (BPC), or UI Integrity unit, report to the UI Director; Seven states have their BPC, or UI Integrity unit, report to the Agency s Commissioner/Director/Secretary; and Four states have their BPC, or UI Integrity unit, report to a different office such as the General Council, separate agency division, or split reporting duties across divisions. Q2) Does your state verify out-of-state work registration? One of the leading causes for improper payments occurs when claimants do not adhere to work registration requirements as prescribed in state laws and policies. The Center inquired as to what states were doing to verify if interstate claimants had registered for work in the agent state s public workforce system. In this assessment, twenty-three states said they verify out-of-state work registrations as a matter of protocol. Five states identify all out-of-state claimants and send a notification of the requirement to register for work. In those states, claimants are given a date to provide proof of their registration. If no proof is provided, a hold, denial, or adjudication appointment is put on the claim. Some states have agreements with neighboring states for an automated interface, whereas other states perform this via a manual process. 7

9 Q3) Does your state have an automated case management system for the UI Benefit Payment Control (BPC) unit? Automated Case Management System (#states) NO 19 YES 30 States that said yes to this question were also asked the following: a) Was the automated system developed in-house or was an outside vendor used? b) What year was the system implemented or last updated? c) How are the automated cases assigned? and d) Does your state tie individual performance standards to the BPC cases assigned to staff? Of the thirty states that have an automated BPC case management system, eleven systems were developed in-house while eighteen states had a vendor create their system. Five companies were named as vendors used: On Point Technology; Deloitte; Fast Enterprises; Pondera Solutions; and RedMane Technology. The average age of state BPC case management systems is 8.3 years with the oldest system developed in 1989, and the newest implemented in Updates to the systems averaged 2.8 years. The responses related to automated case assignment methodology is depicted in the table below. 8

10 How Cases are Assigned Cross match or source dependent Potential dollar amount threshold Other Based on location Previous overpayment Difficulty of the case The twelve states that provided additional comments in the other field, listed assignment factors such as: The last four digits of the claimant s SSN; The perceived validity of the case; The claim type (fraud or non-fraud); Investigator workload; and Wage audit notices. When asked if states tie individual performance standards to the BPC cases assigned to staff, the majority of the thirty states that responded indicated they do tie in quality, timeliness, and quantity to individual performance standards. This tracks closely with Benefits Timeliness and Quality (BTQ) performance evaluation variables for UI adjudication. Some of the respondents stated they also look at monthly expectations, the number of cases opened during a specific time period, and investigation results when measuring staff performance. Q4) Adjudication - Does your BPC Unit issue determinations? Overpayment Determinations issued by BPC Unit Yes No Non-fraud issues 44 5 Fraud issues

11 The majority of states BPC units adjudicate both fraud and non-fraud issues. In addition, one state indicated their agency does not structurally have a BPC unit. In the 2015 assessment, three common procedures for adjudication were mentioned: UI fraud investigative staff adjudicate all issues and refer other nonmonetary issues to another unit; UI fraud investigative staff adjudicate all non-monetary and fraud/nonfraud issues for their cases; and All non-monetary and fraud/non-fraud determinations are made by an adjudications unit outside the UI fraud or BPC unit. The state responses this year reinforced the answers to last year s assessment, that states fraud operations have their investigators adjudicate both fraud and non-fraud issues once the investigation is complete. A few states forward nonfraud issues to their regular UI adjudication process. Q5) Does your state use the SIDES Earnings Verification system (E-Response)? The SIDES Earnings Verification application automatically sends wage verification notices to employers through a state specific employer portal. Using an automated process to communicate and receive information from employers, provides faster information for weekly wage data allowing states to more quickly identify if a claimant is working and subsequently failing to report their weekly earnings. Manual paper responses from employers are less timely and are a leading cause of overpayments. The results of state responses demonstrated confusion in answering this question. It appears respondents didn t distinguish between the SIDES Separation Information exchange and the SIDES Earnings Verification exchange. For more information about SIDES and E-Response, please visit Q6) Has your state modified its integrity messaging in the last twelve months? Forty-four states answered this question. The majority of current upgrades to UI integrity messaging involve the state UI agency s website as the medium for their messaging, with specific letters to claimants and employers as the secondary medium. Eighteen states indicated they had modified their UI integrity messaging on social media accounts. Fourteen states also provided written comments such as, a pop-up messaging ( nudging ) in their claimant 10

12 system, while another state cited their factsheets to employers and claimants on UI Fraud and Integrity. One state said they were starting a radio campaign in the summer of 2016 on misclassified workers. 32 Areas States have Changed their Integrity Messaging in the Last Year State UI Agency's website Letters to claimants Letters to employers Interactive Voice Response (IVR) System Social media accounts Other - Write In 11

13 Protocol (IP) Address Section Anytime a device connects to the internet, the device is assigned a number called an IP address. This IP address is transmitted to any servers visited and is logged on those servers, leaving a trail of Internet activity. As the use of data analytics and data cross matching grows, capturing a claimants IP address is proving to be helpful in combatting improper payments. Tracking IP addresses helps states identify if an individual filing for UI benefits is outside the country or using a suspicious address that is linked to prior fraudulent activities. Many state UI laws prohibit individuals from filing a claim from outside the United States, except in limited circumstances. Q7) As part of your state s UI process, does your state capture a claimant s or employer s IP address? Thirty-eight states answered that the state capture an IP address, while nine indicated they did not, as shown in the chart below. IP Address Captured No 9 Yes 38 12

14 If a state said yes to this question, they were also asked what their UI process(es) were to capture an IP address. 33 Processes that Capture IP Addresses Those states providing comments mentioned additional processes or applications for capturing IP addresses, such as: Each time a claimant uses their UI system; With their on-line claimant self-service address change or when direct deposit enrollment is accessed; Their employer business registration application; Using Splunk software to capture IP address and the state is able to pull the IP addresses as needed for investigations; During log-ins, monetary calculator, address updates, and method of payment updates; When the status of a UI claim is viewed; and For out-of-state initial claims. Q8) Does your state screen IP addresses? States that Screen IP Addresses NO 19 YES 28 13

