OBJECTIVES:

We aim to examine the effects of the Child Development Fund (CDF) program on the long-term psychological, health, social, and financial development of children in poverty.

METHODS:

The current study was a 4-year follow-up survey study (the follow-up study) of a nonrandomized controlled trial (the first study), conducted in 2019 and 2015, respectively, in Hong Kong. Subjects were 902 young adults who were from families living in poverty and receiving financial assistance from the government and who had completed the first study. Approximately 61% of the subjects (N = 546) completed the current follow-up study.

RESULTS:

Of the 546 subjects (46% male), 335 (61%; CDF participants) completed the CDF program between 2011 and 2015, and 211 (39%; CDF nonparticipants) were matched controls in the first study. The mean age was 20.64 years (SD = 2.72). After adjustments of covariates, the CDF participants reported fewer behavioral problems, higher levels of study motivation, higher levels of hope, and more positive attitudes toward their future education. The CDF participants also perceived greater social support from significant others and reported greater amounts of money saved every month. After further adjustments, CDF participants showed greater improvements in health-related quality of life related to emotional functioning. A considerable number of families of the CDF participants who were working for a living were no longer reliant on financial assistance from the government.

CONCLUSIONS:

The positive effects of the CDF program on children in poverty could be sustained throughout adolescence to young adulthood. Findings warrant the promotion of the program to reach more children.

What’s Known on This Subject:

The benefits of Child Development Accounts can be long-term, and researchers have shown that the positive effects Child Development Accounts on financial outcomes among children can last for as long as 7 years.

What This Study Adds:

Young adults who had participated in the Child Development Fund Program showed better health, psychological, and social outcomes and reported higher likelihood of saving habits and greater saving amounts. Working young adults help families leave the poverty line.

Child Development Accounts (CDAs) are built on the assumption that asset holdings promote children’s development in different aspects.1,2  CDAs, unlike traditional antipoverty policies that consist of cash assistance and safety nets only, generally involve subsidized or matched savings to help children accumulate assets for future developmental needs3  and help reduce poverty by encouraging asset holding and accumulation. Asset accumulation and account holding may enhance long-term socio-emotional, educational, and financial development among children through multiple pathways.2,46  Not only may the savings directly help children pay for future education,7  but the asset-building process may also help children learn financial management and nurture their study motivation, confidence, and future orientation.8,9  The benefits of CDAs can be long-term, and research has revealed that the positive effects of CDAs on financial outcomes among children can last for as long as 7 years.10 

Empirical evidence supports the effectiveness of CDA-related interventions in the promotion of children’s development.4  Apart from CDAs, mentorship programs are one of the most-common components in interventions. Findings from meta-analytical research reveal that mentoring could be a powerful strategy that benefits disadvantaged children by providing extrafamilial social support.11  Mentors may serve as significant others to provide guidance for children. Such support may compensate for children’s lack of social resources, which in turn empowers them to overcome challenges in life.12 

CDAs and relevant interventions have been widely implemented in different countries.3  In Hong Kong, the Child Development Fund (CDF) program is a 3-year intervention established by the government in 2008. The CDF program aims to promote the long-term development of children, aged 10 to 16 years, from families receiving government subsidy and to reduce intergenerational poverty by encouraging children to create plans for their future and build tangible and intangible assets to fulfill their goals.13  It rests on the premise that both types of assets could enhance one’s future orientation,14  which then motivates one to assign a greater value to future events and work hard to accomplish one’s ambitions.15  A lack of resources may often hinder children in poverty from having positive future orientations.14  As compensation, the CDF program aims to provide opportunities for these children to build assets and thus develop a positive future orientation and enhance long-term development. To achieve these objectives, the CDF program is designed with 3 major components: the CDAs, the mentorship program, and the creation of personal development plans. It is believed that these interactive components could work closely together to have positive effects on children’s development in various aspects and lead to a reduction in poverty. Details of the components and the conceptual framework of the CDF program are depicted and explained in Fig 1.

FIGURE 1

Conceptual framework of the CDF program.

FIGURE 1

Conceptual framework of the CDF program.

Close modal

To evaluate the CDF program, we have conducted 2 longitudinal nonrandomized controlled trials. The first study was conducted in January to June 2016 to examine the effects of the CDF program.16  The findings demonstrate that CDF participants outperformed the nonparticipants by reporting fewer behavioral problems, greater social support, greater motivation and interest in studying, better communication skills, more-positive future orientation, and greater commitment to their future education and careers. Overall, the impacts of the CDF program on children’s short-term and long-term development have been promising.

Although plenty of empirical evidence supports the positive impacts of CDAs and relevant interventions on children in poverty, research on the sustainability of the effects of these interventions on children’s development through young adulthood is scarce. In this study, we set out to examine the long-term effects on participants’ development ∼5 years after they had completed the CDF program, aiming to compare the health, psychological, and social outcomes between the CDF participants and nonparticipants. Some of the participants have started their career and thus may have impact on poverty reduction. To evaluate the longer-term financial development of the participants, we also evaluated 2 CDA-related indicators among the young adults: saving habits and amount of savings per month. On the basis of the conceptual framework and previous findings,13,16  it was hypothesized that the CDF participants would have more-positive future orientation, leading to better outcomes than nonparticipants.

