Achieving Zero Stunting: A Sustainable Development Goal Interlinkage Approach at District Level

Abstract

The sustainable development goals (SDGs) provide an integrated framework of targets and indicators, including the elimination of stunting, to support better development planning. Indonesia faces a significant challenge as it ranks fourth globally in terms of stunting prevalence, exacerbated by disparities across regions, gender, and socioeconomic status, further compounded by the ongoing COVID-19 pandemic. Given the interlinked nature of SDGs, this study provides empirical support for the prioritization of SDG indicators, primarily in the context of stunting elimination at the district level in Indonesia. This study employed a combination of economic complexity and network theory, utilizing data from a comprehensive set of 54 indicators spanning 28 targets within 13 SDG goals in 514 districts. The analysis is based on network metrics, including revealed comparative advantage (RCA), proximity, centrality, and density to establish the SDG interlinkage network and identify key priority indicators. The findings highlight the importance of prioritizing indicators such as civil registration, health facilities and services, access to basic facilities and housing, and access to ICT in efforts to reduce stunting, particularly among disadvantaged households. Given the unique resources and capacities of each region, our analysis offers district-specific prioritization strategies for stunting elimination.

1. Introduction

In order to create prosperity for people and the environment, in 2015, 191 UN member countries decided to commit to and adopt the 2030 Agenda for Sustainable Development [1,2,3]. SDGs have a broader and more relevant scope in all countries, as well as a growing focus on goals on equity and climate change than its predecessor, the Millennium Development Goals (MDGs) 2000–2015 [3]. The SDGs consist of 17 goals, 169 targets, and 244 indicators, which are comprehensive, holistic, transformative, and based on a sustainable paradigm. The SDG agenda was created integrated, based on the concept that economic prosperity, environmental protection, and social welfare cannot be addressed separately [4,5]. For example, eliminating hunger is linked to poverty reduction, so addressing undernutrition is part of addressing poverty (SDG 2 and SDG 1) [6]. Access to energy can also help women work from home and earn an independent income (SDG 7 and SDG 5) [6]. These examples show that there can be positive links between SDGs, yet improving one may have negative consequences for another, resulting in trade-offs [1]. Therefore, studies on SDG interlinkages are important to understand how specific SDGs connect with others, as this is key to unlocking their full potential [6].

Understanding the interlinkages between the SDGs could direct this study to focus specifically on tackling a specific issue, including stunting. As stunting is caused by multidimensional factors, the interlinkage analysis aims to help policymakers in determining which indicators should be targeted to accelerate the resolution of a region’s stunting problem. Figure 1 shown that the case of stunting in Indonesia in 2021 remains high, at 22.4% [7]. The government has implemented several plans and policies, including the National Movement to Accelerate Nutrition Improvement (Gerakan Nasional Percepatan Perbaikan Gizi/Gernas PPG), the Convergence Action for the 2018 Stunting Prevention Program, Presidential Regulation Number 18 of the 2020 Medium-Term Development Plan (Rencana Pembangunan Jangka Menengah/RPJMN), and Presidential Regulation Number 72 of 2021 concerning Accelerating Nutrition Improvement [8,9]. These efforts resulted in an annual decline in the number of cases of stunting. However, these results are still insufficient and relatively slow, as the rate of stunting in 2022 is far below the target to reduce stunting cases to 10% by 2030 [10]. This case of stunting is classified as urgent and needs to be accelerated because the impact of stunting is not merely a matter of height but rather the low quality of human resources resulting from poor health conditions.

The purpose of this study is to discover the interlinkages between the SDG indicators to identify critical indicators to reduce stunting in Indonesia at district level, using Cirebon District as a case study to be further analyzed. Prior to this, there have been national-level studies on interlinkages between SDG indicators [10,11,12,13,14,15,16,17] but none that studied them at district level. The research conducted at the national level is deemed insufficient to determine the interlinkages between the SDGs at the regional level. Due to variances in regional characteristics, priorities, and legislation, the linkage and efficacy of SDG indicators will vary among areas and cannot be expected to be homogenous as national studies [18]. As a result, the government should take into account the discrepancies in each of these locations to resolve issues with reaching indicators generally. Furthermore, local governance is vital for spearheading and promoting the SDGs [19]. This is due to the fact that the SDGs address societal issues that are visible and felt directly by local communities, organizations, businesses, and families [19].

This study uses economic complexity and network theory to create an interlinkage network for the SDGs in Indonesia using data on 54 SDG indicators from 514 districts. Identifying target relationships is necessitated due to the different impacts of target linkages on stunting problems. Several factors prompted us to conduct research on the Cirebon District. First, Cirebon District’s average readiness to achieve the SDGs remains lower than the average score of districts/cities in West Java Province, particularly in terms of health and education goals [9]. Second, based on data from the 2022 Indonesian Nutrition Status Survey [7], the prevalence of stunting in Cirebon District reaches 18.6% or 15,299 infants distributed across 10 working areas of the public health center, 28 working areas of villages in 9 sub-districts [20].