15 The states that answered yes to this question were also asked: a) Types of IP addresses screened? b) Types of claims checked when screening for IP addresses? c) What actions does your state take if a foreign IP address is detected during claims processing? d) Which additional action(s) are taken when the state prevents a claim from being completed? e) Does your state allow UI eligibility for claimants filing for outside the United States or Canada? f) What UI eligibility must be met if a claim is filed outside the U.S. or Canada? and g) What technology does your state use to identify IP addresses? When asked about actions taken once a suspicious IP address is identified, fifteen states block the claim if the on-line claim is filed from a foreign IP address. Twelve states allow the claim to be filed but either place an issue on it for adjudication, or an alert is sent to staff. Types of IP Addreses Screened Originating from a foreign country Both Originating within the US When asked what types of claims states check when screening IP addresses, close to seventy percent of states (28) checking yes for this question said they screened for all claim types initial, additional, reopened, and weekly continued claims (see chart below). 14

16 Types of UI Claims Checked when Screening for IP Addresses Initial claims Weekly continued claims Additional claims Reopened claims All of the above When foreign IP addresses are encountered, states have several different processes, including: Blocking the claim from being filed; Allow the claim to be filed, but creating an issue for adjudication; Allowing the claim to be filed and sending an alert to staff for review; and Allowing the claim to be filed, but running it through a scoring mechanism to determine whether or not a review should occur. Action Taken on Foreign IP Addresses Allow but Flagged Block Claim 15

17 When asked, does your state allow UI eligibility for claimants filing from outside the United States or Canada, fourteen states allow this, ten of those states allow claimants to file from outside the U.S. or Canada if they meet certain eligibility criteria, while seven of the fourteen states allow claimants to file from outside the U.S. or Canada if they are a trailing military spouse. One state indicated their UI law allows for filing in Federal territories, such as Guam, or other areas where military personnel might be stationed. Another state said their UI law allows if work search requirements are met since some jobs are global. 10 Allow UI Eligibility Outside the U.S. or Canada 7 1 Meets certain A&A requirements Trailing military Spouse Other For technology, fourteen states use commercial IP detection software from the following vendors: GeoIPA databases; MaxMind; F5 ASM Firewall; Oracle Identity Manager; and Cisco SourceFire IPS. IP Address Detection Process - In-house vs Vendor Vendor Software Developed in-house Other 16

18 Additional information provided includes eleven states that block both foreign and suspicious American IP addresses as well as having the IP address check as part of all of their claim functions. Q9) Some states have a repository for a "hot" IP address list for known bad or prior fraudulent IP address. Does your state compile a "hot" IP address list? Compile a "Hot" IP Address List NO 27 YES 20 State s answering yes to this question were also asked: a) Is this list incorporated in your state s data analytics method or process? and b) Does your state use the hot IP address list as a cross match source? Of the twenty states that answered yes to this question, eleven states incorporate the list in their state's data analytics process. When asked to briefly describe the method for incorporating the "hot" IP address list into their state's data analytics process, nineteen states provided comments. The comments varied depending on how the respondent interpreted this question. Some states provided technical detail about their systems integrating the list whereas other states approached their response from a process perspective describing how the data is used. A few states indicated they used a manual process to determine if there is a legitimate reason for the cross match hit. When asked if their state uses the "hot" IP address list as a cross match source, ten states said their system generates a report of flagged claims for further investigation. 17

19 Q10) Does your state have a process for verifying IP addresses that are "masked" or filed from a proxy server? Masking the IP address refers to hiding the actual location of where the computer is connecting to the Internet. The top four reasons people intentionally hide their IP address are: To hide their geographical location; To prevent website tracking; To avoid leaving a digital footprint; and To bypass any bans or blacklisting of their IP address. Verify Masked IP Addresses NO 33 YES 14 Fourteen states said they have a process for verifying IP addresses that are masked or filed from a proxy server. 18

20 Training Section Many states offer training including: adjudication methods, computer system updates, and new processes or procedures. The Center wanted to know if states provide specific training addressing BPC or UI integrity issues. Q11) Does your state UI agency offer training programs specific to addressing BPC or UI integrity issues designed to prevent, detect, and recover improper payments (both fraud and non-fraud)? Thirty-five states said they offer UI integrity and fraud training to their investigators. As a follow up to this question, we asked who provides the training? 31 Who trains your BPC Staff? BPC/Integrity staff Agency, or UI Division, training unit Other Outside training program Other training programs mentioned were: Mathematica s UI integrity training modules, Guest speakers from law enforcement partners, such as State Police, Federal Department of Homeland Security, County Prosecutor's Office, and Attorney General's Office, USDOL s BAM auditor and supervisor training, and One state has developed computer-based training on UI integrity. When asked about what topics they offer for integrity training, the bulk of the responses centered on prevention, detection, and collections. In addition to states saying they train staff on case management and collections, eleven states provided other comments for topics they offer for integrity training. Some of the comments included: Improper payments and criminal prosecution; 19

21 Case sharing between adjudication and fraud unit; UI adjudication; Presenting cases in administrative hearings and testifying in court; Prosecution training, e.g., report writing, interviewing techniques, mock interviews, and mock trial; Fraud detection criteria and new employee investigative training; Cross match types; and Changes in agency policy and procedures. Topics for Integrity Training Detection Prevention Collections or recovery Case management Other 20

