Why Multilingual Research Matters

There are so many barriers that exist when people try to access the social safety net. The process of applying might involve in-person interviews that are difficult to make. The language of an application or renewal might be confusing, or could require documents that are hard to find. Many people experience difficulties and compounding hardships that make it harder to access the help they need in times of crisis.

All of these things are made more challenging by language access barriers. Studies show that 42% of Spanish speakers, 74% of Cantonese speakers, and 84% of Mandarin speakers who are eligible for CalFresh, California’s food assistance program, are currently not enrolled. If we take a cumulative look at language communities who have a primary language other than English (or PLOE populations), this translates into over one million households that aren’t receiving the benefits they deserve. We can and must do better to serve these communities. 

Why addressing language barriers is a critical equity issue

One of Code for America’s main missions is to make the safety net more inclusive and to reduce the participation gap in public benefit programs, particularly among populations who need it the most. That is why our GetCalFresh service is available in English, Spanish, and Traditional Chinese, and we offer texting services and customer support in English, Spanish, and both Simplified and Traditional Chinese. However, taking a look at our GetCalFresh data from the fall of 2021, we found that despite the availability of  Spanish and Chinese application options, approximately 40% of Spanish-speaking clients, 56% of Mandarin-speaking clients, and 68% of Cantonese-speaking clients were applying in English—not their preferred language. It made us wonder: Why are so many people applying in English even though they prefer to speak another language? What does this tell us about language accessibility in our products? And more fundamentally, what does language access and language equity mean? And why does it matter for our work?

Put bluntly, access is about equal treatment in government services. Equity is about providing extra support for populations who may be underrepresented in or excluded from public programs. While we care very much about equality and equity, the ultimate goal of our language work is inclusion; we want to know how our work can remove the systemic barriers that make accessing food assistance so difficult in the first place and have those directly impacted by safety net systems lead the way in systems change.

 

When focusing on language equity, we rely on inclusive design principles, which operate from the reversal of the 80/20 rule.  Instead of focusing most of our efforts on reaching 80% of the population, we spend most of our efforts on the 20% of the population that is “harder to reach.”  We know this will give 100% of our clients a better experience, because improving our products for those who are most marginalized will by default make the product easier to use for everyone—and at the same time, centering our research on PLOE populations will help us more effectively narrow the participation gap.

Building research practices with PLOE communities

This past year, the GetCalFresh team spoke to multilingual staff at community-based organizations (CBOs) and two qualitative research contractors who are native Mandarin and Cantonese speakers to help us better understand the barriers faced by these language populations as they applied to GetCalFresh and moved through the service. This was done concurrently with a longitudinal research study with Spanish-speaking populations where we spoke with clients over the course of six months from the start of their application for the CalFresh program through renewal. 

Having native speakers conduct research with their own language communities both opens trust in the research process and means that researchers carry cultural and linguistic sensitivities that can make the research a better overall experience for clients. These cultural sensitivities show up everywhere in our language use. In our practices with Spanish-speaking clients, for example, we make sure to address clients with the respectful second-person pronoun usted before switching to the more informal tu if that’s what clients prefer; for older clients, we make sure to use the terms señor, señora, or the gender-neutral señore. For Mandarin or Cantonese-speaking clients, we reword formal written questions to sound more conversational, prioritizing the way the languages are spoken rather than written. 

When research is led by native speakers who are embedded in their cultural communities, “harder to reach populations” aren’t actually that hard to reach. We just need the right people and resources available to do this work. 

What our research has found so far

Misinformation and confusing rules can prevent PLOE populations from applying to CalFresh. Widespread misinformation, confusion around eligibility rules, and stigma around public benefits programs discouraged eligible PLOE populations from reaching out to get needed food assistance. In particular, changes in federal policies around public benefit usage and immigration proceedings lead to fear and confusion about potential consequences of accessing benefits. 

Maybe it's going to affect the individual's personal records, tax records or personal work records or what. If you have ever taken one of these benefits, it's kind of like you are not a good person anymore. 

If one day [my daughter] wants to study or something else, then she shouldn’t have anything [applying for public benefits] affect her [future] plans.

PLOE populations rely heavily on people within their social networks to act as translators and/or help navigate the digital application; however for more isolated individuals or individuals with strained social networks, not having access to this kind of assistance can be a barrier. While CBOs play an important role in supporting PLOE communities, only 3-17% of GetCalFresh online applications from PLOE clients are CBO-assisted. This means that the remaining 83-97% of applicants may rely on family, friends, and community members to assist them with completing the GetCalFresh application—and have to structure their application process around a helper’s availability. This reliance on social networks also disadvantages socially isolated people who want to apply for CalFresh online.

Even if the online app is online, they still can’t put it on a computer. So we [the CBO staff] go in and we help them with that. If not us, seniors who have kids or grandkids can help. But without a family (like in the case of some refugees) then it might be harder. 

When I asked [my friend for help] again later, he seemed to be a little impatient, so I was embarrassed to trouble him again.

If PLOE populations are not aware of an application in their own language or if the application is not available in their preferred language, it can lead to on-the-fly translations, which can slow down the application process. Because general or community translators are not necessarily familiar with CalFresh terminology, on-the-fly translation can also reduce the quality of the information being shared.

Translations can be pretty challenging because a lot of the questions we ask and the verbiage we use is very technical, very CalFresh specific. Just your general interpreters are probably not always going to understand them, what they mean and how to translate them. I don't think interpreters are supposed to have to go back and forth with a client but it does end up happening sometimes. It's like, sometimes our agents get left out of that conversation as they try to figure that out themselves, which isn't ideal but it does happen.

Accounting for the nuances in different written and spoken formats of a language can help improve the client experience. Even if an application is available in different formats, the digital application experience can be a challenge if only one type of written format is offered or if people prefer spoken to written language. For example, among Cantonese and Mandarin-speaking communities, there is no default written format—some feel more comfortable with Simplified Chinese; others feel more comfortable with Traditional Chinese. The best practice is to provide services in both written formats, which is currently done on California’s Employment Development Department website. Even so, for Cantonese-speakers, an application in written form can still be difficult to understand, since the tone and clarity of the written language differs so much from the spoken language.

I am not familiar with traditional Chinese characters, I can probably understand it when put it in a sentence, but when I take it out separately, I'm not so sure.

It requires a lot of effort and time for me to digest and fully understand written Chinese. I need to read each word and think about what it means. The application asks about employer info. I couldn't understand 僱主 (employer) in written Chinese, because we say 老細 (boss [less formal]) in Cantonese.

Where do we go next?

Our research so far has already illuminated so many barriers and areas for growth, and we plan to use these learnings to better serve PLOE populations. Our team is currently working on design interventions to make our services more equitable for different language populations, and we’ll keep documenting and sharing those strategies as we implement and learn from them. One thing that has been shown loud and clear from our work so far: in order to reach communities who speak languages other than English, we must engage in research efforts designed with an understanding of the needs of those who have been underrepresented in and excluded from services. To serve these populations inclusively and effectively, we have to engage them in a culturally responsive way. Hearing directly from our clients—especially those who struggle to access the services they need—makes government services more accessible for everyone.