Domestc Savngs and Internatonal Captal Flows Martn Feldsten and Charles Horoka The Economc Journal, June 1980 Presented by Mchael Mbate and Chrstoph Schnke
Introducton The 2 Vews of Internatonal Captal moblty - amed to show the relatonshp between S and I, Determnants of Optmal savng Polcy In order to know the optmal rate of savng and the ncdence of tax changes we consder how moble the supply of captal s, whether captal flows to equalze yelds to nvestors and whether savngs that orgnate from a country reman to be nvested there. 1. Closed economy -the natonal returns on addtonal savngs s the domestc margnal product of captal ncreasng savng rates mples postponng consumpton -naton receves both after tax yeld and the tax revenue -What should nfluence t he natonal savng polcy s thus the pretax margnal product of captal.
Introducton 2. Open Economy -captal s moble and addtonal savngs wll leave the country(captal exporter) -yeld to the home country on the addtonal savngs s only the net of tax return receved by the nvestors and not the pre tax margnal product of captal. After-tax real yelds are about 50% of the pretax yeld= optmal savng polcy s lkely to depend on whether the economy s closed or open(perfect world captal markets) Analyss of Tax Incdence wth respect to captal moblty - A closed economy has fxed captal stock -tax on ncome of captal n producton s borne by owners of captal -Captal moblty mples burden shftng to labor and foregn captal owners. corporate ncome tax borne less by captal -Labor ncome tax shftng of domestc owners of captal s dffcult
Introducton Does captal flow to equate net of taxes? possble for short term securtes but not long term portfolo captal or drect nvestment because for long term nvestments-(unobserved), but skeptcal to equate net of taxes because Rsks and uncertantes of foregn markets Offcal restrctons on captal export and future captal export control Fear of changes n tax regmes n foregn countres Insttutonal rgdtes n domestc country
Introducton Captal does not maxmze nvestors net of tax return. there exst nternatonal dfferences n tax rules and nteracton of foregn and domestc taxaton net return maxmzaton s dfferent f taxes dd not exst. Wthout taxes gross returns wll be the same n all countres and nvestors ndfference on where to place funds. Wth taxes nvestors should specalze ther nvestment n a country dfferent from ther own. Snce ths specalzaton s absent, then there exst restrctons on captal movement to maxmze each nvestors net return. Drect nvestment n foregn market s assocated wth mplementng strateges, explotng producton knowledge and overcomng trade restrctons and not pursut for proft reason why countres are both exporters and mporters of captal.
The emprcal approach Data from 16 OECD countres (n the baselne specfcaton) 1960 1974, yearly observed Gross domestc savng and nvestment rates Both on average around 25% of GDP, standard devaton 0.045. Relatvely stable over tme Use averages over 5 and 15 years horzon Basc equaton:! I $ # &! =! + " S $ # &
The emprcal approach (cont d)
The emprcal approach (cont d) Result: reject hypothess of perfect world captal moblty ncremental savng remans n the domestc country Net specfcaton: measurement error s lkely coeffcents are based towards unty. Consstent estmate can be obtaned by IV approach, usng gross savngs rate as nstrument results reman unchanged
The emprcal approach (cont d) Guess: there s a common mpact of a thrd varable on savngs and nvestments: populaton growth (lfe cycle theory of savngs). But the results do not change when growth s ncluded n the model. No further checks on ths ssue. Another robustness check mplemented on the degree of openness:! I $! # & =! + (" 0 + " 1 X ) S $ # & Results stable
The emprcal approach (cont d) Dsaggregaton nto household, corporate and government! I $ # &! SH $ =! + " H # &! SC $ + " C # & No sgnfcant dfference between the contrbuton of dfferent sectors to total nvestment. Corporate nvestment s most responsve to corporate savng ( nsttutonal rgdtes or portfolo preferences) No evdence of coeffcents sgnfcantly dfferent from unty! SG $ + " G # &
The emprcal approach (cont d)
Crtcs and conclusons Sample sze too small for testng real tme seres approach would be useful Conclusons true nature of world captal markets. approprate to study ncome dstrbuton and tax ncdence excludng captal moblty natonal returns on domestc savng s equal to pretax domestc margnal product of captal