22 Recoupment Section UI Improper payment recoupment methods vary widely by state. Gaining an understanding of processes, procedures, vendors, and software used for recoupment is the intent of these questions. Areas of interested included: the Treasury Offset Program (TOP), skip tracing, case management systems, liens, levying bank accounts, any automated processes, electronic payment options, leveraging business licenses, bankruptcy filing, bonds, and contracting out for UI debt recovery. State responses to the following fourteen questions demonstrated how the various state UI laws, regulations, and policies impact what collection tools states have in recovering UI debts. Q12) Does your state currently participate in TOP to collect delinquent UI employer tax debts? Federal law requires state UI agencies to participate in TOP to recover certain UI benefit overpayments and tax debt. If a state said yes to this question, they were also asked: a) What were your recoveries for 2015? b) Was your state using an automated or manual process to record the UI employer tax debts in your state UI system? and c) Was your state planning on including UI employer tax debt to TOP requests? If so, when? Currently, there are ten states participating in TOP to collect unpaid UI employer tax debts. Asked if states were anticipating implementing the UI tax debt component to TOP, ten said they were planning on implementing it sometime this year for the 2016 tax season, and fifteen states noted after States Particiapting in TOP UI Tax After

23 For more information on the collection numbers from TOP, please visit s Treasury s most recent newsletter at: ts_matter.htm. Q13) Does your state use any outside data sources or tools for locating (skip tracing) debtors' most current address? Skip tracing is a colloquial term used to describe the process of locating a person's whereabouts for any number of purposes. Thirty-six states indicated they use external sources and tools for updating bad addresses. For these states, they were also asked what data sources are used for locating the most current address. 26 Skip Tracing Sources States Use (# states) Other State Directory of New Hires National Directory of New Hires Driver License Agency State Revenue Department Social Media Credit Bureau Company Other responses included the following: County property records; Current employer information; Social services agency data; Postal service; Secretary of State s office; PACER bankruptcy system; Internet searches; LexisNexis/Accurint; and Clear Software. Comment: see Question 22 on page 30 describing PACER. 22

24 Q14) Does your state have an automated collections case management system (separate and apart from your automated billing system)? Collections Case Management System NO 30 YES 19 Nineteen states said they have a collections case management system that helps prioritize assignments and other collection tools. States indicate their systems are used for the following: Billing; Quarterly balance statements; Repayment agreements; Civil processing; Bankruptcy status; Warrants; Some UI Tax work; and One state said their collections case management system was built as part of their UI Benefits system. On-Point Technology s RECOVER software solution was named for its automated billing notices and payment agreements. Other states comments about case management systems included: Creates a task when a claimant s repayment agreement is missed and becomes delinquent; Issues quarterly balance statements; Presents cases when new information is received. It determines the status of cases based on payments; Assigns cases for delinquent employers which are flagged to show priority; Automatically sends correspondence to claimants including: demands, missed payment letters, payment plan agreements, and notice of intent to garnish wages. It also creates alerts to indicate further review is needed prior to sending notice of garnishment letters, second demands that qualify for liens, and missed payments that require review to 23

25 determine next steps. Bankruptcy cases are tracked in the system but handled manually by staff; Advises staff of failed payment agreements or not meeting monthly minimum payments, and generates work items for examiners. Also, notifies of judgments and garnishment opportunities and expired payment agreements where balances exist; Automates the collection cycle and assignments are automatically assigned to Tax Enforcement Officers at the time a Tax Warrant is issued; Allows workload to be managed through the use of follow-up codes by activity type, such as, debt has a bad address, sent to private collection agency, or have bank proceedings underway; Provides timely review and handling of civil actions and auto population of applicable documents; and Tracks a case throughout the collection process - it notifies staff if a payment hasn't been received or if paperwork has not been received. Q15) Does your state have any automated processes for contacting debtors to recover benefit overpayments or tax debts? Nineteen states answered yes to this question. Those states were asked what automated methods are used. 10 Automated Process for Contacting Debtors Targeted mail campaign in addition to your monthly billing statement Other Telephonic dialing (robo calls) - Agentless notifications Telephonic dialing (robo calls) - Agent available Eight states also provided write-in comments that included: Demand or payment plan letters; Billing statements; Automating the notifications to employers for outstanding wage audits and new hire employer compliance; and Initiate garnishments and issue tax liens. 24

26 Q16) Does your state allow electronic payment options on overpayments or tax debts? Thirty-eight states said they offered electronic payment. The assessment then asked if states allowed either credit card and/or electronic check (ACH) payments. Only four states said they offer both electronic check and credit card payment options for claimants and employers. The remaining responses were mixed as technology resources and banking options shape what states are able to do. Some states allow credit card transactions for claimants only, others allow ACH payments from employers but not for claimants. Four states allow ACH payment for both parties, but not credit card options. One state mentioned they had outsourced their payment website to ValuePay, which allows both ACH and credit card payments. Electronic Payment Options on UI Benefit Debt Automated Self-serve Manual Staff-assisted 6 12 Not Applicable 5 8 Electronic Check (ACH) Credit Cards As the chart above demonstrates, twenty-nine states allow for an automated self-service process to electronically pay on UI debt. Eighteen additional states said they assist with electronic payments. A breakdown of state responses for electronic payment options are: Three states offer only credit cards to claimants and only electronic checks to employers; Three states offer only electronic checks to employers and no electronic payments options to claimants; Five states offer only credit cards to claimants and both options to employers; 25

27 One state offers both options to claimants and no electronic payment options for employers; Two states offer only credit cards to both claimants and employers; Two states offer only electronic checks to claimants and both options to employers; Four states offer only the electronic check payment option to both claimants and employers; Two states offer both options to claimants and only electronic checks to employers; and Four states offer both options to claimants and employers. When asked how credit card transaction fees are covered, the majority of the twenty-six states responding said the transaction fees were passed along to the employer and claimant. Six states said they covered the fees for claimants and employers through their penalty and interest fund, while two states said they covered the fees through the state s general fund. Q17) Does your state have a State Income Tax offset program? Thirty-six states have a state income tax offset program, three do not, and nine states do not have a state income tax. Five states indicated they are active in offsetting employer tax debts in addition to claimant overpayments. Q18) Does your state actively recover UI benefit overpayments through wage garnishments issued to the claimant's current employer(s)? Wage Garnishments NO 15 YES 34 Thirty-four states answered yes to this question. When asked about the percentage of total annual overpayment collections attributed to wage garnishments, seven states said their wage garnishment process accounted for 11 to 42 percent of total improper payment collections. 26

