Electronic Theses and Dissertation

Disparitas pendapatan rumah tangga menjadi perhatian karena dapat memicu permasalahan sosial dan ekonomi. Pendapatan rumah tangga seringkali digunakan sebagai indikator perbandingan tingkat kemiskinan pendapatan antar gender Kepala Rumah Tangga (KRT). Melalui pemanfaatan data panel Indonesian Family Life Survey (IFLS), penelitian ini bertujuan untuk mengevaluasi dampak perbedaan gender KRT terhadap pendapatan rumah tangga menggunakan metode Difference in Difference (DID) pada regresi data panel. Data terdiri 9.150 rumah tangga yang diamati dalam dua periode waktu, 2007 dan 2014. Sampel dibagi ke dalam dua kelompok, yaitu kelompok kontrol yang terdiri dari 8.287 rumah tangga dan kelompok “perlakuan” yang terdiri dari 863 rumah tangga yang kepala rumah tangganya berubah dari laki-laki di tahun 2007 ke perempuan di tahun 2014. Terdapat tiga model estimasi yang dapat digunakan untuk data panel, yaitu Common Effect Model, Fixed Effect Model, dan Random Effect Model. Model estimasi tersebut kemudian diimplementasikan dengan menggunakan DID sehingga dapat dilakukan evaluasi dampak perbedaan gender KRT terhadap pendapatan rumah tangga. Hasil penelitian menunjukkan perbedaan gender KRT berdampak pada penurunan pendapatan rumah tangga sebesar Rp8.600.000 jika menggunakan Fixed Effect Model dan sebesar Rp7.100.000 jika menggunakan Random Effect Model. Variabel-variabel yang berpengaruh terhadap pendapatan rumah tangga adalah jumlah anggota rumah tangga yang berusia 15–64 tahun, jumlah pekerja dalam rumah tangga, penghasilan dari selain bekerja, penghasilan nontani, total pengeluaran konsumsi, sisa pinjaman rumah tangga, status kepemilikan rumah, kepemilikan toilet, penggunaan listrik, pembelian air untuk minum, usaha nontani, lama pendidikan KRT, status daerah tempat tinggal, bahan bakar utama memasak, dan usaha tani. Nilai koefisien determinasi pada Random Effect Model adalah 0,35, lebih tinggi dibandingkan Fixed Effect Model pada 0,30. Karena itu diduga Random Effect Model lebih baik dalam mengestimasi dampak perbedaan gender kepala rumah tangga terhadap pendapatan rumah tangga.

Disparities in household income are a concern because they can trigger social and economic problems. Household income is often used as an indicator of comparing the level of income poverty between heads of household (KRT) gender. Through the use of Indonesian Family Life Survey (IFLS) panel data, this research aims to evaluate the impact of gender differences in household heads on household income using the Difference in Difference (DID) method in panel data regression. The data consists of 9,150 households observed in two time periods, 2007 and 2014. The sample was divided into two groups, namely the control group consisting of 8,287 households and the "treatment" group consisting of 863 households whose head of household changed from male in 2007 to female in 2014. There are three estimation models that can be used for panel data, namely the Common Effect Model, Fixed Effect Model, and Random Effect Model. The estimation model was then implemented using DID so that the impact of gender differences in household heads on household income could be evaluated. The research results show that gender differences in household heads have an impact on reducing household income on average by IDR 8,600,000 if using the Fixed Effect Model and IDR 7,100,000 if using the Random Effect Model. The variables that influence household income are the number of household members aged 15–64 years, the number of workers in the household, income from other than work, non-farm income, total consumption expenditure, remaining household loans, home ownership status, toilet ownership, electricity use, purchasing water for drinking, non-farming businesses, length of education of household heads, status of area of residence, main fuel for cooking, and farming businesses. The coefficient of determination value in the Random Effect Model is 0.35, higher than the Fixed Effect Model at 0.30. Therefore, it is suspected that the Random Effect Model is better at estimating the impact of gender differences in household heads on household income.