Modeling urban house-rent variation in Bangladesh: a study of four metropolitan cities

Abstract

Regional variallon is an obvio\IS prevalence in almost e~ery sector in Bangladc,h. The phenomenon is one-slep ahead kadmg 10 lnwr-societnl variation in lhe housing seclor. There have heen limiled ,lmhcs (as wdl as inadequate .,cope for ,tudy) in this field m Bangladesh. How~ver su~h analyses are crucial for'devising any lypes of rcgulaling mechani,m and policy [ormulahon, WIth this view in mind, this study focus~, on the regional pattern of urban house rents ill Bangladesh, The Sllldy analY7,es the t,ends of house rent vanahon in last 15 year" identifLe.' the pattern of such changes across the cities and develops a foreca'ling modd 10 explain lhe probability of a year-on-year renl increase at different rate, across the region,. The sUldy is confined with rent analyses of four metropolitan cilies of Bangladesh. Dhaka. Chittngong, Rajshahi and KhlLlna.Thc dwcllings arc primarily divided into high, medium and low income categorie,. House-renl, of alllypcs of dwellings are lhe highest in Dhaka metropolitan area and lowesl in Khulna, although the highest growth can be ohserved here. In ca>e of l"w"incomc dwellings highest rent ean be ohserved in Dhaka metropolitan city. rhe lowc,l rcnl of high-income dwdhngs ean be found in Rajshahi metropolitan city. Rents of middle-income dwellings arC hIgher in c\'ery eity having the highesl value in Dhaka. Another imporlant finding is that although hou.,e-rents are higher in the two bigger citi", of Dhaka aud Chittagong, the growth pattern of these two cities are more or less 'leady. In COnlIa"l, lhc lower house-rent areas of Khulna and Rajshahi pose an incremental b'[(}wlnpallern. nH,ch higher than Dhaka and Chittagong for all type, of d,,'elliug'. 'J he model e'l\makS >uggest that .type of structul'e' is the mmt influential variable in determining the rent iucrease patlem follow~d by lhe 'location' parameter. The model also sugge;;ts that the probabilily of a year-on-year rent-merease is higher for lower, mLddleand low-income d,,'ellings at nigher rale, (i.e, allhe ralc of 6% - 15%),