This follow-up study is a longitudinal nonrandomized controlled trial evaluating the CDF program in Hong Kong. It was conducted in early 2019 to evaluate the long-term effects on participants who had completed the CDF program between 2011 and 2015. Families of all subjects of the intervention group (CDF participants) and comparison group (nonparticipants) had been receiving financial assistance from the government during the CDF program period (2011–2015), indicating that they were living in poverty. A detailed description of the first study has been published elsewhere.16 

In the follow-up study, we invited all 902 adolescent and young adults (17–27 years old) who had participated in the first study. Of them, 552 were CDF participants and 350 were nonparticipants who formed a matched comparison group. A total of 335 CDF participants and 211 nonparticipants successfully completed the follow-up study (N = 546), giving a response rate of 61%. Detailed profile of the nonrandomized controlled trial is presented in Fig 2. There was no significant difference in age and sex comparing respondents and nonrespondents. Both CDF participants and nonparticipants were asked to complete a structured questionnaire after they had provided written informed consent. Participation was voluntary. Anonymity and confidentiality were ensured. Subjects had reported no harm or negative effect during the intervention period and the studies. No incentives were given to the subjects. Subjects were assured that their eligibility of receiving any government subsidy or financial assistance would never be affected by the responses they gave in this study.

FIGURE 2

Profile of the nonrandomized controlled trial (the follow-up study).

FIGURE 2

Profile of the nonrandomized controlled trial (the follow-up study).

Close modal

Behavioral Problems

Behavioral problems were measured with 4 5-item subscales of the Chinese version of the Strengths and Difficulties Questionnaire, 17  including conduct problems, hyperactivity, emotional problems, and peer problems. All items were rated on a 3-point scale; the higher the score, the more severe the behavioral problems.

Study Motivation

Two items, which had been included in the first study,16  were used to measure individuals’ study motivation. Both groups were asked to separately rate their interest in studying and level of happiness caused by studying on a 10-point Likert scale. Higher scores indicated greater interest in and happiness related to studying and learning.

Hope

The 12-item Hope Scale was used to measure individuals’ feelings of hope related to their future.18  The scale comprised 2 subscales: the agency scale measured individuals’ goal-directed determination, whereas the pathway scale measured individuals’ plans to achieve their goals. All items were rated on a 4-point scale; the higher the score, the higher the level of hope for the future.

Future Orientation

Individuals’ attitudes toward their future education and careers were assessed with 2 subscales of the Chinese version of the Exploration and Commitment Questionnaire.19  Each of the 2 subscales contained 14 items that were rated on a 5-point scale. Higher scores indicate greater commitment to one’s future education and career, and thus more-positive future orientation.

To allow comparisons with the first study to be made, the health-related quality of life (HRQoL) of the young adult subjects in the follow-up study was measured by using the Chinese version of the 23-item Pediatric Quality of Life Inventory, Generic Core Scale.20  There were 4 subscales, which measured HRQoL in regard to physical functioning (8 items), emotional functioning (5 items), social functioning (5 items), and school functioning (5 items). Items were rated on a 5-point scale. All ratings in the follow-up study were converted to a scale from 0 (very poor) to 100 (very good). Higher scores indicate better HRQoL.

Perceived social support was assessed with the Chinese version of the 12-item Multidimensional Scale of Perceived Social Support.21  The scale consisted of 3 subscales, which measured individuals’ social support from family, friends, and significant others. All items were rated on a 7-point scale. Higher scores indicate higher levels of social support.

Saving Habits and Amounts

Saving habits were measured with 2 items. One item assessed whether individuals had any saving habits, and the other recorded the average amount of money they saved each month.

Covariates

Age and sex were measured and controlled for in the analysis. To evaluate the longer-term effects of the CDF program, the scores of the outcome measures in the first study were also used as covariates.

Other Demographic, Work-Related, and Economic Characteristics

Educational attainment, employment status, family income, personal income, length of current employment, and family receiving any form of financial assistance from the government were assessed. Individuals’ views of their current job were also rated on a 5-point scale: “not interesting at all,” “slightly interesting,” “quite interesting,” “very interesting,” and “not sure.”

All statistical analyses were performed by using SPSS 25.0 and R Statistical Software v3.4.3 (IBM SPSS Statistics, IBM Corporation). All P values, unless otherwise specified, were 2-tailed, and the significance level was set as .05. Demographic, work-related, and economic characteristics of all subjects were summarized by using descriptive statistics. A series of regression analyses were performed to test the longer-term effects of the CDF program on all outcome measures, with sex and age adjusted as covariates. For each regression model, Cohen d statistics were computed to indicate the effect sizes by showing the mean difference between the CDF participants and nonparticipants in standardized units. To examine the longer-term effects of the CDF program, a differences-in-differences analysis was performed by one-way analyses of covariance (ANCOVAs), with the between-group differences in the follow-up study compiled and the scores in the first study adjusted for. In the ANCOVAs, age and sex were also controlled as covariates. To indicate the effect sizes, partial η2 statistics were used.

Table 1 presents the demographic characteristics of the sample. Overall, 46% of the subjects were male. The mean age was 20.64 years (SD = 2.72). Between-group differences were only found in educational attainment (P < .001): a greater proportion of CDF participants than nonparticipants had achieved a university degree or above (41% vs 25%). Among subjects who were working for a living during the follow-up study, a significantly smaller proportion of CDF participants’ families than nonparticipants’ were receiving financial assistance from the government (7% vs 25%).

