About the Innovation Initiative

The Canadian Education Bond (CLB) is a financial incentive launched by Canada's Department of Employment and Social Development to encourage low-income parents to save for their children's higher education. This is done through Registered Education Savings Plans (RESPs). Starting in 2015, Only one in 3 eligible children has benefited from the Canada Education Bond. Today, Nearly two million children are entitled to the bond. The Department of Employment and Social Development of Canada runs a research program that includes randomized controlled trials and qualitative design thinking methodologies to further understand the challenge and increase the appetite for Canadian education.

In 2016, The department conducted the first experiment based on behavioral insights in an effort to increase the demand for Sindh by sending mails that motivate eligible families to benefit from the program. The thesis design was based on relevant recent research in the field of behavioral insights, As reviewed by a number of academics, The results, however, were disappointing. Although we have succeeded in conducting a randomized, disciplined trial within the government sector (which is a unique precedent), Many of the messages used during the experiment reduced the turnout for the bond compared to the unified messages that the program used in communication previously. The findings indicated that existing research was not enough to develop solutions to promote savings behaviors among low-income Canadians.

In order to better understand the educational and financial decision-making process of low-income families, We launched a design project that combines for the first time in the Canadian government the qualitative results of the design project, and the Quantitative Behavioral Insights Project. We engaged 146 people in this project through interviews, workshops and ethnographic observations on interaction with services. also We involved parents, Students (including those of indigenous and remote descent) and teachers, Providers of registered education savings plans, And stakeholders from various disciplines in government and community agencies. The joint development of research tools between us, our clients and other stakeholders has ensured the integrated collection of qualitative data.

We were then able to design messages based on findings and insights from qualitative field research. This research has found many insights and ideas, including those described below. These insights or observations were reflected in the intervention procedures that were tested to assess their effectiveness. For example:

Observations from qualitative research → lead to → intervention or hypothesis tested through the letters.

Citizens pointed out that the unconfirmed information contained in the letters (such as the minimum or maximum amount a child could obtain from the Canadian Education Bond) was the most confusing in their eyes. Moreover, Older children can receive higher amounts and claim them retroactively (although most parents were unaware of this) ← ← personalizing the content of the message may reduce confusion and raise the likelihood of the bond being accepted so that the letter displays exactly how much the child is entitled to (to date).

The focus on the concept of "saving" has not resonated with all low-income families. Some of these families think they are unable to save. This ← lead to ← reduced focus on saving in the message, which may increase the demand for the program.

To test the validity of those visions and ideas, We conducted a randomized, disciplined trial targeting more than 140,000 eligible children as young as 12 or 13. Previous experience has indicated that this segment, representing older children who meet the requirements, She was the least to benefit from the bond.

The results of the experiment were very positive, The messages combining the two visions outlined above boosted the uptake of the Canadian Education Bond by 55% compared to the consolidated letter (8.8% compared to 5.7% respectively). In practice, If we send 10,000 standardized messages, We expect 573 beneficiaries to register for the Canadian Education Bond. In total, these families will receive $570,000 in education savings from the government. If we send a message based on behavioral insights and the design project, 878 beneficiaries will register for the program. In total, these families will receive $978,000 in education savings from the government.

The results of this innovation have been implemented into the CLB communications, which are now being received by eligible children.

also The design project resulted in a number of additional insights that will be tested on other segments of children eligible to benefit from the program through randomized controlled trials.

Given the success of this mixed-method methodology, We are considering using them to deal with a number of new challenges faced by management.

What makes your project innovative?

This project is an appropriate example of the successful use of mixed approaches that combine design and behavioral visions, It is the first of its kind in the Canadian government. Specifically, We gained a better understanding of the needs and experiences of low-income Canadians using ethnographic research, We then combined those insights into communication materials based on behavioral information. And then These communication materials were tested using precise methods (randomized controlled trials). Given the success of the insight-based messaging experience from ethnographic research and its results outperform the standardized messages used in communication materials that are consistently sent to eligible families for the programme, This project demonstrates the complementarity between the two methods and the benefits of testing lessons learned from a large-scale (qualitative) design project in experimenting with behavioral insights. The findings from this trial have been more widely adopted in the Canadian Education Support Program.

Where is the innovation initiative?

After the success of the first experiment of behavioral insights that were directly guided by the design project, The Canadian government currently applies the principles of best-yielding letters to regular communication materials sent under the Canada Education Support Program. also Regional governments, municipalities and community entities seek to apply these findings to their communication materials. We are currently participating in a number of new experiments to test and verify other insights from the design project. This further enhances our knowledge. There are still many other insights from the design project that can be applied to intervention procedures based on behavioral insights other than messages. We plan to test and verify those insights through new experiences. In addition We seek to expand the use of hybrid approaches in dealing with other important challenges facing management in light of the success of this innovation. also We are working to disseminate these findings both to the Canadian government and globally in the near future.

Participating entities

The first behavioral insights project involved academic professionals who contributed their expertise in the field of behavioral sciences. also The design project saw collaboration with regional governments and municipalities to better understand internal working conditions and challenges facing the programme. We interviewed 126 Canadian citizens to find out their stories, concerns and future aspirations for higher education. We were also guided by community and educators' views on best practices for dealing with low-income Canadians.