28 The Center will use these responses to gather more information from states with high collection percentages so best practices can be promoted to the UI system. In the 2015 Assessment, states reported: Cost to file the paperwork and time required to send staff to court to obtain judgments were some reasons why several states don t conduct garnishments or do very few; States most successful in collecting outstanding UI debt through garnishments generally have an automated process with the courts or state law doesn t require a court order; Several states use a lien process or an administrative order, rather than a court order, to conduct wage garnishments; and Most of the states not currently administering wage garnishments would require state law changes in order to begin this practice. Q19) Does your state actively process liens or levy bank accounts on overpayments or tax debts? Liens or Levy Bank Accounts NO 6 YES 43 Overall, forty-three states said they actively process liens and/or levy bank accounts on UI benefit overpayments and/or tax debts. Only six states file liens on personal and real property and also levy bank accounts for claimants and employers. The remaining responses included a varying combination of liens and levies on claimants and employers. States were asked how they applied liens and levies. The processes used are shown in the tables below. 27

29 Liens Claimants Employers Automated 7 12 Manual Levies Claimants Employers Automated 2 5 Manual 8 24 Q20) Does your state leverage the issuance and/or renewal of any business or professional licenses to recover UI overpayment and/or delinquent employer tax debts? Levying Licenses for Delinquent UI Debts NO 28 YES 21 Twenty-one states answered yes to this question. When asked what type of licenses states leverage, responses included: Liquor Licenses; Contractors Licenses; Commercial Fishing Licenses; and Lottery Sales permits. None of the states comments specifically identified freezing or taking away licenses for benefit overpayment debts. Q21) What other sources are actively used to recover UI overpayments and/or delinquent employer tax debts? Thirty-six states indicated they pursue other revenue sources for outstanding UI debts as follows: 28

30 23 Other Resources to Recover UI Debt (# states) Lottery winnings Other State Vendor receivables Not applicable Casino winnings Other responses include: Mineral royalties; Collection agencies; Small claims court; Bankruptcy monitoring; Personal liability of Corporate Officers; Interstate benefit offsets; Recoupment liens filed on claimant s social security number; Recreational license suspended; Halt state contracts; Federal vendor offset; Interstate Reciprocal Overpayment Recovery Arrangement (IRORA); Controller warrant hold; and Unclaimed property. 29

31 Q22) Does your state actively file adversary proceedings objecting to the discharge of fraudulent UI overpayments in bankruptcy? Actively Pursue Bankruptcy Filings (# states) NO 11 Yes 37 Generally speaking, fraud overpayments are not dischargeable under federal bankruptcy law. Thirty-seven states confirmed they actively file adversary proceedings objecting to the discharge of fraudulent UI overpayments. Recent discussions with states on this topic have identified a spike in bankruptcy filings, presumably due to TOP and its sixty-day notice. In addition, states have said that barriers to pursuing bankruptcy petitions are caused by a general lack of resources, a strict felony test, new staff, and new procedures. Many states have multiple bankruptcy courts, further compounding the difficulty in monitoring bankruptcy filings. Several states have noted the use of Public Access to Court Electronic Records (PACER) for notification and verification of individuals filing for bankruptcy. Some states pursue criminal charges if the fraud overpayment debt is tied up in bankruptcy court. Q23) Does your state require certain employers to post a cash bond amount to ensure future tax debts are paid? Cash Bonds Required NO 25 YES 22 30

32 Twenty-two states require certain employer groups to post a surety bond for UI covered employment. Seventeen states require it for reimbursable employers, four states utilize this for industry specific employers, and four states make this a requirement for chronically delinquent employers. Four states in this group also commented they require a cash bond on chronically delinquent employer accounts for both reimbursable and contributory employers. Q24) Does your state refer or contract out UI debt recovery? Refer or Contract Out Collections (# states) YES 10 NO 39 The majority of states responding to this question do not contract or refer UI debts for recovery actions. Of the ten states that do, four states indicated they have a threshold on the age of the debt prior to referral, ranging from 45 days to 150 days. Of the same ten states, five states use other state agencies to collect referred UI overpayments, while five states contract out to a third party collector. Five states refer for both UI claimant overpayments and UI tax debt. Of the remaining states, two states refer only the UI tax debt and three states refer only benefit overpayments. Refer or Contract out UI Debt Recovery (# of States) Other State Agencies Contractor Claimants Only 1 2 Employers Only 0 2 Both Claimants and Employers

33 Q25) What other collection actions does your state find motivating for employers or claimants to settle their UI debts? Thirty-two states responded to this question. TOP notices, prosecutions, liens, and wage garnishments were the primary categories mentioned as motivators for individuals to settle their UI debt. Collection Actions that Motivate Individauls to Settle their UI Debt Tax offset Prosecutions Liens/Levies Wage garnishments Specific state responses include: Automated billing letters and prejudgment letters; Developed a unit that is designated to make collection calls to employers with delinquent accounts and claimants with outstanding overpayment balances; Court proceedings; Prosecutions Small claims court (civil actions) and filing with the state attorney's office Offering claimants the option to repay their overpayment to prevent prosecution Federal Income Tax Refund; Sends notices prior to intercepting tax refunds TOP notices and the likelihood of a tax refund intercept Garnishments; Wage garnishments and also wage garnishments for state government employees Offer earnings withholding orders and monthly installment agreements for claimants Repayment requirements; 32

34 State law requires claimants to repay fraudulent debts including penalties and interest in cash before being eligible to receive future UI benefits Claimants who have committed UI fraud are barred from future benefits until all fraud debt is settled Financial record matching Flexible payment plans Installment-payment plans Tax; and End of September deadline to have all employer debt satisfied to maintain eligibility for a reduced UI tax rate Civil penalties for employer s failure to file required reports Negotiation of settlements (employer debt) State law that allows for the responsible party debt transfer making the person liable for the entire UI tax debt in addition to the entity being liable Warrants and Levies. A levy is filed against a person s annual permanent fund dividend Liens are the most motivating action for employer groups Warrants and levies are very effective. 33