TABLE 1

Subjects’ Demographic, Economic, and Work-Related Characteristics by Group

VariableTotal (N = 546)Intervention Group: CDF Participants (n = 335)Comparison Group: Nonparticipants (n = 211)P
Age in y, mean (SD) 20.64 (2.72) 20.80 (2.68) 20.37 (2.77) .07 
Sex, No. (%)    .89 
 Male 249 (46) 152 (45) 97 (46)  
 Female 297 (54) 183 (55) 114 (54)  
Educational attainment, No. (%)    <.001 
 Up to junior high school 10 (2) 5 (2) 5 (2)  
 High school 146 (27) 73 (22) 73 (35)  
 Diploma 44 (8) 21 (6) 23 (11)  
 Higher diploma 84 (15) 54 (16) 30 (14)  
 Associate degree 73 (13) 45 (13) 28 (13)  
 University 189 (35) 137 (41) 52 (25)  
Employment status, No. (%)    .99 
 Student 345 (63) 211 (63) 134 (63)  
 Working for a living 201 (37) 124 (37) 77 (37)  
Monthly family income in USD, mean (SD) 2351.60 (1097.61) 2346.71 (1147.35) 2359.62 (1014.30) .91 
Family receiving any form of financial assistance from the government, No. (%) 173 (32) 106 (32) 67 (32) .98 
 Students (n = 345) 146 (42) 98 (46) 48 (36) .052 
 Working for a living (n = 201) 27 (13) 8 (7) 19 (25) <.001 
Monthly personal income in USD, No. (%)    .02 
 <$650 21 (12) 13 (12) 8 (12)  
 $651–$1300 29 (17) 11 (10) 18 (27)  
 $1301–$1950 59 (33) 35 (32) 24 (36)  
 $1951–$2600 47 (27) 36 (33) 11 (16)  
 >$2601 20 (11) 14 (13) 6 (9)  
Length of current employment, No. (%)    .86 
 <1 mo 18 (10) 11 (10) 7 (11)  
 1–3 mo 23 (13) 13 (12) 10 (15)  
 4–6 mo 39 (22) 25 (23) 14 (21)  
 6 mo–1 y 32 (18) 22 (20) 10 (15)  
 1–2 y 36 (21) 23 (21) 13 (19)  
 >2 y 28 (16) 15 (14) 13 (19)  
Views on their job, No. (%)    <.001 
 Not interesting at all 9 (5) 3 (3) 6 (9)  
 Slightly interesting 30 (17) 14 (13) 16 (24)  
 Quite interesting 84 (48) 65 (60) 19 (28)  
 Very interesting 20 (12) 15 (14) 7 (10)  
 Not sure 31 (18) 12 (11) 19 (28)  
Saving habits, No. (%)    <.001 
 Yes 286 (52) 206 (61) 80 (38)  
 No 260 (48) 129 (39) 131 (62)  
VariableTotal (N = 546)Intervention Group: CDF Participants (n = 335)Comparison Group: Nonparticipants (n = 211)P
Age in y, mean (SD) 20.64 (2.72) 20.80 (2.68) 20.37 (2.77) .07 
Sex, No. (%)    .89 
 Male 249 (46) 152 (45) 97 (46)  
 Female 297 (54) 183 (55) 114 (54)  
Educational attainment, No. (%)    <.001 
 Up to junior high school 10 (2) 5 (2) 5 (2)  
 High school 146 (27) 73 (22) 73 (35)  
 Diploma 44 (8) 21 (6) 23 (11)  
 Higher diploma 84 (15) 54 (16) 30 (14)  
 Associate degree 73 (13) 45 (13) 28 (13)  
 University 189 (35) 137 (41) 52 (25)  
Employment status, No. (%)    .99 
 Student 345 (63) 211 (63) 134 (63)  
 Working for a living 201 (37) 124 (37) 77 (37)  
Monthly family income in USD, mean (SD) 2351.60 (1097.61) 2346.71 (1147.35) 2359.62 (1014.30) .91 
Family receiving any form of financial assistance from the government, No. (%) 173 (32) 106 (32) 67 (32) .98 
 Students (n = 345) 146 (42) 98 (46) 48 (36) .052 
 Working for a living (n = 201) 27 (13) 8 (7) 19 (25) <.001 
Monthly personal income in USD, No. (%)    .02 
 <$650 21 (12) 13 (12) 8 (12)  
 $651–$1300 29 (17) 11 (10) 18 (27)  
 $1301–$1950 59 (33) 35 (32) 24 (36)  
 $1951–$2600 47 (27) 36 (33) 11 (16)  
 >$2601 20 (11) 14 (13) 6 (9)  
Length of current employment, No. (%)    .86 
 <1 mo 18 (10) 11 (10) 7 (11)  
 1–3 mo 23 (13) 13 (12) 10 (15)  
 4–6 mo 39 (22) 25 (23) 14 (21)  
 6 mo–1 y 32 (18) 22 (20) 10 (15)  
 1–2 y 36 (21) 23 (21) 13 (19)  
 >2 y 28 (16) 15 (14) 13 (19)  
Views on their job, No. (%)    <.001 
 Not interesting at all 9 (5) 3 (3) 6 (9)  
 Slightly interesting 30 (17) 14 (13) 16 (24)  
 Quite interesting 84 (48) 65 (60) 19 (28)  
 Very interesting 20 (12) 15 (14) 7 (10)  
 Not sure 31 (18) 12 (11) 19 (28)  
Saving habits, No. (%)    <.001 
 Yes 286 (52) 206 (61) 80 (38)  
 No 260 (48) 129 (39) 131 (62)  

Table 1 also shows that a greater proportion of the CDF participants reported higher personal income than the nonparticipants. The CDF participants tended to report greater interest in their current jobs than nonparticipants. Furthermore, a greater proportion of the CDF participants had saving habits than their counterparts (61% vs 38%).