2. Literature Review

2.1. SDG Interlinkages

Several efforts have been conducted to analyze the interlinkages of the SDGs. They have used different tools and approaches to model the network of interlinkages and analyze clusters of issues to better understand the connections. Le Blanc [12], as the pioneer in SDG interlinkage analysis, built a matrix that connects every SDG target to all the goals to which its wording references and showed that some goals and targets are well connected. UNESCAP [17] conducted a network analysis of 82 indicators from 174 countries and found a dense core of highly connected socioeconomic indicators and a periphery of environmental indicators. IGES [14] used social network analysis (SNA) and found that agriculture activities, food production, safe drinking water, sanitation, energy, and resilient infrastructure are the network’s most important priorities due to their essential roles in connecting other targets. ICSU [6] used a scoring approach to identify causal and functional relations underlying progress in or the achievement of the sustainable development goals and targets. Allen et al. [11] adopted a multi-criteria analysis (MCA) decision framework that assesses and prioritizes SDG targets based upon their “level of urgency”, “systemic impact”, and “policy gap”. El-Maghrabi et al. [13] employed network theory and economic complexity and found that the goals tend to have the strongest relationship with the other goals’ SDG indicators, especially in low- and middle-income countries. SDGs Bappenas [10] used interpretive structural modeling (ISM) and showed that free primary and secondary education, double energy efficiency, universal health coverage, and enhanced renewable energy will help accomplish other SDG targets. Recently, UNDP [21] provided a visualization of SDG interlinkages that represents their synergies and trade-offs taken from various literature.

2.2. Stunting

A large and growing body of literature has investigated various factors related to stunting. For instance, the malnutrition that contributes to stunting is not only related to a lack of family food security (SDG 2) but also the effects of poverty (SDG 1), poor health services (SDG 3), maternal awareness (SDG 4), and access to clean water and sanitation (SDG 6) [22,23,24]. The linkage of these targets results in different relationships between each target based on geographical conditions and different periods [1]. In Malawi and Nepal, the availability of handwashing facilities and clean drinking water does not synergize with stunted children’s growth. It has been proven that increasing access to handwashing facilities or drinking water did not reduce stunted children’s growth [25,26,27]. Meanwhile, in rural areas of India, the linkage between the availability of handwashing facilities and clean water and sanitation (SDG 6) reduced malnutrition issues in children (stunting) [28]. According to Agostoni et al. [29], apart from the health facility factor (SDG 3), malnutrition in children is related to several other factors, such as mother education (SDG 4), agriculture and food systems (SDG 2), network social safeguards (SDG 1), and the availability of clean water and adequate sanitation (SDG 6).

2.3. Cirebon District at a Glance

This analysis focuses on the case study of Cirebon District, located in West Java, Indonesia (map of the study area can be found in Figure 2). Cirebon District spans an area of 1070.29 km² and comprises 40 sub-districts and 424 villages [30,31]. The district has a population of approximately 2.291 million people, resulting in a population density of 2141 individuals per square kilometer. In terms of economic performance, the district experienced a growth rate of 8.3% in 2022, with a Gross Regional Domestic Product (GRDP) of IDR 56.65 trillion or USD 3.8 million [30]. The processing industry sector contributes significantly to the district’s economy, accounting for 20.6% of the GRDP [30].

Poverty remains a concern in Cirebon District, with 12.01% of the population living below the poverty line in 2022, equating to a monthly per capita income of less than USD 28 or IDR 416,914 [30]. This poverty rate is relatively high compared to other districts and cities in West Java. Additionally, the district faces a high prevalence of stunting, with 18.6% of children affected, corresponding to 15,299 infants across 10 working areas of public health centers and 28 working areas of villages in 9 sub-districts [7,20]. The government of Cirebon District has set a target to reduce stunting cases by 3% by 2024 and aims to eliminate stunting entirely by 2030 [32].

Figure 2. Map of Cirebon District. Source: Wikimedia [33,34,35].

Figure 2. Map of Cirebon District. Source: Wikimedia [33,34,35].

3. Materials and Methods

3.1. Data

This study uses SDG indicator data of 2017 and 2018 from 514 districts in Indonesia. The data were collected from the Indonesia Socioeconomic Survey (SUSENAS) and Labor Force Survey (SAKERNAS). In addition, some indicator data are taken from official publications such as Central Agency on Statistics (BPS) and the Ministry of Health and Statistics Indonesia (RISKESDAS). All these data are compiled following the definition and methodology in the Indonesia SDG metadata [36]. Given the availability of the data, this study could only compile data for 70 indicators of 13 goals. Three goals are not included in this study, namely SDG 12 Responsible Consumption and Production, SDG 14 Life Below Water, and SDG 15 Life on Land. However, there are duplicated indicators and several indicators that have no variation across districts and hence could not be analyzed further. Gross enrolment rate for upper-secondary school/equivalent, the percentage of the population living below the national poverty line, the proportion of people who use the internet, and the proportion of people who control/own mobile phones are among the duplicated indicators. After conducting the selection process, a total of 54 indicators were deemed appropriate for further analysis, representing 28 targets across 13 goals (

Table 1. Summary of Selected SDG Indicators.

Goal	Number of Target	Number of Indicator
1	2	8
2	2	4
3	3	5
4	6	15
5	1	2
6	2	2
7	1	1
8	4	9
9	2	2
10	2	2
11	1	1
16	1	2
17	1	1
Total	28	54

). The full list of indicators can be found in the Supplementary file (Table S1: List of SDG Indicators).

3.2. Economic Complexity and Network Theory

The SDG interlinkages in this study are based on the economic complexity and network theory following [13,17] that was first developed as a country development diagnostic framework [37]. The analysis is based on a set of network metrics, including revealed comparative advantage (RCA), proximity, centrality, and density. Inspired by these studies, the current study attempted to adopt the method into the research.