35 Prosecution Section The following four questions provide information on hoe states prosecute claimants and employers that are suspected of committing benefit and tax fraud. Factors such as resources, judicial structure, and case priorities influence the number of cases states pursue for criminal prosecution and convention. Many claimants and employers agree to restitution or a payment plan before the court system gets involved. Q26) Does your state prosecute for criminal charges on UI fraud overpayment cases? Forty-four states said they pursue criminal charges on fraud cases and the majority have prosecuted less than 25 cases in the past 12 months. Prosecute UI Benefit Fraud NO 5 YES 44 Average Number of UI Overpayment Cases Prosecuted Criminally per Calendar Year cases or less over 100 cases 26 to 50 cases 51 to 75 cases 76 to 100 cases 34

36 Q27) Does your state prosecute for criminal charges on employer tax cases? Prosecute UI Tax Cases NO 32 YES 13 Thirteen states said they pursue criminal charges on delinquent employer tax cases. Based on state responses, states prosecute fewer criminal charges for employer tax cases than for UI Benefit fraud. The complexity of UI tax fraud cases, staff knowledge, staff resources, and current state laws are stated barriers for the small number of states pursuing criminal charges. Average number of UI Tax debt cases prosecuted criminally per calendar year Five cases or less Over 40 cases Six to 39 cases 35

37 Q28) What purposes does your state use, or message, fraud criminal prosecution cases or statistics? Although the question itself asked about the purposes states use or advertise fraud criminal prosecution or statistics, the response options also incorporated responses on where states display this information. Ten states selected five or more of the categories listed indicating a broad array of means to communicate to the claimant population the consequences for committing UI fraud. As the chart shows, media coverage or newsprint, agency s website, posters, and written communication to claimants were the top four messaging groups. 30 Messaging UI Fraud Criminal Prosecutions Other comments included: Agency s Facebook page; Successful UI fraud prosecution cases are sent to prosecutors who have never handled UI fraud cases to show that criminal prosecution of UI fraud is routinely accomplished; The agency periodically issues or contributes information for news releases (usually in conjunction with law enforcement agencies or the AG's Office). Educational materials sent to employers; and 36

38 Fraud consequences and warnings are posted to online claims and IVR systems. Q29) Does your state use law enforcement officers to support the criminal prosecution of UI fraud cases? Use of Law Enforcement NO 17 YES 32 Thirty-two states responded that they utilize law enforcement in helping support the criminal prosecution of fraud cases. This support includes serving search warrants, arrest warrants, or performing or assisting in investigations. In the 2015 Assessment Report, four states indicated they also have sworn law enforcement officers as part of their UI integrity structure and embedded in their state s UI investigative-type unit. 37

39 Data Analytics and Predictive Modeling Section The use of Data Analytics and Predictive Modeling is increasing in the UI arena as states continue to experiment and find success with this relatively new approach to payment integrity. States are also using data analytics and predictive modeling to prioritize workload from cross matches. With this question, the Center is seeking a broader knowledge of the state and vendor developed applications currently being used and the processes and procedures employed in utilizing the results of these tools. Data Analytics examines raw data for the purpose of drawing conclusions about that information while Predictive Modeling looks at past data to determine the likelihood of certain future outcomes. For this assessment, processing a cross match, or a structured query language (SQL) routine, is not considered operating a data analytics or predictive modeling program. Q30) Does your state use predictive modeling for fraud prevention and detection? Using Predictive Modeling YES 12 NO 37 States that answered yes to this question were also asked: a) Who developed the model? b) At what frequency does it run? c) How does your process operate? d) Is it proprietary or open source? e) Do you investigate a UI claim for possible fraud without a corresponding cross match hit? f) What are the top three useful data elements in your model? 38

40 When asked who developed their application, two states developed an inhouse solution and ten states contracted with a vendor. Vendors currently being used by these states include: Catch Intelligence; Deloitte; On-Point Technology; Pondera; and SAS. When asked if their predicative modeling process runs in real-time or routine batch processes, three states noted their process runs in real-time. When asked how their process operates, one state specifically uses their model on the front end, such as the initial claims process, while five states solely use their model on the back end, or after the claim has been adjudicated and/or benefits have been issued. Six states execute their models on both the front end and back end. 11 Predicative Modeling Process 7 6 At the back end, such as continued claims On the front end, such as an initial claim Both When asked if their application was proprietary or open source, ten states said their predictive modeling software was proprietary and one state said their software was open source. When asked if states investigate a UI claim because of a high probability recommendation from their predictive modeling process, nine states said they are flagging those claims for further investigation. 39

41 States were asked to select the data elements their state uses in their predictive modeling process. 7 Data Elements used in Predictive Modeling Program Other Address Employer Phone Separation reason Taxes Payment type Banking IP addresses, wages reported by employers, incarceration data, death records, employment separation dates, and security questions/answers were listed under other. Q31) Does your state use data analytics for fraud prevention and detection? Eighteen states said they use a data analytics tool for UI fraud prevention and detection. If a state said yes to this question, they were also asked: a) Who developed the model? b) At what frequency does it run? c) How does your process operate? d) Is it proprietary or open source? e) Do you use a risk or ranking score? f) Do you investigate a UI claim for possible fraud without a corresponding cross match hit? g) What are the top three useful data elements in your model? When asked who developed their UI Agency s data analytics model, six states said they built it in-house while twelve contracted with a vendor. 40

42 Vendors mentioned included: Catch Intelligence; Deloitte; On-Point Technology; Pondera; RSI; and SAS. When asked if their data analytics process runs in real-time or batch (frequency), four states noted their process runs in real-time and fourteen states run their data analytics process in a batch method. 15 Data Analytics Process At the back end such as continued claims On the front end such as an initial claim Both When asked how their process operates, and as the chart demonstrates above, fifteen states use their models on the back end, such as the continued claims process, while thirteen states use their models on the front end, such as the initial claims process. Ten states said they use their models on both initial and continued claims. When asked if their data analytics software was proprietary or open source, fifteen states use a proprietary tool one state uses an open source solution. When asked if states use a risk or ranking score for case prioritization from their data analytics, eleven states responded yes and six states do not. When asked if states investigate a UI claim because of a high probability recommendation from their data analytics process, nine states said they are flagging those claims for further investigation. 41