Results of the regression analyses are summarized in Table 2. Significant predictive effects of group (CDF versus non-CDF) on different developmental outcomes were observed. After adjustments for age and sex, the CDF participants reported fewer behavioral problems (B = −2.04, P < .001), particularly in regard to emotional symptoms (B = −0.69, P < .01), peer problems (B = −0.46, P < .001), and conduct problems (B = −0.67, P < .001). They also showed higher levels of motivation to study (B = 0.74, P < .01) and experienced more happiness when studying (B = 0.22, P < .01). A higher level of hope (B = 1.11, P < .001), in particular greater hope related to better planning to achieve future goals (B = 0.74, P < .001), and more-positive attitudes toward their future education (B = 5.11, P < .001) and future careers (B = 5.71, P < .001) were reported among the CDF participants. The CDF participants perceived greater social support from significant others than their counterparts (B = 0.24, P < .01), and they reported saving greater amounts of money every month (B = 69.08, P < .01). The latter between-group difference was only significant among subjects who were working for a living (B = 190.71, P < .001). There was no significant effect of CDF participation on HRQoL in terms of physical, emotional, social, and school functioning.

TABLE 2

Effects of the CDF Program in the Follow-up Study

VariableEffect of Participationa (95% CI)PCohen d
Psychological outcomes    
 Behavioral problems    
  Emotional symptoms −0.69 (−1.10 to −0.28) <.01 −0.18 
  Peer problems −0.46 (−0.72 to −0.19) <.001 −0.18 
  Conduct problems −0.67 (−0.93 to −0.40) <.001 −0.27 
  Hyperactivity −0.23 (10.55 to 0.09) .17 −0.08 
  Total difficulty −2.04 (−2.93 to −1.14) <.001 −0.24 
 Study motivation    
  Motivation to study 0.47 (0.18 to 0.76) <.01 0.17 
  Happiness caused by studying 0.22 (0.08 to 0.35) <.01 0.17 
 Hope    
  Pathway 0.74 (0.47 to 1.02) <.001 0.29 
  Agency 0.36 (0.03 to 0.70) .03 0.12 
  Total 1.11 (0.57 to 1.64) <.001 0.22 
 Future orientation    
  Future education 5.11 (3.39 to 6.83) <.001 0.32 
  Future career 5.71 (3.96 to 7.46) <.001 0.35 
Health outcomes    
  HRQoL    
  Physical functioning 1.25 (−1.18 to 3.69) .31 0.06 
  Emotional functioning 0.90 (−2.54 to 4.34) .61 0.03 
  Social functioning 1.14 (−1.75 to 4.02) .44 0.04 
  School functioning −1.42 (−4.34 to 1.51) .34 −0.05 
  Total 0.47 (−1.95 to 2.88) .70 0.02 
Social outcomes    
 Perceived social support    
  From significant others 0.24 (0.07 to 0.40) <.01 0.15 
  From family 0.19 (0.02 to 0.37) .03 0.12 
  From friends 0.17 (0.00 to 0.34) .05 0.11 
  Total 0.20 (0.05 to 0.35) <.01 0.14 
CDA-related outcomes    
 Savings amount, USD    
  Overall (N = 546) 69.08 (27.50 to 110.66) <.01 0.18 
  Among students (n = 345) 13.92 (−0.09 to 27.93) .52 0.11 
  Among subjects who were working for a living (n = 201) 190.71 (84.61 to 296.81) <.001 0.19 
VariableEffect of Participationa (95% CI)PCohen d
Psychological outcomes    
 Behavioral problems    
  Emotional symptoms −0.69 (−1.10 to −0.28) <.01 −0.18 
  Peer problems −0.46 (−0.72 to −0.19) <.001 −0.18 
  Conduct problems −0.67 (−0.93 to −0.40) <.001 −0.27 
  Hyperactivity −0.23 (10.55 to 0.09) .17 −0.08 
  Total difficulty −2.04 (−2.93 to −1.14) <.001 −0.24 
 Study motivation    
  Motivation to study 0.47 (0.18 to 0.76) <.01 0.17 
  Happiness caused by studying 0.22 (0.08 to 0.35) <.01 0.17 
 Hope    
  Pathway 0.74 (0.47 to 1.02) <.001 0.29 
  Agency 0.36 (0.03 to 0.70) .03 0.12 
  Total 1.11 (0.57 to 1.64) <.001 0.22 
 Future orientation    
  Future education 5.11 (3.39 to 6.83) <.001 0.32 
  Future career 5.71 (3.96 to 7.46) <.001 0.35 
Health outcomes    
  HRQoL    
  Physical functioning 1.25 (−1.18 to 3.69) .31 0.06 
  Emotional functioning 0.90 (−2.54 to 4.34) .61 0.03 
  Social functioning 1.14 (−1.75 to 4.02) .44 0.04 
  School functioning −1.42 (−4.34 to 1.51) .34 −0.05 
  Total 0.47 (−1.95 to 2.88) .70 0.02 
Social outcomes    
 Perceived social support    
  From significant others 0.24 (0.07 to 0.40) <.01 0.15 
  From family 0.19 (0.02 to 0.37) .03 0.12 
  From friends 0.17 (0.00 to 0.34) .05 0.11 
  Total 0.20 (0.05 to 0.35) <.01 0.14 
CDA-related outcomes    
 Savings amount, USD    
  Overall (N = 546) 69.08 (27.50 to 110.66) <.01 0.18 
  Among students (n = 345) 13.92 (−0.09 to 27.93) .52 0.11 
  Among subjects who were working for a living (n = 201) 190.71 (84.61 to 296.81) <.001 0.19 
a

