This Program Letter provides guidance on LSC grantees’ use of automated client-intake systems (such as a web-based or telephone system) to collect financial-eligibility information from a prospective client and make a determination of financial eligibility under 45 C.F.R. Part 1611 without review by a person.[fn]This program letter does not address other screening issues, such as conflicts, alienage, or nature of the legal problem.[/fn]

Background

LSC grantees “may provide legal assistance supported with LSC funds only to individuals whom the recipient has determined to be financially eligible for such assistance.” 45 C.F.R. § 1611.4(a).   The financial-eligibility requirements of 45 C.F.R. Part 1611 also apply to uses of non-LSC funds that are allocated to LSC requirements, such as Private Attorney Involvement activities under 45 C.F.R. Part 1614.[fn]For example, non-LSC funds, when used for PAI purposes, are functionally treated as if they were LSC funds with all of the requirements and restrictions applicable to the LSC funds attached, and clients served by PAI attorneys must be LSC-eligible regardless of the source of funds supporting the provision of those services. LSC Advisory Opinions 2009-1004 (June 19, 2009) and 1998-06-21 (June 21, 1988) (both citing to 45 C.F.R. § 1614.1).[/fn]

LSC has been asked whether, consistent with LSC’s regulations, grantees may use automated systems to collect financial-eligibility information and determine financial eligibility without review by a person. 

The answer to the question is yes, subject to the conditions and limitations described in this Program Letter.  With proper design and implementation, automated systems can not only meet LSC’s requirements, they can streamline the information collection process, eliminate unnecessary barriers for applicants, maintain or improve the accuracy of financial screening compared with other methods, and expedite intake decisions.  They can improve the efficiency and the effectiveness of the delivery of LSC-compliant legal services.

Automated systems must meet the same requirements as other forms of intake, including in-person and telephone interviews.  The requirements do not vary with the form of intake.

Application of LSC’s Regulations to Automated Systems for Information Collection and Screening for Financial Eligibility

Part 1611 sets out four requirements for any system through which a grantee gathers financial-eligibility information.  Once those requirements are met, the grantee can then determine if the applicant meets the grantee’s financial-eligibility requirements.

First, the system must “make reasonable inquiry regarding sources of income, income prospects, and assets.”  45 C.F.R. § 1611.7(a).  Second, the system must do so using “simple” inquiries and procedures “in a manner that promotes the development of trust between attorney and client.”  Id. § 1611.7(b).  Third, the system must have the capacity to determine whether there is “substantial reason to doubt the accuracy of the financial-eligibility information provided,” such as internal inconsistencies or missing material information.  Id. § 1611.7(c).  Fourth, “[i]f there is substantial reason to doubt the accuracy of the financial-eligibility information provided by an applicant,” then the system must “make appropriate inquiry to verify the information a manner consistent with the attorney-client relationship.”  Id.  If an automated system meets some but not all of these requirements for a particular application, then the grantee must employ other means to comply with the remaining requirements.

The grantee must keep reasonable records regarding the structure and operation of any automated system, and it must use that information for testing and review of system operation and of each applicant’s financial eligibility (e.g., branching-logic structure, question trees, procedures manuals to guide staff in the use of the system, and completed applications for clients provided services).  See generally 45 C.F.R. § 1611.7(b).

The following guidelines provide illustrative examples of how LSC’s regulations would apply to hypothetical automated intake systems.  The term “applicant” refers to the person providing the information, who may be the potential client, or, where authorized, a family member, a household member, or other person in a position to assist with the application process, depending on the circumstances and the grantee’s policies.