43 Variables Used in Data Analytics Process Address Employer Other Phone Payment type Separation reason Taxes Banking The twelve states responding yes to this question were asked to select the top risk variables (data elements) used in their data analytics process to identify possible UI fraud. Address on record and employer information were the top two variables selected. Other write-in comments included Wages; IP address; IP provider; Separation dates; Username configuration; Same security questions/answers; Incarceration data; State employees; Death lists; Banking Information; and address. Q32) Does your state have any Standard Operating Procedures or other documentation of work processes used in conjunction with your data analytics and predictive modeling applications that could be shared with the Center? Three states indicated they could share their standard operating procedures or other process documentation for their data analytics and/or predictive modeling applications. The Center will follow up with these three states for this information. 42

44 Cross Match Section With the increasing availability of data that can be used for cross matching and analysis in the UI Integrity process, the Center wanted to get a better understanding of specific state activities in this area. The eleven questions covered the following areas: State and National Directory of New Hires; Prison information; Death records; Fictitious employers; Frequency of the matches; How cross match hits are prioritized for investigation; Changes to existing cross matches; and Implementing new cross matches. Responses revealed that states are expanding their use of cross matches to prevent and detect possible fraudulent or suspicious actions. Various states are contracting with vendors to access national databases and/or additional public information data to strengthen their existing cross match activities. States are exploring and experimenting with processes to prevent improper payments before they are issued, and are implementing cross matches during the initial claim process to help stop UI fraud. Q33) Does your UI agency cross match with your state s State Directory of New Hires (SDNH)? Forty-four states said they cross match with their state s SDNH database. If a state said yes to this question, they were also asked: a) The frequency, b) Statutory authority to assess penalties, c) Actively enforcing the penalty assessment, and d) Identifying employers who are non-compliant. 43

45 Frequency of Cross Matches (# states) Real-time 2 Other 4 Weekly batch 17 Daily batch 22 When asked about the frequency of the cross match run, and as the above chart shows, two states connect in real-time, twenty-two states perform a daily batch, and seventeen states run a weekly batch. Four other states mentioned they do a batch run at different times, such as twice a week or every two weeks. When asked if their state has statutory authority to assess SDNH penalties on employers that fail to report in a timely manner, twenty states said yes. When asked about the penalty assessment they assess on employers per violation, sixteen states said they do asses penalties. Ten states assess the $25 maximum allowed under federal law and four states assess less than the maximum amount in a range of $15 to $24 per new hire violation. Several states have larger penalties for incidents of conspiracy as allowed under federal law. Note: Federal law indicates states have the option of imposing civil monetary penalties for noncompliance. Federal law mandates that if a state chooses to impose a penalty on employers for failure to report, the fine may not exceed $25 per newly hired employee. If there is a conspiracy between the employer and employee not to report, that penalty may not exceed $500 per newly hired employee. States may also impose non-monetary civil penalties under state law for noncompliance. When asked if states actively enforce the statutory SDNH penalty assessment, four states actually enforce the requirement. Twelve states have a process for identifying employers who are non-compliant with the SDNH reporting but do not actively assess penalties. 44

46 States were asked what process they use to identify employers who are noncompliant with the SDNH reporting requirements. Three states have automated their cross match using quarterly wage data and the NDNH database. Five states have automated their cross match using quarterly wage data, the SDNH database, and "multi-state filer" database provided by Office of Child Support Enforcement. One state finds non-compliant employers as a result of regular scheduled employer UI Tax audits. Four states use their process for identifying non-compliant employers solely for marketing and outreach efforts to those employers to make them aware of the reporting requirements. Q34) Which UI claims does your state run against the National Directory of New Hires (NDNH)? State and federal law requires employers to report newly hired and re-hired employees to the NDNH. The U.S. Department of Health & Human Services Office of Child Support Enforcement (OCSE) operates the NDNH, a database established pursuant to the Personal Responsibility and Work Opportunity Reconciliation Act of The primary purpose of the SDNH and NDNH is to assist state child support agencies in locating parents and enforcing child support orders. Congress also has authorized specific state and federal agencies, including UI agencies, to receive information from the NDNH for authorized purposes. The chart below shows the responses for this question. Claim Types Cross Matched Against NDNH (# states) Continued claims Initial claims Additional and Reopened claims Other Other claim types written in included: transitional claims, episodic claims, and outstanding UI overpayments. 45

47 Q35) What is the frequency of your cross match with the NDNH? Eight states run the NDNH cross match daily, thirty-nine states perform this weekly, and two states run the NDNH cross match quarterly. Q36) Does your state UI agency cross match against incarceration data? Cross Match Prison Data (# states) NO 11 YES 38 Thirty-eight states said they run cross matches against incarceration data. States that answered yes to the question were asked for the following information: a) What type of institution - local, county, state, or federal? b) The frequency and method of the cross match. i.e. batch, real time, or manual? c) The source of the incarceration data - State/local government, Appriss, SSA/Prisoner Update Processing System (PUPS), or other? d) The number of average quarterly hits? and e) Type of UI claims cross matched - Initial and/or continued? The following table shows the type of institution, the frequency of those queries, data sources states are using to verify if individuals are incarcerated while claiming UI benefits, and the claim type used for cross matching. Unfortunately, the average quarterly hits category did not receive sufficient response to include in the table. The number in the claims column indicate those states that cross match both initial and continued claims. 46