The effects of participation in the CDF program were adjusted in regard to age and sex. Values reported are unstandardized regression coefficients (B) associated with the predictor “group” (1 = CDF group; 0 = Non-CDF group).

Table 3 shows the ANCOVA results. After adjustments to age, sex, and the scores in the first study, significant between-group differences remained in aspects of behavioral problems, study motivation, hope related to better planning to achieve goals, future, and the amount of money saved every month (all P < .05). The between-group differences in social support from significant others became nonsignificant, whereas that in HRQoL related to emotional functioning became significant after additional adjustments were made to the scores in the first study. Specifically, the CDF participants reported an improvement in emotional HRQoL in the follow-up study, whereas nonparticipants reported a deterioration in this respect (P < .01).

TABLE 3

ANCOVA Results for Outcome Factors Between the First Study and the Follow-up Study by the Group of Participants in the CDF Program

VariableMean (SD)PaPartial η2
Intervention Group: CDF Participants (n = 335)Comparison Group: Nonparticipants (n = 211)
First StudyFollow-up StudyFirst StudyFollow-up Study
Psychological outcomes       
 Behavioral problems       
  Emotional symptoms 3.01 (2.25) 2.52 (2.34) 3.50 (2.20) 3.24 (2.43) .03 0.01 
  Peer problems 3.09 (1.52) 2.98 (1.52) 3.41 (1.41) 3.44 (1.63) .03 0.01 
  Conduct problems 2.27 (1.60) 2.09 (1.44) 2.97 (1.78) 2.80 (1.80) .03 0.01 
  Hyperactivity 3.36 (2.14) 3.19 (2.01) 3.54 (1.72) 3.46 (1.67) .27 <0.01 
  Total difficulty 11.73 (5.34) 10.78 (5.05) 13.41 (5.28) 12.94 (5.57) <.01 0.01 
 Study motivation       
  Motivation to study 6.69 (1.76) 6.94 (1.69) 6.45 (1.62) 6.43 (1.69) <.01 0.02 
  Happiness caused by studying 3.57 (0.79) 3.72 (0.82) 3.47 (0.78) 3.49 (0.75) <.01 0.02 
 Hope       
  Pathway 11.91 (1.60) 12.22 (1.58) 11.31 (1.61) 11.47 (1.60) <.01 0.02 
  Agency 11.04 (1.84) 11.08 (1.89) 10.62 (1.88) 10.69 (1.96) .68 <0.01 
  Total 22.96 (2.99) 23.30 (3.07) 21.93 (2.99) 22.16 (3.12) .07 0.01 
 Future orientation       
  Future education 62.30 (9.59) 63.50 (9.42) 59.52 (10.43) 58.48 (10.80) <.001 0.05 
  Future career 62.50 (10.22) 64.07 (10.14) 58.50 (9.88) 58.20 (10.17) <.001 0.04 
Health outcomes       
  HRQoL       
  Physical functioning 86.93 (13.43) 86.62 (12.88) 86.08 (15.50) 85.43 (15.73) .37 <0.01 
  Emotional functioning 74.88 (20.31) 77.55 (19.55) 78.67 (18.58) 76.47 (20.38) <.01 0.02 
  Social functioning 84.81 (17.19) 85.55 (16.72) 84.60 (16.19) 84.24 (16.76) .20 <0.01 
  School functioning 75.36 (17.54) 83.52 (18.54) 75.62 (18.42) 83.58 (19.55) .44 <0.01 
  Total 80.49 (14.42) 83.31 (13.36) 81.24 (14.86) 82.43 (15.35) .16 <0.01 
Social outcomes       
 Perceived social support       
  From significant others 5.60 (0.97) 5.61 (0.97) 5.39 (1.03) 5.38 (0.98) .08 0.01 
  From family 5.41 (0.97) 5.37 (0.98) 5.20 (1.02) 5.18 (1.10) .44 <0.01 
  From friends 5.55 (0.94) 5.52 (0.97) 5.31 (1.09) 5.35 (1.04) .91 <0.01 
  Total 5.52 (0.85) 5.50 (0.84) 5.30 (0.96) 5.30 (0.94) .32 <0.01 
CDA-related outcomes       
 Savings amount, USD       
  Overall (N = 335) 56.68 (121.84) 142.06 (309.03) 30.02 (81.21) 58.76 (133.78) <.01 0.01 
  Among students (n = 345) 28.50 (58.60) 38.46 (78.80) 14.83 (36.50) 18.86 (38.34) .33 <0.01 
  Among subjects who were working for a living (n = 201) 104.63 (175.41) 318.34 (446.08) 56.44 (121.56) 128.21 (197.95) <.01 0.04 
VariableMean (SD)PaPartial η2
Intervention Group: CDF Participants (n = 335)Comparison Group: Nonparticipants (n = 211)
First StudyFollow-up StudyFirst StudyFollow-up Study
Psychological outcomes       
 Behavioral problems       
  Emotional symptoms 3.01 (2.25) 2.52 (2.34) 3.50 (2.20) 3.24 (2.43) .03 0.01 
