The impacts of covid-19 on Indonesian poverty and unemployment

This study estimates the negative impacts of the COVID-19 on poverty and unemployment in Indonesia. In doing so, this study develops and uses the SD model that reproduced similar patterns in terms of GDP, poverty, and unemployment. Estimated unemployment and estimated poverty under the COVID-19 are calculated for three different economic scenarios: the optimistic, the mid, and the pessimistic scenarios. This study concludes that Indonesia will experience rising unemployment and poverty in ranges (9-10) and (25-26) million people respectively by the end of this year – depending on projected economy growths. This study suggests that existing financial aids are sufficient to support rising unemployment and rising poverty level. However, if the Indonesia government cannot carefully slow the COVID-19 flow, higher financial supports are required to curb the negative impacts of the COVID-19.


Introduction
The COVID-19 has affected almost all countries in all continents across the world. Despite low fatality rates, the negative impacts have reached large-scale socioeconomic impacts such as low employment and possible economic recessions. In reality, regardless of country income, all countries have struggled to escape from the COVID-19 impacts. Recently, despite lower infection rates upon preparedness of more open social and economic activities, the possible second waves have shadowed policymakers. Although some countries have slowed the COVID-19 flow, those affected countries have struggled to curb the negative impacts of the COVID-19.
As one of the emerging markets, Indonesia has attempted to properly handle the pandemic through some policies such as the social restriction policy, restricted travels, and relaxed policies relate to taxes and consumer credits. However, Indonesia has still struggled to sustain its positive economic growths.
Statistics show that the COVID-19 cases were firstly identified on March 1 st , 2020 after two female Indonesians returned from the International dance festival (Wijaya, 2020; https://covid19.go.id). Following the first two identified cases, the Indonesian COVID-19 cases have increased rapidly, despite the social distancing policy. Up to date, there are about 50,000 cases, 2,000 recoveries, and 3,000 fatalities due to the COVID-19 (https://covid19.go.id).
Despite beneficial findings of those studies, there is no available study in investigating the impacts of the COVID-19 on employment, and poverty. Some studies noticed the impacts of the pandemic on unemployment and economic growth (Boissay et al., 2020;Caracciolo et al., 2020;Fornaro, & Wolf, 2020;Gumede et al., 2020;Loayza, & Pennings, 2020;McKibbin, & Fernando, 2020). Even, the IMF itself has offered financial support to affected countries in minimizing the negative impacts (https://www.imf.org/en/Topics/imf-and-covid19/Policy-Responses-to- . The IMF has also provided a link -explaining possible policies tackling the negative impacts in affected countries (https://www.imf.org/en/Topics/imf-and-covid19/COVID-Lending-Tracker#APD). The aforementioned studies enhance our awareness of understanding the negative impacts of the COVID-19 on socioeconomic factors, especially on employment and poverty. This study aims to fill this gap, focusing the negative impacts of the COVID-19 on unemployment and poverty in Indonesia.

Data and Methods
Socioeconomic data were collected from the Indonesian statistics bureau (www.bps.go.id). For instance, Gross Domestic Product (GDP), poverty, and employment were collected in the period 2000-2019. This study uses the 2000 constant price's GDPs which are also available in the BPS website. The poverty definition used in this study follows the BPS definition, stating that disadvantaged people who cannot fulfill their daily calorie intake about 2,100 per daily capita (https://www.bps.go.id/subject/23/kemiskinan-dan-ketimpangan.html).
To estimate the impacts of the COVID-19 on poverty and unemployment, this study assumes that the economic downturn starts from reduced consumption. This study takes this way as Indonesia GDP has been mainly dominated by consumption about 61% of the total GDP (BPS, 2018;Soemartini, 2007;Wiranthi, 2014). So when the Indonesian economy is under pressure, consumption is expected to experience a significant loss (Suryahadi et al., 2020).
The application of the system dynamics (SD) approach has been applied in some subjects such as macro economy and health issues (Sterman, 2001). Furthermore, Sterman (1991) affirmed that one benefit of the SD approach is enabling policymakers to solve critically complex issues through the system-thinking concept. The concept captures important features such as stock-flow, time delays, non-linearity, and feedbackwhich are also underlying features of the real world. Owing to this, this study applies the SD approach in understanding the impacts of the COVID-19 on poverty and unemployment in Indonesia.

The Indonesia economy
Indonesia has experienced significant economic growths since the Asian monetary crisis of 1998. Since an economic contraction about 13% in 1998, Indonesia has experienced promising economic growths about 5-7 % per year as seen in figure 1. Consumption has dominated Indonesian Gross Domestic Product (GDP) about 60% of the total GDP (Bahri, 2008, BPS, 2018. Owing to dominance of household consumption, Indonesia has still led a positive economic growth during the economic crisis of 2008. Some studies (Suryahadi et al., 2006;Van Leeuwen, & Földvári, 2016;Yusuf, & Sumner, 2017) confirmed the positive impacts of Indonesian economic growth on poverty reduction. It was found that every 1 percent economic growth has led to 0.5 percent of poverty reduction in Indonesia. As seen in figure 1, the number of poor has reduced subject to positive economic growths.

