The introduction of a financial incentive scheme tied to on-time birth registration (the PAPP) from August 2018 until July 2020 had immediate and substantial effects on improving the timeliness of birth registration in Fiji throughout the two-year period during which the payment was available. Among children born from January 2016 to July 2018, prior to the introduction of the PAPP, the average time for a child’s birth to be registered was almost two years (665 days) and only 57% of births were registered on-time (≤ 365 days). The most significant delay in birth registration during the pre-PAPP period was among iTaukei children compared to non-iTaukei children, followed by single mothers compared to married mothers, and then younger mothers (10–19 years) compared to older mothers (≥ 20 years). Female children were found to have greater delays in birth registration than male children, however, the difference was very small and of questionable significance. Following the introduction of the PAPP, the average time for a child’s birth to be registered declined sharply to approximately four months (124 days), and 93% of births were registered on time. The greatest improvements in the timeliness of birth registration were among iTaukei children, mothers aged 10–19 years, and single mothers. However, even during the PAPP, a substantial disparity remained between iTaukei and non-iTaukei children, and, to a lesser extent, between single mothers and married mothers. Whereas disparities between older and younger mothers and male and female children were not present during the PAPP.
When the PAPP was discontinued in August 2020, the immediate increase in the birth-to-registration interval rose to approximately half of the pre-PAPP level. While this may indicate a long-term positive effect of the PAPP on social and cultural norms concerning on-time birth registration, it is also partly explained by artefact due to lower completeness of birth registration data during the post-PAPP period. For 2020, where birth registration completeness was estimated at 82%, the mean birth-to-registration intervals can be expected to increase over time, as unregistered children become registered. Given the clear effect of the PAPP for 2018-19, it would be expected that this would apply more to children born in the latter (August-December post-PAPP) part of 2020. Therefore, a longer follow-up time is required for birth registration completeness to reach ≥ 95% for 2020 onwards before it can be determined whether the mean birth-to-registration intervals during the post-PAPP period return to similar levels as those before the PAPP’s introduction, and the extent to which a possible residual effect of the PAPP on timely birth registration exists. For this reason, analysis of mean birth-to-registration intervals and the likelihood of on-time birth for the post-PAPP period were not undertaken in this study due to a lack of sufficient follow-up time (≠ 5 years).
We estimated the national completeness of birth registration during 2016-17, prior to the introduction of the PAPP, as ≥ 95%. While high levels of completeness can be achieved in the absence of financial incentives, the mean birth-to-registration interval during 2016-17 was almost two years, whereas during the PAPP (August 2018 to July 2020), the mean birth-to-registration interval was four months. In the absence of the PAPP, the main requirements for birth registration and a birth certificate in Fiji are primarily for school enrolment or for obtaining a child’s passport. These are not linked to timely birth registration, with school enrolment not occurring before the child is at least four years of age, and with many children never obtaining a passport. In the absence of incentives tied to timely birth registration, such as the PAPP, it could be expected to take five or more years for the national completeness of birth registration in Fiji to reach ≥ 95%. For instance, among children born in 2020, birth registration completeness may not reach ≥ 95% until 2025, once those children have reached primary school age and require a birth certificate for enrolment.
An increase in birth registration completeness when children are around five years of age has been documented in many countries where birth registration is required for school enrolment, and is often the first incentive for parents to undertake the registration process. This has been observed in Nepal [6], India [7], Uganda [8], and Nigeria [9]. Delayed birth registration, however, leaves a child without a legal identity and has been shown to increase a child’s risk of statelessness, trafficking, and child labour [7, 10, 11] and exclusion from essential services including health care [12, 13]. Delayed birth registration also, importantly, hinders the availability of complete and timely vital statistics on fertility and child mortality that are essential for accurate monitoring and national planning across multiple government sectors, or for reporting progress against global development agenda frameworks such as the Sustainable Development Goals.
The largest disparity in the timeliness of birth registration prior to the introduction of the PAPP, and also whilst the PAPP was available, was between iTaukei children and non-iTaukei children. The reasons why iTaukei children had much larger mean birth-to-registration intervals throughout 2016-20 could be various and need to be better understood. Fijians of Indian descent make up the majority of the non-iTaukei population and may have a greater need for a passport, which may partly contribute to the observed higher timeliness of birth registration in the non-iTaukei population. The intersection between wealth quintile and ethnicity is not well known in Fiji, and it is also possible that mothers of iTaukei children are more likely to be in the lower wealth quintile, and thus poverty may have a greater impact on birth registration due to direct and indirect costs, rather than cultural aspects of ethnicity. Furthermore, geography may play a part since the iTaukei population live in more rural and remote areas than the non-iTaukei population, with greater transport barriers in accessing civil registration offices. Standardisation of geographic birth registration data (currently being undertaken by the civil registry) and disaggregation of household income and expenditure data by ethnicity, will facilitate a better understanding of the current ethnic-specific disparities in the timeliness of birth registration in Fiji.
Direct and indirect costs associated with birth registration are well documented barriers for parents and caregivers completing the birth registration process [14]. Many studies have identified that loss of wages and transportation costs are the main barriers to birth registration, particularly among poorer households [7]. Financial incentives have been shown to be effective in increasing birth registration through a variety of mechanisms in Asia, Africa and Latin America [15]. The financial incentives have either directly aimed to improve birth registration timeliness and completeness, or have included birth registration as an administrative requirement for access to other financial incentive programs or schemes. In India, the implementation of the Majoni scheme in 2009 provided financial incentives for the registration of female children, and within one-year female birth registration increased from 24–39% in the target population [16]. In Zimbabwe, the implementation of a cash transfer program in 2010, which mandated birth registration as a condition of enrolment in the scheme, documented an increase in birth registration from 8–25% within one year among children aged 0–4 years in the target population [17]. South Africa’s Child Support Grant is a nationwide cash transfer to households under the government-determined income threshold, and has been cited as a significant contributor to increasing South Africa’s birth registration from 21% in 1992 to 84% in 2012 [18].
Limitations
The dataset used in this study did not enable individual identification of which birth registrations also accompanied an application for the PAPP (79% overall), and which did not (21%), during the two-year period the PAPP was available. Exceeding the combined parental income threshold likely explains a portion of the latter category. Although the intersection between wealth quintile and ethnicity is not well understood in Fiji, married mothers who had to declare both parents’ income, compared with single mothers, were likely to constitute a larger portion of the group that exceeded the PAPP’s income threshold. This may have resulted in a lower overall effect of the PAPP among married mothers than otherwise if this group had similar access to the PAPP.
The maternal marital status used in this study was recorded at the time of birth registration. When birth registration is delayed, particularly by several years, the maternal marital status may be different to what it was at the time of giving birth. In this study, the number of married women was 10% lower during the PAPP compared to before the PAPP, which suggests that before the PAPP up to 10% of women in the married category may have been single at the time of giving birth. The effect of this proportion of potentially misclassified marital status records would not be expected to substantially change the large and statistically significant difference in the timeliness of birth registration identified between single and married mothers.