  Peer problems 3.09 (1.52) 2.98 (1.52) 3.41 (1.41) 3.44 (1.63) .03 0.01 
  Conduct problems 2.27 (1.60) 2.09 (1.44) 2.97 (1.78) 2.80 (1.80) .03 0.01 
  Hyperactivity 3.36 (2.14) 3.19 (2.01) 3.54 (1.72) 3.46 (1.67) .27 <0.01 
  Total difficulty 11.73 (5.34) 10.78 (5.05) 13.41 (5.28) 12.94 (5.57) <.01 0.01 
 Study motivation       
  Motivation to study 6.69 (1.76) 6.94 (1.69) 6.45 (1.62) 6.43 (1.69) <.01 0.02 
  Happiness caused by studying 3.57 (0.79) 3.72 (0.82) 3.47 (0.78) 3.49 (0.75) <.01 0.02 
 Hope       
  Pathway 11.91 (1.60) 12.22 (1.58) 11.31 (1.61) 11.47 (1.60) <.01 0.02 
  Agency 11.04 (1.84) 11.08 (1.89) 10.62 (1.88) 10.69 (1.96) .68 <0.01 
  Total 22.96 (2.99) 23.30 (3.07) 21.93 (2.99) 22.16 (3.12) .07 0.01 
 Future orientation       
  Future education 62.30 (9.59) 63.50 (9.42) 59.52 (10.43) 58.48 (10.80) <.001 0.05 
  Future career 62.50 (10.22) 64.07 (10.14) 58.50 (9.88) 58.20 (10.17) <.001 0.04 
Health outcomes       
  HRQoL       
  Physical functioning 86.93 (13.43) 86.62 (12.88) 86.08 (15.50) 85.43 (15.73) .37 <0.01 
  Emotional functioning 74.88 (20.31) 77.55 (19.55) 78.67 (18.58) 76.47 (20.38) <.01 0.02 
  Social functioning 84.81 (17.19) 85.55 (16.72) 84.60 (16.19) 84.24 (16.76) .20 <0.01 
  School functioning 75.36 (17.54) 83.52 (18.54) 75.62 (18.42) 83.58 (19.55) .44 <0.01 
  Total 80.49 (14.42) 83.31 (13.36) 81.24 (14.86) 82.43 (15.35) .16 <0.01 
Social outcomes       
 Perceived social support       
  From significant others 5.60 (0.97) 5.61 (0.97) 5.39 (1.03) 5.38 (0.98) .08 0.01 
  From family 5.41 (0.97) 5.37 (0.98) 5.20 (1.02) 5.18 (1.10) .44 <0.01 
  From friends 5.55 (0.94) 5.52 (0.97) 5.31 (1.09) 5.35 (1.04) .91 <0.01 
  Total 5.52 (0.85) 5.50 (0.84) 5.30 (0.96) 5.30 (0.94) .32 <0.01 
CDA-related outcomes       
 Savings amount, USD       
  Overall (N = 335) 56.68 (121.84) 142.06 (309.03) 30.02 (81.21) 58.76 (133.78) <.01 0.01 
  Among students (n = 345) 28.50 (58.60) 38.46 (78.80) 14.83 (36.50) 18.86 (38.34) .33 <0.01 
  Among subjects who were working for a living (n = 201) 104.63 (175.41) 318.34 (446.08) 56.44 (121.56) 128.21 (197.95) <.01 0.04 
a

The one-way ANCOVA tests were adopted to examine the effects of CDF projects by compiling the differences between the participants and nonparticipants of the follow-up study, controlling for the scores of the first study, as well as the age and sex of participants and nonparticipants.

This study is one of few in which the long-term effects of CDF program on children in poverty are comprehensively evaluated by tracking their health, psychological, social, and CDA-related outcomes from childhood to young adulthood. The findings suggest that the CDF program reduced 18% to 27% of the SDs in behavioral problems and increased 17% of the SDs of study motivation, 29% of the SDs of perceived hope, 32% to 35% of the SDs of future orientation regarding education and careers, 14% of the SDs of extrafamilial social support, and 18% of the SDs of savings per month. A greater proportion of families with CDF participants who were working for a living were no longer reliant on financial assistance from the government than families with nonparticipants (7% vs 25%). More CDF participants have developed saving habits and thus contributed to their family income, helping families leave the poverty line. In addition, the CDF participants outperformed nonparticipants by showing better improvements in HRQoL related to emotional functioning. Despite the nonsignificant impacts on longer-term physical functioning, the effects of the CDF program on children’s development in these aspects were demonstrated as being long-term, because they could be sustained throughout adolescence to young adulthood (ie, at least 5 years after children completed the program).