Firstly, the performance of SDG indicators is measured using RCA. RCA measures the performance of SDG indicators in each region based on the capacity of the specified region proxied by HDI’s expenditure per capita. The kernel-weighted local polynomial estimator approach is used to plot SDG indicator data against expenditure per capita. The graph can be used to see whether a region is classified as having an over- or under-performed indicator. A region is considered over-performed if the SDG indicator excels performance expectations based on the region’s capacity. The opposite is used to determine if any SDG indicators are under-performed. Mathematically, RCA consists of binary numbers (0 and 1) describing each indicator’s comparative advantage. The way to get RCA is as follows:

$$RCA\left(i,r\right)=\left\{\begin{array}{c}1,\\ 0,\end{array} \begin{array}{c}{x}_{i,r}>Z_{\alpha =0.05}+E\left(x_{i,r}\right)\\ otherwise\end{array}\right.$$

where RCA(i,r) is the RCA value in specific indicator (i) and region (r), which is 1 when the statistical confidence level exceeds 95% or when the indicator is over-performed and 0 when it is under-performed.

The second step involves measuring the proximity between each SDG indicator. Proximity in this context refers to the degree of closeness or interconnectedness between the SDG indicators. Their proximity to other indicators determines the power to succeed in the SDG indicator. Proximity also indicates a region’s ease of capacity to be used together to achieve the two indicators [13,17]. Based on the set of RCA for each indicator in each region, the proximity is measured by the minimum of the two conditional probabilities; that is, the probability of A performs better given the performance of B, and the probability of B performs better given the performance of A, as in the following equation [13]:

$$P\left(A|B\right)=\frac{P\left(A\cap B\right)}{P\left(B\right)} \ne P\left(B|A\right)=\frac{P\left(B\cap A\right)}{P\left(A\right)}$$

The third step entails measuring centrality. Centrality refers to the importance or prominence of each SDG indicator within the network of interlinkages. It helps identify indicators that play a significant role in connecting and influencing other indicators within the SDG framework. The term centrality refers to the total number of calculated SDG proximity pairs [13,17]. Centrality is used as a connectivity metric, with a high centrality indicating that an SDG is strongly linked with other SDGs [13,17]. The greater the centrality value, the more related (central) the indicator is to other SDG indicators. The centrality could also be an indicator of the potential variety and contribution to the overall performance of the region’s SDG program [13]. It is strongly advised that SDGs with high centrality be redeployed and improved to achieve the SDG goal [13,17].

$$Centralit{y}_j={\displaystyle \sum }_{i}Proximit{y}_{ij}$$

In the fourth step, density is measured to assess the efficiency of achieving an indicator, considering the region’s capabilities and performance on other metrics [13]. The density indicator shows how close an “unsuccessful” SDG is to a region’s successful SDGs. Notably, RCA and density are both indicator- and region-specific, whereas proximity and centrality are merely indicator-specific [13]. The density of region r on specific SDGs’ indicators j is equal to the total proximity of an SDG indicator j to all other SDGs that over-performed divided by the centrality of SDG j, as shown in the equation below [13]. Any SDG has a density value between 0 and 1. The greater the density value of an under-performed SDG indicator, the closer its necessary capacities are to the region’s current capabilities [13].

$$Densit{y}_{rj}=\frac{{\sum }_{i}Proximit{y}_{ij}RC{A}_{ri}}{{\sum }_{i}Proximit{y}_{ij}}$$

Subsequently, a network visualization is generated based on the methodology employed in UNESCAP [17] to illustrate the interlinkages across the SDG indicators. Each indicator is connected to another based on its proximity value, which serves as the weight of the link. The process begins by using the maximum spanning tree (MST) algorithm to create the initial “skeleton” of the network. The MST algorithm establishes N-1 links (N being the number of indicators) that connect all network nodes to their closest partner [17]. Next, all links with a proximity value greater than or equal to 0.4 are added to the minimum spanning tree (MST) to differentiate between nodes with numerous strong connections and those with fewer connections. Once the network is constructed, it is visualized using the Kamada Kawai technique. This visualization method aids in representing the interrelationships between indicators in a visually informative manner.

Finally, drawing from the approach proposed by [13], the selection of priority indicators takes into account various considerations, including (1) indicators with high centrality as its effect on the performance of other SDG metrics is the greatest and (2) indicators with high density, which means they have a sufficient capacity to be over-performed. By considering these factors, the analysis helps identify priority indicators that hold strategic importance and have the potential to drive progress across multiple SDGs.

4. Results

4.1. Performance in SDG Indicators

By using kernel-weighted polynomial smoothing, this analysis can determine the RCA for each indicator in each region. A comparative advantage exists in an indicator if the indicator’s RCA value is 1. This comparative advantage demonstrates that the indicator in the region has better performance relative to other regions’ performance with the same capacity.

Since this study focuses on the stunting indicator, Figure 3 illustrates the correlation of stunting on per capita spending in all Indonesian districts. The position of Cirebon District in the plot is marked by a blue dot. Meanwhile, the other districts in West Java are shown by red dots. This indicator allowed this study to compare the performance of each SDG indicator in the Cirebon District against that of other districts in Indonesia. The red line serves as the fitted line, whereas the shaded region represents the 95% confidence interval. This red line represents the average SDG indicator achievement for each level of per capita spending.

As shown in Figure 3, the position of Cirebon District (blue dot) in the plot is below the fitted line and the shaded region. This position holds that the performance of Cirebon District in resolving stunting is much better (outperforms) than numerous regions in Indonesia (grey diamonds), including most areas in West Java (maroon diamonds) with similar capacity.