1. Data Collection (Requirements One and Two)

Grantees may collect complete financial-eligibility information through automated systems that provide a “reasonable inquiry regarding sources of the applicant's income, income prospects and assets.” 45 C.F.R. § 1611.7(a)(1).  Those systems must:

1. Provide instructions and explanations that are accurate, clear, and intelligible to applicants;
    
2. Elicit information regarding household (1) size, (2) composition, (3) income, and (4) assets that is sufficient to make a financial-eligibility determination.
   • • Example 1-A: The system asks the name of each person living in the same household (as that term is defined by the grantee) as the applicant.  For each person, the system then asks about relationship, income, assets, and income prospects.  The system asks sufficient questions for an applicant to communicate that she lives alone, her only income comes from part-time employment at a retail store that reports all her wages on her W-2, she has no vehicles or other property (other than personal possessions in her rental apartment without any significant resale value), and her cash and other assets are worth less than $800.
3. Provide an opportunity for the applicant to indicate if she does not have some of the requested information or if she cannot fully provide the information through the automated system. In those situations, the system must process the application as in need of further review for financial eligibility by a person.
   • • Example 1-B: The system provides a summary of the financial and household information collected.  It asks the following questions.  If the applicant answers yes, it either goes to a new set of automated questions or it routes the application for review by a person and possible follow-up.
       • • Do you think anyone in your home receives money that is not on this list (even if you don’t know how much)?
         • Do you think anyone in your home has something of value (over $200) that is not on this list (even if you don’t know how much its worth)?
     • Example 1-C: If the household’s total income and assets exceed the eligibility guidelines, then the system asks more detailed questions relevant for a determination of whether, for eligibility purposes, the grantee would exclude some of the occupants in the applicant’s home (e.g., rent-sharing by two independent families) and/or whether the household has expenses that could qualify for authorized exceptions to the income ceiling under 45 C.F.R. § 1611.5.  The system includes opportunities for the applicant to ask for help understanding the questions.  The system also includes opportunities for the applicant to say that he has additional information or questions and to either provide that information or schedule direct contact with the grantee (by phone, e-mail, text, or other method). 
        
     • Example 1-D: A self-employed applicant’s income is at 130% of the federal poverty guidelines.  The system flags her application based on the self-employment and asks if she has both reported and unreported income and business expenses. She says yes to all three.  Because the system is not programmed to handle those areas of inquiry, it flags the application for direct contact (by phone, email, text, or other method).

2. Further Inquiries (Requirements Three and Four)

1. Automated systems may make further inquiries when there is “substantial reason to doubt the accuracy of the financial information provided.” Error-checking is a common feature of automated data-entry systems.  Often, online systems for commercial purchases will not proceed if the information submitted does not meet preset parameters.  Grantees may program systems to ask follow-up questions or indicate which information entered is insufficient.  Grantees may set limits on the number of changes that may be made to a particular entry so as to require human review.
    
2. The grantee does not need to review the household and financial information or the financial-eligibility determination unless the grantee has substantial reason to doubt its accuracy. The fact that the automated system collected the information and made the determination is not, by itself, a substantial reason for such a doubt.
   • • Example 2-A: The system first asks for the names of every person living in the same home, including minors and infants.  It then asks about income for each person.  The applicant cannot proceed without entering an income for each person or selecting $0 or [Unknown].
        
     • Example 2-B: The grantee programs the system to compare total household income with the federal poverty guidelines and also has chosen to automatically cross-reference that information with general information about any benefits programs that the household participates in.  For example, the total household income is 150% of poverty and the applicant indicated that she receives TANF.  The system notes that TANF is normally not available for households at 150% of poverty in that county.  The information could be incorrect, or the people in the home could constitute more than one household for TANF purposes.  The system may ask further questions or route the application for review by a person.

3. Determinations

Once the automated system has elicited financial-eligibility information in a manner that meets all four requirements above, the automated system could then make the financial-eligibility determination under the following circumstances (without review by a person unless an applicant’s response triggers additional review).  These determinations must follow the grantee’s financial eligibility policies as adopted pursuant to Part 1611.