The effect sizes of the CDF program on children’s development are comparable to other antipoverty interventions around the world, thus providing supportive evidence for the effectiveness of CDA-related interventions in the long-term. For example, the SEED for Oklahoma Kids program had small but significant effects on child social-emotional development 4 years after the children received the CDAs (effect size = 0.08),2  whereas another CDA-related intervention showed a large effect on asset accumulation among adolescents in 7 years.10  Furthermore, the effectiveness of the CDF program is comparable to other non-CDA interventions. In particular, its long-term effects on reductions in children’s behavioral problems are similar to those of Head Start, a comprehensive early childhood intervention for disadvantaged children.22  Given the comparatively low costs regarding time and resources required and the high levels of long-term effectiveness in terms of promoting children’s development, the CDF program was demonstrated in this follow-up study and the previous 2 studies as a successful intervention to help children in poverty.

The effects of the CDF program may be accounted for by the direct and indirect influences of each component in regard to shaping the participants. In the program, children created long-term goals for their future, were guided as they made specific action plans, and saved money to achieve their goals using the action plans. This process, which lasted for 3 years, helped children to gain a sense of hope and empowerment regarding their futures. Previous researchers have demonstrated that, with an enhanced future orientation, children have fewer problematic behaviors, better academic achievements, and better well-being.23  The enhanced psychological, health, and academic outcomes correlated to future orientation may then lead to better employment outcomes during young adulthood. Indeed, the CDF participants in this study were found to be more likely than nonparticipants to finish university education, earn a higher income, and find their job interesting. These better academic and employment outcomes are essential to breaking the cycle of poverty.24 

Enhanced social support and networks may be another pathway through which the CDF program can promote children’s development. With mentoring, children could receive intangible support from mentors who serve as guides and companions. Social support is of particular importance for children, especially those from disadvantaged backgrounds; it has long been demonstrated to promote children’s psychological and social well-being.25  In the follow-up study, CDF participants perceived significantly higher levels of support from significant others than nonparticipants, even 5 years after they had completed the program, reflecting the role played by mentors in participants’ lives.12 

The CDAs in the CDF program helped participants start accumulating assets during childhood and educated them about useful ways to spend their money to fulfill their life goals. The current findings are consistent with previous studies and indicate the positive effects of a matched savings account combined with mentoring in regard to psychological, health, and academic outcomes.3  It is believed that savings may help children’s long-term development both directly and indirectly. Not only may the assets increase one’s level of future orientation and hope, they may also help pay the fees for education or lifelong learning, to facilitate better equipment among disadvantaged children for their futures and help children develop savings habits to further accumulate assets in adulthood.

Findings of this study need to be interpreted with the following limitations. First, findings were limited by the use of self-reported outcome measures. Reporting bias may appear. Second, as a government-commissioned study to evaluate their CDF program, this study is not able to be a randomized controlled trial, and residual confounding may be possible. The findings may not be generalizable to other settings, and careful interpretation is needed. Yet, the CDF participants (who had been selected before the evaluation studies) and nonparticipants (who were matched comparison group selected in the study) were from the same pool in the government register, and their families had been receiving social security assistance before the start of the CDF program. We believe that the 2 groups shared similarities in their demographic and socioeconomic background, and this was demonstrated by the nonsignificant differences in the demographic variables tested in the study. Although this was not a perfect design academically, findings have made an impact on government policy on child poverty, as the promising results have already motivated the Hong Kong government to scale up the CDF program from pilot intervention to a routine program. Nonetheless, future research may consider using a carefully designed randomized controlled trial and track the outcomes longitudinally.

The CDF program was effective in helping children in poverty from childhood to young adulthood. The promising findings of this study warrant the promotion of the program to a wider population to help more children and their families in the territory.

Prof Chan conceptualized and designed the study, drafted the initial manuscript, and reviewed and revised the manuscript; Drs Lo and Yu designed the data collection instruments, collected data, conducted the initial analyses, and reviewed and revised the manuscript; Drs Ho and Ip coordinated and supervised data collection; Drs Ho, Yan, and Ip critically reviewed the manuscript for important intellectual content; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

This trial has been registered with the HKU Clinical Trials Registry (http://www.hkuctr.com/) (identifier HKUCTR-2847).

Deidentified individual participant data (including data dictionaries) will be made available, in addition to study protocols, the statistical analysis plan, and the informed consent form. The data will be made available after publication to researchers who provide a methodologically sound proposal for use in achieving the goals of the approved proposal. Proposals should be submitted to koling.chan@polyu.edu.hk.

FUNDING: Supported by the Labour and Welfare Bureau of the government of the Hong Kong Special Administrative Region.

