Abstract
Considering the poor socioeconomic status of the Spanish surnamed population in Texas, the Spanish surname neonatal mortality rate seems surprisingly low. During five of the last 10 years, the neonatal mortality rate for Texas Spanish surnamed population was actually lower than for Anglos. It is often suggested that the low Spanish surname neonatal mortality rate is due in part to reporting problems peculiar to the Spanish surname population in Texas. Linked birth and neonatal death records of Texas residents are examined for evidence of underreporting or misreporting of Spanish surnamed neonatal deaths. It is found that discrepancies in coding race on the birth and death records cause a minor deflation of Spanish surname neonatal and infant mortality rates. Indirect evidence indicates that there may be a substantial amount of underreporting of Spanish surname neonatal deaths. This underreporting appears to be associated with the presence of a large number of Mexican nationals misidentified as Texas residents, and the greater reliance upon lay midwives by the Spanish surname population in Texas. Reasons proposed to explain a reluctance to report neonatal deaths include fear of contact with authorities, and fear that a reported death would diminish the value of the birth certificate.