Other indicators are also shown in the Supplementary file (Figures S1–S54: Performance of SDG indicators on per capita income). Figure 4 depicts a summary of the classification. This study found that Cirebon District has 28 indicators that are classified as over-performed while the other 26 indicators have below-average performance (under-performed) (see Appendix

Table A1. Summary of RCA, Proximity, Centrality, and Density of SDG Indicators in Cirebon District.

No	Goal	Code	Indicator	RCA	Proximity to Stunting	Rank of Proximity to Stunting	Centrality	Rank of Centrality	Density	Rank of Density
1	1	1.2.1	Poverty	0	0.37	42	21.43	38	0.53	44
2	1	1.4.1.(a)	Birth in Health Facility (B40)	1	0.46	16	26.02	7	0.58	3
3	1	1.4.1.(f)	Urban Slum Households (B40)	1	0.45	20	20.04	45	0.54	33
4	1	1.4.1.(g)	Net Enrollment Rate Primary School (B40)	0	0.40	34	24.02	22	0.57	11
5	1	1.4.1.(h)	Net Enrollment Rate Lower-secondary School (B40)	0	0.50	4	24.65	16	0.55	21
6	1	1.4.1.(i)	Net Enrollment Rate Upper-secondary School (B40)	0	0.47	13	24.90	15	0.56	15
7	1	1.4.1.(j)	Birth Certificate (B40)	1	0.47	11	26.31	1	0.55	23
8	1	1.4.1.(k)	Electricity (B40)	1	0.46	14	26.18	5	0.57	4
9	2	2.1.1.(a)	Underweight	0	0.69	1	21.58	37	0.54	31
10	2	2.2.1	Stunting	1	0.00	54	22.19	35	0.54	39
11	2	2.2.2	Wasting	0	0.50	3	21.40	39	0.55	22
12	2	2.2.2	Overweight	1	0.40	36	21.14	41	0.54	37
13	3	3.1.2	Birth by Skilled Health Personnel	1	0.47	9	26.25	3	0.57	5
14	3	3.1.2.(a)	Birth at Health Facility	1	0.46	19	26.17	6	0.58	2
15	3	3.8.1.(a)	Unmet Need in Health Services	0	0.50	5	22.72	29	0.54	35
16	3	3.8.2	Health Insurance	1	0.42	28	22.27	34	0.53	46
17	3	3.a.1	Smoking	0	0.35	46	20.35	43	0.56	13
18	4	4.1.1.(d)	Gross Enrollment Rate of Primary School	0	0.36	44	20.62	42	0.54	36
19	4	4.1.1.(e)	Gross Enrollment Rate of Lower-secondary School	0	0.46	18	22.82	28	0.55	27
20	4	4.1.1.(f)	Gross Enrollment Rate of Upper-secondary School	1	0.41	30	23.92	23	0.53	41
21	4	4.1.1.(g)	Mean Year School	0	0.38	38	23.60	25	0.54	38
22	4	4.2.2.(a)	Gross Enrolment Rate of Early Childhood Education	1	0.44	23	22.62	32	0.56	16
23	4	4.3.1.(b)	Gross Enrollment Rate of University	0	0.37	43	23.19	26	0.56	17
24	4	4.4.1	Youth with ICT	1	0.51	2	26.28	2	0.57	9
25	4	4.4.1	Adults with ICT	1	0.47	12	25.65	8	0.56	18
26	4	4.5.1	Gender gap (F/M) Net Enrollment Rate of Primary School	1	0.40	35	23.86	24	0.54	32
27	4	4.5.1	Gender gap (F/M) Net Enrollment Rate of Lower-secondary School	0	0.32	48	17.22	50	0.52	48
28	4	4.5.1	Gender Gap (F/M) Net Enrolment Rate Upper-secondary School	1	0.29	51	12.60	53	0.51	51
29	4	4.5.1	Gender Gap (F/M) Net Enrolment Rate of University	0	0.23	53	12.51	54	0.53	40
30	4	4.6.1.(a)	Literacy Rate ≥ 15	0	0.41	31	24.12	21	0.55	24
31	4	4.6.1.(b)	Literacy Rate 15–24	1	0.46	15	25.07	12	0.55	25
32	4	4.6.1.(b)	Literacy Rate 15–59	1	0.45	22	24.91	14	0.53	43
33	5	5.3.1	Married Women < 15	1	0.44	24	24.30	20	0.55	26
34	5	5.3.1	Married Women < 18	1	0.37	41	19.63	47	0.53	45
35	6	6.1.1.(a)	Access to Safely Drinking Water	1	0.42	27	24.35	18	0.56	12
36	6	6.2.1.(b)	Sanitation	1	0.45	21	25.61	9	0.57	10
37	7	7.1.2.(b)	Household Gas	1	0.44	25	25.50	11	0.56	14
38	8	8.1.1	Annual growth rate of GDP per capita	0	0.30	50	13.53	52	0.57	6
39	8	8.1.1.(a)	GDP per capita	0	0.36	45	18.44	49	0.51	52
40	8	8.2.1	Annual growth rate of GDP per employed person	1	0.42	26	22.68	30	0.55	28
41	8	8.3.1	Informal Employment (%)	0	0.33	47	20.29	44	0.52	47
42	8	8.3.1.(a)	Formal Employment (%)	1	0.41	32	24.32	19	0.53	42
43	8	8.3.1.(b)	Informal Agricultural Employment (%)	0	0.26	52	14.89	51	0.52	49
44	8	8.5.1	Hourly Earnings	0	0.31	49	19.16	48	0.51	53
45	8	8.5.2	Unemployment Rate	0	0.41	33	19.93	46	0.55	29
46	8	8.5.2.(a)	Underemployment Rate	1	0.48	6	22.64	31	0.54	30
47	9	9.2.2	Manufacturing employment	1	0.48	7	23.00	27	0.57	7
48	9	9.c.1.(a)	Mobile phone	0	0.42	29	24.55	17	0.55	20
49	10	10.1.1	Gini Coefficient	0	0.38	39	21.69	36	0.51	50
50	10	10.2.1	<50% median income	0	0.37	40	21.24	40	0.50	54
51	11	11.1.1(a)	Adequate Housing	1	0.47	10	25.53	10	0.57	8
52	16	16.1.3.(a)	Physical Violence	0	0.40	37	22.61	33	0.54	34
53	16	16.9.1	Birth Certificate < 5	0	0.48	8	26.21	4	0.59	1
54	17	17.8.1	Access to Internet	1	0.46	17	25.00	13	0.56	19