     
  • ANCOVA

    analysis of covariance

  •  
  • CDA

    Child Development Account

  •  
  • CDF

    Child Development Fund

  •  
  • HRQoL

    health-related quality of life

1
Sherraden
M
.
Assets and the Poor: A New American Welfare Policy
.
Armonk, NY
:
ME Sharpe
;
1991
2
Huang
J
,
Sherraden
M
,
Kim
Y
,
Clancy
M
.
Effects of child development accounts on early social-emotional development: an experimental test
.
JAMA Pediatr
.
2014
;
168
(
3
):
265
271
3
Sherraden
M
,
Lissa
J
,
Cheng
LC
, et al
.
International Child Development Accounts
.
St Louis, MO
:
Washington University in St. Louis
;
2016
4
Shanks
TRW
,
Kim
Y
,
Loke
Y
,
Destin
M
.
Assets and child well-being in developed countries
.
Child Youth Serv Rev
.
2010
;
32
(
11
):
1488
1496
5
Kim
Y
,
Sherraden
M
,
Huang
J
,
Clancy
M
.
Child development accounts and parental educational expectations for young children: early evidence from a statewide social experiment
.
Soc Serv Rev
.
2015
;
89
(
1
):
99
137
6
Elliot
W
,
Jung
H
,
Friedline
T
.
Math achievement and children’s savings: implications for child development accounts
.
J Fam Econ Issues
.
2010
;
31
(
2
):
171
184
7
Nam
U
,
Kim
Y
,
Clancy
M
,
Zager
R
,
Sherraden
M
.
Do child development accounts promote account holding, saving, and asset accumulation for children’s future? Evidence from a statewide randomized experiment
.
J Policy Anal Manage
.
2013
;
23
(
1
):
6
33
8
Nam
Y
,
Huang
J
,
Sherraden
M
. Asset Definitions. In:
McKernan
SM
,
Sherraden
M
, eds.
Asset Building and Low-Income Households
.
Washington, DC
:
Urban Institute Press
;
2008
:
1
32
9
Shapiro
TM
. The Importance of Assets. In:
Shapiro
TM
,
Wolff
EN
, eds.
Assets for the Poor: The Benefits of Spreading Asset Ownership
.
New York, NY
:
Russell Sage
;
2001
:
11
33
10
Clancy
MM
,
Beverly
SG
,
Sherraden
M
,
Huang
J
.
Testing universal child development accounts: financial effects in a large social experiment
.
Soc Serv Rev
.
2016
;
90
(
4
):
683
708
11
Eby
LT
,
Allen
TD
,
Evans
SC
,
Ng
T
,
Dubois
D
.
Does mentoring matter? A multidisciplinary meta-analysis comparing mentored and non-mentored individuals
.
J Vocat Behav
.
2008
;
72
(
2
):
254
267
12
Erickson
LD
,
McDonald
S
,
Elder
GH
 Jr.
.
Informal mentors and education: complementary or compensatory resources?
Sociol Educ
.
2009
;
82
(
4
):
344
367
13
Chan
CC
,
Lai
M
,
Ng
ECW
,
Lau
WSY
.
A review of features and outcomes of the Hong Kong Child Development Fund
.
China J Soc Work
.
2013
;
6
(
2
):
127
148
14
Riccio
J
,
Dechausay
N
,
Greenberg
D
, et al
.
Toward Reduced Poverty Across Generations: Early Findings From New York City’s Conditional Cash Transfer Program
.
New York, NY
:
MDRC
;
2010
15
Chen
P
,
Vazsonyi
AT
.
Future orientation, impulsivity, and problem behaviors: a longitudinal moderation model
.
Dev Psychol
.
2011
;
47
(
6
):
1633
1645
16
Chan
KL
,
Lo
CKM
,
Ho
FKW
,
Zhu
S
,
Lai
SMK
,
Ip
P
.
The longer-term psychosocial development of adolescents: child development accounts and the role of mentoring
.
Front Pediatr
.
2018
;
6
:
147
17
Lai
KY
,
Luk
ES
,
Leung
PW
,
Wong
AS
,
Law
L
,
Ho
K
.
Validation of the Chinese version of the strengths and difficulties questionnaire in Hong Kong
.
Soc Psychiatry Psychiatr Epidemiol
.
2010
;
45
(
12
):
1179
1186
18
Snyder
CR
,
Harris
C
,
Anderson
JR
, et al
.
The will and the ways: development and validation of an individual-differences measure of hope
.
J Pers Soc Psychol
.
1991
;
60
(
4
):
570
585
19
Zhang
LL
,
Zhang
WX
,
Ji
LQ
,
Nurmi
JE
.
Psychometric analysis of adolescent future orientation questionnaire in Chinese cultural context
.
Xinli Fazhan Yu Jiaoyu
.
2006
;
14
:
7
11
20
Lau
JTF
,
Yu
XN
,
Chu
Y
, et al
.
Validation of the Chinese version of the Pediatric Quality of Life Inventory (PedsQL) cancer module
.
J Pediatr Psychol
.
2010
;
35
(
1
):
99
109
21
Chou
KL
.
Assessing Chinese adolescents’ social support: the multidimensional scale of perceived social support
.
Pers Individ Dif
.
2000
;
28
(
2
):
299
307
22
Puma
M
,
Bell
S
,
Cook
R
,
Heid
C
.
Head start impact study: final report. 2010. Available at: www.acf.hhs.gov/sites/default/files/opre/hs_impact_study_final.pdf. Accessed April 3, 2020
23
Chen
P
,
Vazsonyi
AT
.
Future orientation, school contexts, and problem behaviors: a multilevel study
.
J Youth Adolesc
.
2013
;
42
(
1
):
67
81
24
Gofen
A
.
Family capital: how first-generation higher education students break the intergenerational cycle
.
Fam Relat
.
2009
;
58
(
1
):
104
120
25
Chu
PS
,
Saucier
DA
,
Hafner
E
.
Meta-analysis of the relationships between social support and well-being in children and adolescents
.
J Soc Clin Psychol
.
2010
;
29
(
6
):
634
645

Competing Interests

POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.