48 Incarceration Cross Matching (# of States) Data Sources Frequency Claim Type State/Local Gov SSA + PUPS Vendor Other Real Time Batch Manual Initial Continued Both Local Jail County Jail State Prison Federal Prison The Center plans on following up with states claiming to cross match against federal prison information. Our understanding is vendors that promote access to prison information are not able to share federal prison information with states. Q37) Does your state have an agreement in place with the U.S. Social Security Administration to access their Prisoner Update Processing System (PUPS) information? Six states have an agreement in place with the Social Security Administration to access the PUPS information. Comment: The New York State Department of Labor has developed a helpful overview of the steps required to secure an agreement with SSA to access PUPS information. The Center will provide this information to states as requested. You also can find it on the UI Community of Practice. Q38) Does your state cross match Continued Claims against death records or the U.S. Social Security Administration's Death Master File (DMF)? Less than half of the states responding to the assessment use state, federal, or other sources to determine if weekly claims may have been filed in the name of a deceased individual. Twenty-three states cross match their continued claims against SSA s death master file or state death records. Fifteen states access their state department for vital records for this information, while six states receive this data from a vendor. 47

49 Death Records Cross Match Source Other State Agency Vendor Other Q39) Does your state run a cross match on death records or the Death Master File against claimants with existing overpayments? The majority of states responding to this question do not use death records to help inform overpayment debt recovery actions. Existing UI Debt against Death Records NO 31 YES 17 For the states that perform this cross match, the assessment asked about the processes used when they receive a positive match. Nine states write-off the debt. Other state responses include the following: Before debt is written off the claims must be investigated to identify who is claiming against the deceased individual s SSN; Collections will try to recover debt from the IRS, state refund, or family trust. After recovery efforts, the debt is written off; Determine if there is an estate and file a proof of claim for overpayments that meet the threshold; If the person dies within the current calendar year, the file is left active for potential tax refund offset. After the end of the calendar year, the file is marked deceased and collections efforts stop; If they have an estate and probate is filed, a claim is filed; 48

50 Notate file as deceased and then the UI claim is stopped for active collection; Submit to estate for collection; and Deceased claimants with overpayments are submitted for write-off consideration through the state write-off approval process. Q40) Does your state conduct fictitious employer detection cross matches? Twenty-six states said they do conduct a fictitious employer cross match. The top four cross matches states said they use to identify potential fictitious employers include the following: UI Benefits filed against a relatively new employer s UI account (18 states), Filing quarterly wages late (17 states), Employer s mailing address (17 states), and Employer s phone number (13 states). Fictitious Employer Cross Match Attributes

51 Six states also provided additional comments on data sources they use for fictitious employer queries or cross matches as follows: State Department of Revenue or Tax Commission; Manual review of employer information on quarterly wages; No SSN for owners or officers, no FEIN for employer, or no predecessor for employer; Part of the analysis compares the employer IP address, mailing address and phone number against the employee's claim information. Also compare the employer's state tax ID to how new it is in the system and if it has a preceding account; Registration in person; and Similar wages reported for all employees of the business. The descending numeric order of the other information states said they use are listed here: Filing prior quarterly wages as a new employer 18 states; First UI benefits being filed against relatively new employer s UI account 18 states; Filing quarterly wages late 17 states; Mailing address 17 states; Multiple wage reports submitted on same date 15 states; High percentage of employer s employees have filed claims against the employer 14 states; Phone number 13 states; Employer has not paid all taxes due 12 states; Same wages being reported per worker multiple quarters 12 states; High percentage of employer s claims are for Maximum Weekly Benefit amount 8 states; IP address 8 states; Federal employer identification number 6 states; and State employer identification number 6 states. Q41) Does your state's quarterly wage cross match run more than once a quarter? Thirteen states run the quarterly wage cross match more than once a quarter. If a state said yes to this question, they were also asked how often they match claims against employer wages. The responses are indicated in the chart below. 50

52 Matching claims with Employer wages Monthly 5 Daily 4 Weekly 4 An additional question asked if any of the variables in the cross match model had been updated recently and what the variables were. Three states provided comments: Depending on the number of hits generated, the number of weeks claimed, and the amount of wages reported, each run criteria is reviewed and adjusted each quarter; Wage threshold was raised from $500 to $1,000 to reduce the number of false positives; and Tweaked the scoring matrix based on industry code, modifications to child support scoring, number of cross matches mailed per quarter, and multiple employers in a quarter. States were asked to briefly describe their process for prioritizing the wage cross match hits. The comments we received are listed below: All the cross match hits are sent to an investigator for review; Automated assignment process. Top priority is given to "in pay status" claims. Once the claim is assigned, investigators also prioritize based on dollar amount and number of weeks claimed; Hits on active claims receive a higher priority; If the maximum payment exceeds the threshold of $75, it is investigated; Manual review the number of weeks claimed with unreported wages and if employer raises an eligibility or separation issue; Most of the work on cross match hits is automated through the creation of a reporting requirement issue. Our system sends requests for information to a claimant and employer about the hit and requests a response, if no 51

53 response is received the issue is adjudicated as a failure to meet reporting requirements; System uses a proprietary algorithm; Vendor tool assigns the priority scores. The highest scores are audited first along with the amount of overpayment; Vendor tool provides probability logic which we use based upon the number of cases we are in a position to pursue; We don t prioritize the cross match hits. All wage audit notices are mailed and attempts are made to obtain information from all parties involved. Once the responses are received, those wage audit notices are investigated; Work items are prioritized using data analytic risk assessment scores; and When the employer returns earnings information, investigation cases are created, assigned, and worked. Q42) As it relates to cross matches being run, does your state have a process (automated or not) for prioritizing hits to investigators as workload assignments? Prioritizing Cross Match Hits NO 22 YES 26 The trend observed in this question is that most states prioritize investigations based on the amount of the potential overpayment, whether that be weeks claimed or wages reported; eight states use this method. Four states prioritize based on the type of cross-match, and three states give the highest priority to new hire cross-matches. One state prioritizes incarceration and death file crossmatches, one state prioritizes first pay claims and one state investigates all cross match hits without priority. Four states rely on technology to prioritize and or make immediate assignments, three states score hits on cross-matches for priority, while one state does a quarterly cross match which then assigns the cases. Four states manually review 52