System Dynamics Model
The system dynamics model was derived from other studies (Bahri, 2008;Forrester, 1985). In understanding the Indonesian macroeconomic model, the SD model is separated into three sub-models as seen in figures 2-4. The first sub-model is the potential output (PTY) and output (Y or GDP) sub-model that determines relationships between PTY and GDP. Please note that the model documentation is available in appendix A. The second sub-model is called the capital and labor (CL) sub-model that describes relationships between population, labor, and capital. In this sub-model, a relationship between GDP, poverty, and employment is defined based on historical data. The last sub-model is called the Aggregate Demand (AG) sub-model, explaining relationships between GDP, AG, and poverty level.   The model performance in estimating observed GDP, observed poverty, and observed employment can be seen in figures 5-7. Figures 5-7 show that the SD model can reproduce similar observed GDP and observed poverty. For unemployment, the SD model can reproduce similar unemployment patterns in the last decade (2010-2019). As seen in table 1, MAPEs of the SD model are less than 20% which mean the SD model can predict the observed system relatively accurate (Hanke et al., 2001;Hoshmand, 2009

Estimated unemployment and poverty level
To estimate the poverty level and unemployment during the COVID-19, this study evaluates three different scenarios. The first scenario is the optimistic scenario where Indonesian economic growth will be about 2.5% by the end of this year (Wardhana et al., 2020). The second one is the mid scenario where Indonesia will experience an economic growth of about 0.5% (Wardhana et al., 2020). The last one is the pessimistic scenario in which Indonesia will experience a negative GDP growth of about -3.5% (Wardhana et al., 2020). To simulate each scenario, the SD model is embedded by a variable namely "suppress consumption factor" as seen in figure 4. This variable represents a consumption drop in 2020. Consumption is projected to decrease about 11.5%, 19%, and 33.5%. Reduced consumptions correspond with the economic growth of the first, the second, and the third scenario respectively.

No
As seen in figure 9, the number of poverty and unemployment tends to increase for the first, the second, and the third scenario. As expected, the number of unemployed and poverty are the highest ones in the third scenario. Figure 9. Comparison outputs for different scenarios by 2020 Figure 9 displays the estimated unemployment and poverty for each scenario by 2100. It is estimated that unemployment will be about 6.76% to 7.89% for the first to the third scenarios by 2100. This corresponds with 8.9 million and 10.3 million unemployed people respectively. This study also estimated rising poverty about 8.79% to 9.46% depending on the scenarios by 2100. This means that rising poverty will be about 23.75 million to 25.56 million by 2020. Figure 10. Estimated unemployment and poverty by 2020 The Indonesia government has provided support to rising poverty and rising unemployment under the pandemic. The proposed funding is about IDR 677.2 trillion that is about IDR 87.55 trillion to support medical facilities, and IDR 385.75 trillion to relief industriesplease see table 4 for details. For the safety net, the government provides about IDR 203.9 trillion. Table  2 shows that the potential poverty level is about 25.56 million people or 5.6 million households assuming a family size is about 4.58 people (BPS, 2019). This means that the social safety net can support affected families about IDR 700,000.monthly for three months in the worstcase scenario. Figure 11. The government support for the Covid-19 (Kementerian Keuangan RI, 2020) The first scenario The second scenario The third scenario Estimated unemployment and poverty by 2020 Poverty (2020) Unemployment (2020) 123 Funding support in health, economy, and social safety nets (IDR trillion) Small enterprises Corporate/state owned companies Tax relief Sectoral aids Health Social safety net The Indonesian poverty line is about IDR 440,500/person/month or about 2 million/family/month (BPS, 2019). Assuming that a consumption drop is similar to an income loss (33.3%), the safety net of IDR 700,000 per family could relatively curb the income loss during the COVID-19 time.
As this support is only available for about 3 months, this study also suggests that the government should consider and monitor the impacts of the COVID-19 regularly to anticipate rising unemployment and rising poverty. The most importantly, the government should work together with other stakeholders to hamper the COVID-19 spread across the country.

Conclusion
This study applied the SD approach to estimate the impacts of the COVID-19 on Indonesian unemployment and poverty. The SD model can reproduce similar GPD, unemployment, and poverty in the period 2000-2019. As such, the model can be used to forecast projected unemployment and poverty owing to the COVID-19.
This study also suggests that available funding for the safety net is sufficient to support affected households for about three months. In that case, the government should monitor and handle the COVID-19 flow properly so that the society and the industries can return to their normal activities at the expected time. If the impacts of the COVID-19 is longer than expectation, the government should prepare higher financial aids to support rising unemployment and poverty level.    emplyment/tae*adjustment factor2,(labour supply/expected labour supply *((delayed unemployment-relationships between Y output and employment)*labour supply/tae) )) ,MIN(DE desired emplyment/tae*adjustment factor1,(labour supply/expected labour supply*adjustment factor1*((delayed unemployment-relationships between Y output and employment *adjustment factor1)*labour supply/tae))) ))Description: changes in employment Present in 1 view: PTY potential output (IDR/Year) = Yo output initial*((Lw labour/Lo labour initial)^(1-"alpha (capital intensive index)"))*((Kw kapital/(Ko capital initial))^("alpha (capital intensive index)")) Description: the potential output Present in 1 view: