Pengukuran kemiskinan di Indonesia masih terfokus pada pendekatan moneter yaitu mengukur kemiskinan dengan nilai uang sedangkan menurut Sen Poverty is Capability Deprivation, ?Deprivation Intrinsically Important but Income Poverty Instrumentally Significant" yaitu kemiskinan moneter lebih mudah diukur dari pada pendekatan deprivation. Deprivation dapat diartikan sebagai keadaan tidak berpunya pada salah satu indikator kesejahteraan (well being) yang dapat diobservasi tidak tergantung kepada pendapatan, dan dialami oleh mereka yang miskin. Pendekatan ini lebih baik untuk melihat kemiskinan anak karena akan melihat kemiskinan berdasarkan indikator kebutuhan tumbuh kembang anak.
Dengan mempergunakan data gabungan Susenas 2008, Sakernas 2008 dan Podes SP 2010 dengan analisis multilevel logistik biner akan melihat pengaruh karakteristik individu, rumahtangga dan wilayah terhadap peluang timbulnya kemiskinan anak. Hasil menunjukan bahwa faktor latar belakang rumahtangga cukup dominan terhadap peluang munculnya kondisi anak miskin yaitu jenis kelamin KRT, pendidikan KRT, status bekerja KRT, status kemiskinan rumahtangga, jumlah anggota rumahtangga dan tempat tinggal. Jenis kelamin anak dan umur juga memberikan perbedaan terhadap peluang munculnya peluang kemiskinan anak. Faktor tingkat pengangguran, fasilitas kesehatan dan tingkat pendidikan kepala desa juga berpengaruh.
The measurement of poverty in Indonesia is still focused on the monetary approach that measures poverty by monetary standards. According to Sen, poverty is capability deprivation, ?Deprivation Intrinsically Important but Income Poverty Instrumentally Significant?. Poverty by monetary approach is easier to measure then the deprivation approach. Deprivation can also mean lack in several observable variables in the wellbeing indicator that is does not depend on income but is definitely experienced by the poor. This approach is appropriate to seek child poverty because it is based on deprivation of needs that are important for child to grow.
This study utilize a combination of Susenas 2008, Podes SP 2010 and Sakernas 2008 data using binary multilevel logistic analysis to study deprivation by their individual characteristics, households characteristics and regions on the probability incidence child poverty. Results shows that household background that dominates the probability of the child to be poor is the head of household's sex, head of household's education, head of household's working status, household's poverty status, number of household members and household?s place. The sex and age of the child also differentiates the probability incidence of child deprivation. Level of unemployment, health facility ratio and education of local authority officials also has effects in probability of child deprivation incidence.
