This paper inves­ti­ga­tes the rela­ti­onship betwe­en pri­va­te and public invest­ments in Bra­zil. By esti­ma­ting a line­ar Vec­tor Auto­re­gres­si­ve Model (VAR) for 1996–2022, we find a crow­ding-in effect: an incre­a­se of 10% in public invest­ment gene­ra­tes a 1.48% and 2.5% rise in pri­va­te invest­ment after one and four quar­ters, res­pec­ti­vely. To inves­ti­ga­te the role of inco­me dis­tri­bu­ti­on in this result, we employ a Threshold Vec­tor Auto­re­gres­si­ve Model (TVAR). We find the crow­ding-in effect only occurs in the rela­ti­vely low-inco­me ine­qua­lity regi­me: a 10% incre­a­se in public invest­ment results in a 0.8% and 3.4% incre­a­se in pri­va­te invest­ment after one and four quar­ters, res­pec­ti­vely. Con­ver­sely, in the rela­ti­vely high-ine­qua­lity sce­na­rio, the res­pon­se is not sta­tis­ti­cally dif­fe­rent from zero. Thus, from a macro­e­co­no­mic stand­point, dimi­nishing ine­qua­lity holds the poten­ti­al to enhan­ce the res­pon­si­ve­ness of pri­va­te invest­ment to public investment.
Keywords: Public invest­ment. Pri­va­te invest­ment. Inco­me Ine­qua­lity. TVAR. Crow­ding-in effect.
JEL: C32, E22, E25, H54