Note: RCA = 1 classified as over-performed; RCA = 0 classified as under-performed.

for the full-information of indicator status).

4.2. Proximity

Table 2. Summary of Proximity by Goal.

Goal	1	2	3	4	5	6	7	8	9	10	11	16	17
1	-	0.44	0.49	0.45	0.45	0.51	0.56	0.37	0.48	0.44	0.53	0.54	0.48
2	0.44	-	0.43	0.39	0.37	0.42	0.44	0.36	0.44	0.37	0.43	0.44	0.43
3	0.49	0.43	-	0.42	0.44	0.53	0.50	0.38	0.46	0.39	0.53	0.50	0.50
4	0.45	0.39	0.42	-	0.42	0.45	0.47	0.35	0.45	0.40	0.46	0.44	0.48
5	0.45	0.37	0.44	0.42	-	0.46	0.48	0.35	0.41	0.41	0.46	0.45	0.42
6	0.51	0.42	0.53	0.45	0.46	-	0.55	0.40	0.52	0.37	0.79	0.50	0.54
7	0.56	0.44	0.50	0.47	0.48	0.55	-	0.41	0.48	0.45	0.55	0.55	0.50
8	0.37	0.36	0.38	0.35	0.35	0.40	0.41	-	0.40	0.38	0.40	0.38	0.43
9	0.48	0.44	0.46	0.45	0.41	0.52	0.48	0.40	-	0.37	0.53	0.47	0.63
10	0.44	0.37	0.39	0.40	0.41	0.37	0.45	0.38	0.37	-	0.36	0.45	0.35
11	0.53	0.43	0.53	0.46	0.46	0.79	0.55	0.40	0.53	0.36	-	0.48	0.57
16	0.54	0.44	0.50	0.44	0.45	0.50	0.55	0.38	0.47	0.45	0.48	-	0.48
17	0.48	0.43	0.50	0.48	0.42	0.54	0.50	0.43	0.63	0.35	0.57	0.48	-

Source: Authors’ calculations.

shows the average proximity values of 13 Goals, while the proximity matrix of the 54 indicators used in this study can be found in the Supplementary file (Table S2: Proximity Matrix of (54 × 54) Indicators). The greater the proximity value between two SDG indicators, the stronger the interlinkages between them. Strong interlinkages between the two indicators necessitate a comparable level of capability to attain each objective.

Given their proximity value, there are 8 goals out of 13 goals having a strong association with Goal 2 (Zero Hunger). The eight goals are Goal 1 (No Poverty), Goal 3 (Good Health and Well-Being), Goal 6 (Clean Water and Sanitation), Goal 7 (Affordable and Clean Energy), Goal 9 (Industry, Innovation, and Infrastructure), Goal 11 (Sustainable Cities and Communities), Goal 16 (Peace, Justice, and Strong Institutions), and Goal 17 (Sustainable Cities and Communities) (Partnerships for the Goals) (Table 2). With a proximity score of 0.44, Goal 16 (Peace, Justice, and Strong Institutions), Goal 7 (Affordable and Clean Energy), Goal 1 (No Poverty), and Goal 9 (Industry, Innovation, and Infrastructure) have the closest relationship to Goal 2 (Zero Hunger). On the other hand, Goal 8 (Decent Work and Economic Growth) has the lowest proximity value with a value of 0.36.

From the 54 SDG indicators listed, 34 SDG indicators are directly connected to the stunting indicator, as defined by indicators with a proximity value exceeding the 0.4 threshold. Among the 34 indicators are 2.1.1.(a) Underweight, 4.4.1 Youth with ICT, and 2.2.2 Wasting with values of 0.69, 0.506, and 0.502 (see Appendix Table A1). This shows that stunting is strongly related to the indicators of underweight, youth that have access to technology, and wasting in children under five.

4.3. Centrality

As shown in Appendix Table A1, 1.4.1.(j) Birth Certificate, 4.4.1 Youth with ICT, 3.1.2 Birth by Skilled Health Personnel, 16.9.1.(b) Birth Certificate (child), and 1.4.1(k) Electricity (Bottom 40) are among the indicators with the highest centrality, implying that if Cirebon District successfully provides access to these indicators, it may highly influence the success in other SDGs. However, 4.5.1 Gender gap (F/M) Net Enrolment Rate of Lower-secondary School, 8.3.1.(b) Informal Agricultural Employment (%), 8.1.1 Annual growth rate of GDP per capita, 4.5.1 Gender Gap (F/M) Net Enrolment Rate Upper-secondary School, and 4.5.1 Gender Gap (F/M) Net Enrolment Rate of University may not be strongly related to the other SDGs. This suggests that each indicator has very specific, non-transferable conditions for the SDG delivery mechanism [13]. As for 2.2.1 Stunting, this indicator is ranked 35th out of 54 indicators, meaning that the effectiveness of addressing stunting may not be sufficient to promote the achievement of other SDGs.