Users, stakeholders, and beneficiaries

Low-income Canadian families are the primary beneficiaries of innovation, because of it, More are currently benefiting from the Canadian Education Bond. also Other community and government entities at various levels benefit from lessons learned, We continue to share results so that everyone can better promote the Canadian Education bond among eligible families. The implementation of this project within the government sector has contributed to enhancing interest in the use of mixed methods in other areas.

Results and impacts

Outcomes include increased uptake of the program (i.e., increased education savings for low-income Canadians) and improved understanding of users (e.g., understand their perceptions of education and higher education savings) and the ability of management to serve this group. A controlled randomized trial was conducted to compare the effectiveness of different messages sent on an experimental basis. Standardized messages (original without alerts) increased the uptake of the Canadian Education Bond (5.7%) compared to a group that did not receive any messages (2.2%). Based on the insights from the design project, Determining how much a family will receive when opening an account within the program has led to an increase in footfall compared to using unified messages (turnout increased to 8%), Reduced emphasis on "saving" (the term did not resonate with many low-income respondents) also contributed to an increase in footfall (to 8.8%). also The combination of behavioral and design insights resulted in a 55% increase in footfall compared to unified messaging.

Challenges and failures

This project would not have been implemented had the first letter experiment conducted in 2016 not failed. Where that experience showed that the best alerts that were prepared achieved similar results to the results of using the unified messages of the program, Other alerts significantly reduced the turnout. This failure exposed the lack of proper understanding of the clients and highlighted the need to develop the design project. The launch and implementation of the design project required enormous investments in terms of time, It involved going through a lengthy process of obtaining permission from the Privacy Department to authorize us to speak with Canadian citizens. Despite the successes we have achieved by integrating behavioral thoughts into messages and increasing the turnout, However, it is clear that the messages will not achieve the desired goals in terms of enrollment in the program. So we suggested other lessons learned resulting from the design project that could further transform the nature of the program. However, Our wide-ranging proposals have met with a degree of opposition from those in charge of the programme.

Conditions for success

The Innovation Lab combined design and behavioral insights, These are two complementary methodologies that are of great importance in the completion of the project. Another essential element in the success of this project is access to data related to the fulfillment of conditions and enrollment in the Canadian Education Support Program. This allowed us to conduct a randomized controlled trial and analyze its results internally. On the other hand, The Innovation Lab has received great support and cooperation from senior management, Without it, support and assistance in conducting experiments and innovation in the government sector, This project would not have been realized. also We were able to hire the right competencies in the fields of design, ethnographic research, behavioral insights, and data analysis. The funding included travel costs to interview citizens in their homes and with community entities. With regard to values, This project would not have been possible without the openness and acceptance of the challenges related to innovation by the organizers of the Canadian Education Support Program. Had it not been for the trust of the citizens we interviewed in the government and their belief in the importance of education.

The possibility of replicating the experience

This is the first project of this magnitude implemented by the Canadian government. Given its success, We are currently studying in the Department of Employment and Social Development the use of this mixed-method methodology in other programs, Such as unemployment benefits and student loans. also There are other similar efforts to implement projects in mixed ways in other federal departments, But they are smaller, less complex projects. It is worth noting that this methodology can be replicated in government agencies or other institutions (such as private entities and non-profit entities). However, the Innovation Lab creates a unique environment conducive to innovation and experimentation, It also provides some skills that are not usually available in other government agencies. However, This project sends a strong message that demonstrates the importance of building innovation and experimentation capabilities at the internal level in government entities.

Lessons learned

We have come a long way in gathering design and behavioural insights to better understand low-income citizens. We have come to realize how low-income earners interact with programs that stimulate savings. For example, A significant increase in the turnout for these programs can be achieved by focusing on the benefits, Instead of stimulating saving, Clarifying uncertainties, Such as mentioning approximate amounts that the beneficiary may receive. Finally, Canadian citizens expect a certain pattern in government letters. We learned from this project that "simplicity is the best way" (modern ornate letters with lots of pictures reduce turnout) and that we must respect this in our correspondence so that it looks serious and acceptable (and not confused with spam).

Experimental methods can highlight how effective programs and services are. When management realizes how effective the experiment is, She becomes willing to learn and innovate more.

The success of this project depended on non-traditional skills that are rarely applied in government, These include ethnographic research, design, and behavioral sciences. Looking from a broader perspective, We will find that the successes and failures we have seen in our experiments revealed that it is not possible to use a single methodology when relying on behavioral insights. Although decades of research supports the principles of behavioral insights, However, the reality is that interventions based on these principles are not appropriate for all contexts, behaviors and population segments. Our research suggests that this applies specifically to vulnerable populations (low-income families) facing unique challenges. Given that this population was rarely included in research on behavioral insight-based interventions, There is a need to follow a customized methodology to meet their needs and adapt their experiences. The mixed-style methodology has proven to be a powerful tool to help us tailor messages to the segments involved. Specifically, Qualitative research has helped us to better understand the beneficiaries of the program. This, in turn, resulted in fundamental insights that we relied on to formulate more effective messages.