54 cross-match hits for priority. Two states indicated that cross matched claims in current filing status are seen as the highest priority. Q43) Has your state implemented a new cross match or revised an existing cross match in the last 12 months? Twenty-five states said they had implemented a new cross match or revised an existing cross match over the last year. Fifteen of those states implemented a new cross match and ten states updated an existing cross match. Type of Cross March Implemented or Improved in the Last 12 Months Existing 10 New 15 Cross matches to identify incarcerated claimants and cross matches against death records were the most commonly added or expanded cross matching efforts. Other newly added cross matches reported were as follows: Wage records; NDNH/SDNH; State driver licenses or identification cards; Claimant registration in a border state; State licensing boards; Payroll data from Equifax s The Work Number; Federal wage data contained in the NDNH; Incarceration and deceased individual information; System to check wages on a daily basis; and Real time cross match with state identification card & driver s license records during claim intake. 53

55 When states were asked what cross matches were modified, comments included: Automating the incarcerations cross-match with a vendor application to automatically create the cases and issue a contact notification when the claimant tries to file; Changed from a monthly to a weekly cross-match with incarcerated file from state corrections department; NDNH cross match was enhanced; Owner or Officer filing claim; Social Security Administration - added cross-match against name and Date of Birth; State and county prison cross match was automated to pend claims and automatically generate fact-finding forms to the claimant's home address of record, as well as to the prison where the claimant is reportedly located; and Wage cross match performed on older quarters 54

56 ID Verification Section The following two questions explore what steps states have taken to detect potential identity theft. Responses to our inquiry showed states are making strides in this effort. Q44) Does your state have a verification process for Identity Theft? In the 2015 Integrity Assessment Report, eleven states responded that they had identity verification software as part of the initial claim process and thirteen states were in the process of implementing software or exploring the options available to them. In the 2016 assessment, twenty-one states said they have an identification verification process an increase of ten states since last year s report. When asked how the ID verification process was developed, fifteen states developed their solution in-house and six states contracted with a vendor. States using a vendor solution named LexisNexis most frequently as the vendor. 13 Data Sources to Check an Individula's Identity 8 1 Other LexisNexis Equifax The assessment then asked what data sources were being used to verify an individual's identity. Eight states use data from LexisNexis. Other data sources included the following: Real time verifications with Social Security Administration (SSA); 55

57 State agency for driver's license and state identification for verification of ID number and facial recognition; State s Secretary of State office; Claims with same or similar usernames and passwords; IRS, and employer wage records; Accurint public records; and Appriss data. When asked about a state s process when a claimant responds incorrectly to the identity verification crosscheck question: Five states said the individual has to contact the state UI agency; Two states indicated the individual has to report to a local office or American Job Center; Four states noted the individual has to fax or mail a copy of their driver license and other proof of identity documents to state UI agency; and One state requires the submission of government issued identification for all initial claims and identity verification is renewed at each contact during the course of a claim. Q45) Does your state cross match against driver license or state identification card information? Sixteen states responded that they cross match a claimant s information against their state s driver license or state identification card information database for purposes of ID verification. The data elements that states match against are listed in the chart below. 16 Data Elements Matched Driver license or identification number Physical attributes (such as gender, date of birth, etc.) SSN Address Driving Class 56

58 When asked about the processing frequency, eight states run the cross match in real-time and five states cross match driver licenses or state identification data through a daily batch. One state uses a manual process for this cross match. Six states said they had on-line access to detailed state driver s license information. The assessment also probed states process when the social security number or other personal information doesn t match with the state driver s license data. The responses are below. 12 Process for when there is a Discrepancy 5 4 Require claimant to mail (or fax) verifying documentation Require claimant to appear inperson Other The other responses included: Submission of government issued identification for all initial claims; Identity verification is renewed at each contact during the course of a claim; The claimant will have to go through our identity alert process and is required to provide documents as proof of their identity; and Motor vehicle cross match results are combined with the UIQ cross match results. The assessment also asked about the average volume of quarterly hits through this process. Unfortunately, responses were not provided. 57

59 Integrity Data Hub Section The Center took advantage of this assessment to include four questions for the UI Integrity Data Hub (IDH)task effort. The four question topics were: a) What are the national database cross matches that provide the most value? b) Do states currently collect bad (suspicious) actor information? c) Are states willing to submit claimant and employer data to the IDH? and d) Do states prefer that the IDH pushes information to states or that states pull information from the IDH? The Center appreciates states comments on a national data hub and is taking this input under advisement. The Center is working with nine states to pilot a suspicious actor repository targeting fraudulent common claim information. Q46) What specific national database cross matches, if made available by the IDH, would provide the most value to your state? As the chart shows, suspicious individual data, prisoner updating processing system, and death records files were said to be most valuable. Databases to Cross Match if Made Available by the IDH (# states) SSA's Prisoner Update Processing System Death record files Foreign IP Addresses Treasury's Do Not Pay Portal NDNH quarterly wage data NDNH UI Suspicious claim Individual information data/records Other - please list 58

60 Other comments on additional data bases indicated the following: IRS 1099 records; Bad physical address or bad SSN list; Debit cards mailed to vacant addresses identified by the US Postal Service as potential fraud; DMV along with picture ID; Existing fraud disqualification overpayments; Fraud directory for all states, which include stolen ID information as well as fictitious employer information; Out-of-State DMV; Systematic Alien Verification for Entitlements (SAVE); and SSA and database of federal employees. Q47) Does your state currently collect bad (suspicious) actor information? States that collect Suspicious Actor Informaton NO 17 YES 32 Thirty-two states said they collect suspicious actor data such as SSNs, IP addresses, and employer names associated with fraudulent UI claims, schemes, or overpayments. Q48) Would your state be willing to submit claimant and employer UI data to the IDH? Forty-one states said yes they would be willing to supply claimant and employer UI data for purposes of cross matching and building a suspicious actor repository if all the appropriate data sharing agreements and security protocols are in place. 59