4.4. Density

As the density value is indicator- and region-specific, this analysis will focus on the Cirebon District density. Cirebon District has 26 indicators that are still below their expected performance based on their capacity (

Table 3. Summary of Centrality and Density of Cirebon District.

		RCA = 0	RCA = 1	Total
	n	26	28	54
Centrality	max	26.21	26.31	26.31
	mean	21.06	23.72	22.44
	min	12.51	12.60	12.51
Density	max	0.59	0.58	0.59
	mean	0.54	0.55	0.55
	min	0.50	0.51	0.50

Source: Authors’ calculations.

). 16.9.1 Birth Certificate <5, 8.1.1 Annual growth rate of GDP per capita, 1.4.1.(g) Net Enrolment Rate Primary School (B40), 3.a.1 Smoking, and 1.4.1.(i) Net Enrolment Rate Upper-secondary School (B40) are those with the highest density values among under-performed indicators (see Appendix Table A1). However, only 16.9.1.(b) Birth Certificate <5 has a high centrality and density value, which means that the region already has the required capacity to make it a successful indicator. As for the rest of the indicators, governments might be able to engage successfully with a high probability of success, but the returns would be relatively small in terms of having a broader impact on various SDG indicators.

Moreover, 16.9.1 Birth Certificate < 5, 3.1.2.(a) Birth at Health Facility (B40), 1.4.1.(a) Birth in Health Facility (B40), 1.4.1.(k) Access to Electricity (B40), and 3.1.2 Birth by Skilled Health Personnel are among the indicators that have the highest density and high centrality values compared to the other indicators. Investing in closely linked and higher-density indicators is likely to be more cost effective, as there will be positive spill-over effects as tightly clustered SDGs will inevitably create synergies, particularly in tackling the stunting problem.

5. Discussions

The findings of the current study indicate that there are indicators in the analysis with a high potential to contribute to the achievement of the SDGs in Cirebon District, particularly in dealing with stunting. In the following section, the analysis discussion is divided into two parts: first, an explanation of the network visualization created to highlight the interlinkages between the SDGs’ subject areas and, second, a discussion of the crucial indicators in addressing stunting.

5.1. Network Visualization

In this study, a visual representation of the SDG indicator network is constructed using proximity values, following the methodology established by UNESCAP [17]. Figure 5 shows a core–periphery pattern, with indicators for health, education, energy, and employment occupying a fairly dense network in the center. More specifically, these indicators are electricity and gas, health facilities, health workers, safe drinking water and sanitation, formal employment, education, and public services. These indicators represent basic needs that will influence the achievement of other network indicators. The network analysis supports the findings of several studies that discuss interlinkages between SDGs [10,11,12,13,14,15,16,17].

In an earlier study, Le Blanc [12] was able to identify a clear indication that SDG 4 (education) belongs to the “core” of the SDG network, besides SDG 2 (hunger), SDG 10 (inequality), SDG 12 (sustainable consumption and production), and SDG 1 (poverty). In addition, their findings also suggest that SDG 8 (growth and employment) is among those at the top of the list, and all have links with 10 other goals or more. A network study by UNESCAP [17] shows the same core–periphery structure, with SDG 3 (health), SDG 2 (hunger), SDG 9 (infrastructure), and SDG 1 (poverty) indicators prominent in the densely connected core. Similarly, Kunčič [15] network mapping shows the strongest positive connections between SDG 3 (health) and the others. Moreover, IGES [14] discovered that SDG 6 (safe drinking water and sanitation) and SDG 7 (energy) are the network’s top priorities together with SDG 9 (resilient infrastructure agriculture) and SDG 12 (food production) due to their critical roles in connecting other targets. Pereira et al. [16] also found that the least complex goals are SDG 9 (industry, innovation, and infrastructure), SDG 3 (health), and SDG 7 (energy). Their findings suggest that the best strategy for countries might have to achieve these less complex goals before moving on to more complex goals in order to fully achieve the 2030 Agenda. Clearly, focusing on these goals will have the greatest impact on the network. This implies that the best strategy for the region to fulfill the 2030 Agenda is to consider the links between each of them.

5.2. Priority Indicators

The priority indicators are chosen by ranking each metric that has already been calculated in the results section. Indicators with proximity values greater than the threshold value of 0.40 and a direct relationship with the stunting indicator are selected, as shown by the green line in Figure 5. Out of the 54 indicators analyzed, 34 indicators meet this criterion.

Following the initial selection of 34 indicators, a subset of those indicators with the highest centrality and density values is chosen. Based on this assessment, eight variables are chosen as priority indicators for dealing with stunting, specifically in Cirebon District, Indonesia (

Table 4. Priority Indicators for Stunting Handling in Cirebon District.

Indicator	Ranking
	Proximity to Stunting	Centrality	Density
Birth Certificate (child)	8	4	1
Birth at Health Facility	19	6	2
Birth in Health Facility (B40)	16	7	3
Electricity (B40)	14	5	4
Birth by Skilled Health Personnel	9	3	5
Adequate Housing	10	10	7
Youth with ICT	2	2	8
Sanitation	21	9	9

Source: Authors’ calculations.

