Model Parametrik Batas Atas dan Bawah Tarif Tol di Indonesia Menggunakan Stochastic Frontier Analysis

Andreas Wibowo; Sudiro Roi  Santoso

doi:10.5614/jts.2022.29.1.9

Authors

Andreas Wibowo Universitas Katolik Parahyangan
Sudiro Roi Santoso Direktorat Jenderal Pembiayaan Infrastruktur Pekerjaan Umum dan Perumahan Rakyat, Kementerian Pekerjaan Umum dan Perumahan Rakyat

DOI:

https://doi.org/10.5614/jts.2022.29.1.9

Abstract

Abstrak

Menentukan tarif tol yang wajar untuk menarik pembiayaan swasta di sektor jalan tol merupakan isu yang penting. Namun, studi eksisting pada ranah ini masih belum tersedia. Tulisan ini bertujuan mengisi kekosongan pengetahuan tersebut dengan menawarkan model parametrik untuk mengestimasi batas atas dan bawah tarif tol menggunakan Stochastic Frontier Analysis. Data yang dikumpulkan meliputi biaya investasi, internal rate of return, tarif tol yang berlaku saat ini, durasi konsesi, tahun pertama operasi, dan panjang ruas jalan tol sampel. Dengan data set yang ada (n=50), skrining awal menggunakan Korelasi Pearson menghasilkan biaya investasi ternormalisasi menjadi satu-satunya parameter yang signifikan memengaruhi tarif tol. Statistik Mahalanobis-Distance menyimpulkan tidak ada kasus outlier multivariat. Piranti lunak STATA dimanfaatkan untuk membantu perhitungan mengestimasi parameter dan penyusunan model log-linier. Hasil uji Likelihood Ratio memperlihatkan hipotesis nol tidak dapat ditolak untuk model batas bawah, yang mengakibatkan model ini tereduksi menjadi Model Cobb-Douglas. Dua persamaan log-linier dibangun untuk batas atas dan batas rerata tarif tol. Kedua persamaan ini dapat dimanfaatkan sebagai acuan estimasi awal tarif tol sebelum kajian yang lebih detail tersedia. Tulisan ini juga mengidentifikasi beberapa keterbatasan yang dapat menjadi arah bagi studi berikutnya guna mendapatkan hasil estimasi yang lebih bermakna dan akurat.

Kata-kata kunci: Batas atas, batas bawah, tarif tol, stochastic frontier analysis, cobb-douglas, Indonesia

Abstract

Determining a fair toll rate is essential to attract private financing in the toll road sector. However, there is a paucity of existing studies in this area of investigation. Therefore, this paper aims to fill this knowledge gap by offering a parametric model to estimate upper and lower toll rate frontiers using the Stochastic Frontier Analysis. The collected data comprised investment cost, internal rate of return, prevailing toll rate, concession duration, the first year of operation, and length of the sampled toll roads. Given the dataset (n=50), a preliminary Pearson correlation screening suggested that the normalized investment cost be the sole parameter statistically affecting toll rates. The Mahalanobis-Distance statistic also concluded that no cases were considered multivariate outliers. The STATA statistical software package was used to aid in estimating parameters. The Likelihood Ratio test failed to reject the null hypothesis for the lower frontier model, thereby reducing it to the ordinary Cobb-Douglas Model. Two log-linear equations were developed for determining the upper and average levels of toll rates. These equations can be helpful to early toll road estimates before more detailed studies are available. This paper acknowledges the present study?s limitations, which can guide future studies to more meaningful and accurate estimates.

Keywords: Upper frontier, lower frontier, toll rates, stochastic frontier analysis, cobb-douglas, Indonesia

Author Biography

Andreas Wibowo, Universitas Katolik Parahyangan

Dr. -Ing. habil. Wibowo is the Chair of the Center of Excellence in Urban Infrastructure Development and a senior lecturer in construction management at the Department of Civil Engineering of Parahyangan Catholic University. He was the Head of the Secretariat for the Agency for Promoting Water Supply Systems (BPPSPAM) and a research professor at the Agency for Research and Development, Ministry of Public Works Indonesia. His specialty areas include public/private/partnership in infrastructure development; financial engineering; risk modeling; simulation and analysis; construction economics. Dr.-Ing. habil. Wibowo has authored and co-authored more than 100 journal and conference papers. He obtained his Habilitation degree (a post-doctoral degree) from Bauhaus Universitaet Weimar (2016), a doctoral degree from Technische Universitaet Berlin (2005), and a bachelor's and a master's degree from Institute of Technology Bandung (1995, 1998), all majoring in civil engineering. He won a number of renowned scholarships and awards, including the Alexander von Humboldt re-invitation postdoc fellowship (2018), Emerald Literati Award (2016), Alexander von Humboldt fellowship (2011-2014), DAAD doctoral (2001-2005), and re-invitation scholarships (2009), Australia-Indonesia Governance Research Partnership (2007), URGE World Bank scholarship (1996-1998), and best paper awards in several conferences.