1. The information provided shows that the grantee can determine the applicant’s financial eligibility based entirely on the information provided without any steps that require the exercise of judgment.
   • • Example 3-A: An adult woman applies and states that she lives alone with a salary-based income at 400% of the federal poverty guideline.  The grantee programmed the system to determine that applicants with incomes above 250% of poverty for the household size are not financially eligible.  The system provides her with contact information if she has questions about that determination.  For incomes between 200% and 250% of poverty, the system sends the application to a person to review in case there is an obvious error that might put the application in the below-200% category.
   • The following examples (3-B through 3-D) involve a single woman who lives in the grantee’s service area with only her minor child.  Their only income is TANF (the grantee does not count SNAP for financial eligibility).  They received a letter that their TANF will end in thirty days.  They have applied for assistance to maintain their TANF.
     • Example 3-B: Under the grantee’s § 1611.4(c) policy, receipt of TANF is sufficient to demonstrate financial eligibility for residents in the grantee’s service area. The grantee programmed the system to make that determination when an applicant lives in the grantee’s service area and states that her only household income is TANF.
        
     • Example 3-C: Instead of a § 1611.4(c) policy, the grantee has a § 1611.5(a)(1) policy that the income ceiling does not apply to applicants who (1) seek to “maintain benefits provided by a governmental program for low income individuals or families” and (2) whose “assets do not exceed the [grantee’s] applicable asset ceiling.”   The applicant states that she owns no vehicles or real property, has one bank account with $800, owns nothing outside of her household, and that the total value of her household possessions is less than $1000.  The information is sufficient for the system to add up $800 and $1000 to conclusively determine that her maximum household assets are $1800, which is below the grantee’s asset ceiling.  The grantee programmed the system to make the financial-eligibility determination when:
       • • the only legal problems checked off (from a list that includes many problems and “other”) are “I received a letter that I won’t get TANF anymore” and
         • the applicant entered actual or maximum dollar values for all listed assets; and
         • the applicant checked “no other assets.”
     • Example 3-D: The grantee has the same § 1611.5(a)(1) policy as in Example 3-C and no § 1611.4(c) policy. The applicant checks “I don’t know” for the value of her household assets.  The grantee programmed the system to send to a human reviewer all applications for which the applicant checked “I don’t know” for the value of any household assets.
2. The system checks that the information provided is internally, mathematically, and logically consistent for purposes of the financial-eligibility determination.
   • • Example 3-E: The same facts as Example 3-B above. The total monthly TANF benefit for a household of two in the grantee’s service area never exceeds 125% of poverty.  The applicant enters in the monthly value for her TANF.  For applications in which TANF is the only income on the application, the grantee programmed the system to check that the total monthly benefit is below 125% of poverty for the household size.  If so, then it automatically approves the applicant as financially eligible.  If it exceeds 125% of poverty, then an inconsistency exists and the system notes the application for review by a person.
3. The system checks for any missing material information (Requirement Three).
   • • Example 3-F: The grantee programs the system to send the application for review by a person if the applicant checks “I don’t know” for the income of any household member.  The system requires that the applicant state an income for each household member or check “I don’t know.”
4. The grantee must have a process for determining when and how to improve the automated system based on any detected problems.

Development and Review of Automated Systems

LSC encourages grantees to work with LSC staff and with other legal services providers in developing, testing, implementing, evaluating, and refining automated intake systems. 

Grantees should thoroughly test automated intake systems. Grantees should design their systems to enable applicants to identify when the questions are unclear to them or have not fully captured the applicant’s financial situation.

Grantees must ensure that their intake systems provide access for people facing barriers because of disabilities. 45 C.F.R. § 1624. Technology can provide ways to overcome such barriers (such as text-based communications like typing in a chat dialog that people with hearing impediments could use).  Technology can also pose new barriers (such as graphical interfaces requiring pointing to images on a screen that people with limited or no vision cannot easily use).

Contacting LSC

Please direct questions or requests for further information to Mark Freedman, Senior Associate General Counsel, at 202-295-1623 or mfreedman@lsc.gov.