). High centrality indicates that these eight indicators have high connectivity with other indicators in the network. The achievement of indicators with high centrality can contribute to other linked indicators. In addition, if indicators have a high density, it means that these indicators are within reach or, in this case, Cirebon District has most of the capacity needed to succeed in these selected/priority indicators. These priority indicators are divided into four aspects, i.e., civil registration, health facilities and services, access to basic facilities and housing, and access to ICT.

The empirical finding of this study shows that the role of a birth certificate is vital to prevent and reduce stunting. Birth certificates ensure that people (in our case, children) are registered in the system [38]. In a survey of 31 low-middle-income nations, it was discovered that children without birth certificates primarily come from low-income, uneducated households [39]. The absence of birth registration violates children’s fundamental rights, particularly their right to nationality, and may also affect young children’s ability to access specialized health treatments and social welfare programs (such as cash transfer schemes), as well as school enrolment [39]. The findings of a previous study in sub-Saharan Africa evidenced that children aged 2–5 who have birth certificates are less likely to suffer from malnutrition [40]. The reason is that birth certificates can help ensure the accuracy of a child’s nutrition assessment, and the data from the evaluation can help measure child nutrition as a form of child malnutrition prevention or reduction.

Table 5. SDG Priority Indicator Statistics in Cirebon District (%).

Aspect	Indicator	2017	2018
Civil registration	Proportion of children under 5 years of age whose births were registered by the civil registration	79.43	71.36
Health	Proportion of births at health facility	85.97	86.43
	Proportion of births in health facility (B40)	87.25	81.37
	Proportion of births attended by skilled health personnel	97.72	99.49
Access to basic facilities and housing	Proportion of urban population living in adequate housing	50.33	56.97
	Proportion of population using safely managed sanitation	79.24	81.83
	Access to electricity (B40)	100	99.55
Access to ICT	Proportion of youth with ICT	69.92	80.14

Source: SDGs Bappenas [36].

indicates that the proportion of Cirebon District children under 5 years of age that owned birth certificates in 2018 was 71.36%, a decrease of 8.07% from the previous year. This decrease demonstrates that Cirebon District has not done a good job of protecting individual rights and ensuring that everyone has access to justice and social services by issuing adequate birth certificates. Hence, improvement in birth certificate coverage is urgent and has to be hastened to support stunting elimination in Cirebon District.

Next, the health facility and services factors, such as Birth at Health Facility, Birth in Health Facility (B40), and Birth by Skilled Health Personnel, are among health-related SDG indicators associated with stunting. These findings are as expected, since access to health facilities and services is essential to decrease the risk of child stunting by providing quality health services. This is in line with a previous study by Mtoi & Nyaruhucha [41] who reported that the place of birth is significantly associated with reductions in stunting, particularly when mothers delivered their babies at health facilities. Similarly, Ayelign & Zerfu [42] and Budhathoki et al. [43] found that children born at a health facility had a lower risk of stunting than those born at home. Aside from the birthplace, the findings are consistent with previous research that has found that the role and presence of skilled health personnel have a significant impact on stunting reduction. Mtoi & Nyaruhucha [41] also found that the kind of assistance the mother received during delivery was related to lower chances of malnourishment. A study by Torlesse et al. [44] in Indonesia reported that the odds of stunting in children 0–23 months old were more than double if a doctor or midwife did not provide antenatal care. These two studies concluded that having qualified health personnel on hand is vital. According to Potter et al. [45], health personnel have an essential role as educators. The involvement of health personnel as educators will be essential for preventing and reducing stunting. After all, adequate access to health facilities and services will allow mothers to receive all necessary treatment and information to protect their children from infections that can cause stunting.

As shown in Table 5, the proportion of births at health facilities (B40) increased by 0.46% between 2017 and 2018. With current growth rates, this SDG target will most likely be met gradually. Meanwhile, the proportion of births in Cirebon District health facilities has decreased from 87.25% in 2017 to 81.37% in 2018, representing a 5.88% decrease over a one-year period. Furthermore, the Cirebon District has done an excellent job of maintaining and increasing the proportion of births attended by skilled health personnel. Given the current achievement and importance of these indicators, it is imperative for Cirebon District to expedite progress in order to ensure that all children, youth, and women have equitable access to high-quality healthcare and services.

Furthermore, this study shows an interlinkage between stunting and basic facilities and housing, such as adequate housing, sanitation, and access to electricity. Several studies attempted to investigate the relationship between these factors and stunting, and their findings were consistent with ours. First, we will discuss how adequate housing can affect stunting reduction. Home is more than shelter; it is vital to the well-being of children and families. One of the many advantages that come with having access to adequate and affordable housing is improved health [46]. However, because poor households have a more limited ability to obtain adequate housing, children are at a higher risk of malnutrition [47,48,49]. A study of five health centers in Konawe District, Southeast Sulawesi Province, discovered that poor housing conditions, such as limited access to clean water, the absence of a sanitary latrine, and the use of poor housing roof material, all contributed significantly to underweight among children [50]. Adequate housing can also depend on the area people live in. Kumar et al. [51] found that in low- and middle-income countries, children living in urban areas are generally less likely to be stunted and underweight than children living in rural areas. This advantage is mainly due to better housing, greater food availability, electricity, piped water, sanitation, and transportation in urban areas [52].

For the case of Cirebon District, there is an increase in the proportion of the urban population living in adequate housing from 50.33% in 2017 to 56.97% in 2018, representing a 6.64% increase (Table 5). The current path shows that Cirebon District is making moderate progress toward ensuring access for all to adequate, safe, and affordable housing and upgrading slums, despite the fact that it remains low. Accordingly, in order to effectively eliminate stunting in the region, it is crucial to expedite efforts aimed at improving access to quality housing facilities.

Moreover, previous research has highlighted the importance of sanitation in lowering the risk of stunting. Danaei et al. [53] provided a consistent and comparable set of global estimates of the impact of risk factors on stunting. They discovered that poor sanitation is one of the leading risk factors for stunting around the world. Ikeda et al. [48] found that improvements in sanitation facilities made the largest contribution to the decrease in the prevalence of stunting. A previous study in Indonesia by Torlesse et al. [44] found that unimproved sanitation in households with untreated water was associated with more than twice the odds of stunting in their children. Similarly, a recent study by Mulyaningsih et al. [54] attempted to investigate the causes of childhood stunting in Indonesia. Their findings also stated that children who live in areas without access to water, sanitation, and hygiene are at a higher risk of stunting.

Regarding sanitation, Cirebon District was able to increase safely managed sanitation services by 81.83% in 2018 (Table 5), leaving 18.17% of the population with unimproved sanitation. It is imperative that Cirebon District must ensure improved sanitation access in order to contribute to the reduction in stunting prevalence and make significant progress by 2030.

Next, several previous studies in Bangladesh, India, and Vietnam show that basic housing needs, i.e., access to electricity in the household, could substantially lower the chance of stunting in children [55,56,57]. Numerous electronic appliances, such as refrigerators and electric ovens, facilitate the preparation and storage of food in the household and aid in the decrease in child malnutrition [55]. In addition, other electronic devices such as televisions and radios act as media information to address the nutritional requirements of children [55,58]. In fact, with the rapid evolution of technology, media-related information can now be accessed in real time through the internet. These findings are linked to prior studies that suggest that household internet access is associated with reduced stunting [59,60,61]. Internet-based information dissemination and awareness raising tend to increase the likelihood that mothers will provide adequate nutrition for their children, thereby reducing stunting [60].

In the case of Cirebon District, as shown in Table 5, universal access to electricity among poor and vulnerable households was reached in 2017. Given that access to electricity is a priority indicator for reducing stunting, it is important to maintain this achievement and prevent any declines as was the case in 2018. Moreover, a significant improvement is shown in the proportion of youth with ICT. In 2018, 80.14 percent of adolescents in the Cirebon District had access to ICT. This significant rise in the Youth with ICT indicator might be attributed to the rapid pace of technological advancement. This development can be utilized to improve performance on other SDG indicators, such as reducing the prevalence of stunting in the region.

6. Conclusions

This study employs an integrated methodology to investigate the interlinkages between stunting and other SDG indicators. The SDG interlinkages in this study are based on the economic complexity and network theory following El-Maghrabi et al. [13] and UNESCAP [17] that was first developed as a country development diagnostic framework. Applying this methodology to specific problems and regions can result in the identification of key targets, connections, and networks that can be utilized for SDG prioritization [62]. The multidimensional nature of the stunting issue necessitates an interlinkage analysis method between SDG indicators. Given the disparities in geographical conditions and the performance of each SDG indicator, the results obtained may differ from other studies because this study employs a case that is more focused on the district level.

Based on the findings, this study suggests that the effort to eliminate stunting in Cirebon District can be focused on four key aspects, i.e., civil registration, health facility and services, access to basic facilities and housing, and access to ICT. Primarily, priority can be given more to the poor households. It is worth noting that out of the eight priority indicators identified, seven have shown significant progress in the Cirebon District. However, access to adequate housing lags behind the other indicators. Therefore, the Cirebon District government needs to intensify its efforts to improve the achievement of these priority indicators. This is crucial as these indicators have a direct impact on other indicators, particularly in reducing the prevalence of stunting in the district. Furthermore, these findings provide additional evidence of the interconnectedness among SDG indicators, underscoring the multifaceted nature of stunting elimination.

The implications of this study suggest that the Cirebon District government can prioritize specific policy options to address the issue of stunting. These policy options are drawn based on the understanding that stunting is a complex issue with multiple dimensions and that the indicators within the SDGs are interconnected. These policy options include (1) ensuring birth certificate ownership to enable effective monitoring of child growth and development; (2) ensuring equitable access to health facilities and healthcare personnel to enhance healthcare services for all residents; (3) increasing access to clean water, proper sanitation, and effective wastewater management to ensure equitable health conditions for the population; (4) improving access to essential services in order to enhance the availability of adequate housing; and (5) investing in ICT education for the youth to foster their skills and knowledge in the digital age.

Moreover, the findings of this study can have implications for policymakers in terms of prioritizing their limited financial resources. It is crucial to have clear policy directives and adequate funding assistance to facilitate and promote practical linkages between SDG indicators, thereby fostering more integrated programming. By implementing the policy options identified in this study, the Cirebon District government could make significant progress in reducing stunting cases and, at the same time, achieving the sustainable development goals more effectively.

This study reported findings on how to apply an interaction approach to SDGs at the district level in Indonesia, where previous research had mostly focused on analyzing at the national level. Nevertheless, this research has some limitations. Due to data limitations and difficulties in finding data at the district level, this study could only cover 54 SDG indicators, introducing a possibility for bias. This study only conducted a correlation analysis of the SDG indicators. This means that this study can serve as a valuable resource for understanding the interconnections among the SDGs and can provide valuable insights to districts that have not been previously explored yet. Therefore, future research agendas are expected to enrich interlinkage research with more complete and high-quality data, followed by a more complete analysis that is not limited to correlations and also includes synergies and trade-offs to generate more impactful results.