In recent years the number of deliveries brought forward for non-medical reasons has increased in many countries. In the average of the OECD they increased from 15% in 1990 to 27.9% in 2015 (OECD 2011, 2017). In Spain, the rate of induced births stood at 19.4% in 2010, nearly double the rate recommended by the World Health Organisation, which situates it at 10% (Ministry of Health, 2012). The specialised literature alerts to the risks for infant health of bringing forward the birth date for non-medical reasons and the Spanish case offers a unique opportunity to examine this.

As an incentive to increase the number of births, in 2007 the Spanish government introduced a universal subsidy of €2,500 per new-born child, which was popularly known as the “baby cheque”. The subsidy was given in a one-off payment and was universal, so all women who were legally resident and who gave birth from that moment onwards received the bonus. Three years later (May 2010), and as a consequence of the economic recession, it was announced that the benefit was to be eliminated and would no longer be awarded after 31 December of that year. 

In other words, from July 2007 (when the measure was implemented) to 31 December 2010 (when it was eliminated), families received a €2,500 bonus for the birth of a baby, independently of the mother’s economic or employment status. However, if the babies were born on or after 1 January 2011, then no benefit would be received. This meant an important discontinuity took place at the turn of the year from 2010 to 2011. The change in the regulations became, unexpectedly, an incentive to bring forward the delivery date to December for those births expected to occur around the turn of the year. It stands, in short, as an interesting natural experiment for evaluating the effects of preterm birth on babies’ health.

Many mothers were mindful of the possibility of bringing forward their delivery date in order to qualify for the benefit, and some of the media reported this. For example, the newspaper ABC published a report with accounts from pregnant women whose estimated due date was around that time (Fominaya, 2010). Some of those expecting to give birth in January were absolutely against bringing their delivery date forward to December, due to the risks that might be involved, along the lines of the risks alerted to by medical professionals. However, other women felt it was “understandable” that, if their due date fell on one of the first few days in January, they agreed with their doctor to bring the delivery Planned early delivery of babies for non-medical reasons may have negative consequences for babies’ health 30 date forward a few days so that the baby would be born in December. The context of economic crisis faced by the pregnant women at that time should not be overlooked, as it meant the benefit was even more valuable to families.

At first glance it might seem difficult to know whether deliveries were brought forward to benefit from the €2,500 bonus and, if that were the case, how many deliveries were brought forward. It might seem even more difficult to know whether preterm deliveries could have affected the health of those children. However, it is possible to answer both questions based on the high-quality administrative data available on births and hospital admissions of all babies born in Spain between 2000 and 2012.

Specifically, there are two sources of data provided by Spain’s National Institute of Statistics (INE). One is its Childbirth Statistics, which is filled out at hospitals for every birth (date and time of birth, birth weight, weeks of gestation, delivery method and some sociodemographic characteristics of the parents) and the second is the Hospital Morbidity Survey, with information on 99% of all hospitalisations of at least one night (date and length of hospital stay, patient’s date of birth, gender, main diagnosis and province). To estimate the number of births brought forward, comparison was made between the number of births that took place around the turn of that year (end of December 2010 and start of January 2011) and those of previous and subsequent years as a reference. 

In the years prior and subsequent to the subsidy’s cancellation, the number of births in the last and first weeks of the year is very similar. In other words, under normal conditions approximately 50% of babies born close to the turn of the year are born in December and the other 50% in January. However, at the turn of the year from 2010 to 2011, some 56% were born in December (9,946 babies) and 44% in January (7,845), so this is clearly an atypical case. This means that during the last week of 2010 there was an upturn in the number of births followed by a downturn at the start of January 2011, which was when the baby cheque benefit disappeared.

In short, analyses suggest that effectively, the cancellation of the subsidy generated, unexpectedly, incentives to schedule a considerable number of deliveries early. Around two thousand families brought their baby’s delivery date forward from January to December as the result of the cancellation of the subsidy. These 2,000 births represent approximately 6% of all births in January. 

However, there was no marked increase in the number of caesarean sections during the last week of December. This suggests that deliveries that were brought forward and took place by caesarean section probably would have been by this method in any case. The change in the policy provided an incentive for families to bring forward deliveries already planned to be by caesarean, but did not lead to more caesarean sections being performed than would have been performed in any other context.

As for the profile of those who brought the delivery date forward and perhaps contrary to what one might think, the families who tended to a greater extent to bring delivery forward were not precisely those with the lowest socioeconomic levels. Rather to the contrary; advancing of the delivery date took place more frequently among parents with a university education, women of Spanish nationality and in provinces with a greater presence of the private hospitals network. Delivery was also brought forward more frequently among those who had given birth previously or those expecting multiple births. 

The medical literature has documented that a lower gestational age, i.e. preterm delivery, is associated with various medical problems post-partum in babies, especially respiratory problems. For this reason the natural experiment created by the change in the baby cheque benefit is a unique and very valuable opportunity to analyse the effects of preterm birth on two indicators: birth weight and health in the first months of life, measured by the number of hospitalisations of the babies. Specifically, the health results of babies born close to the 2010-11 turn of the year are compared with those born close to the turn of other years. It would be incorrect to compare children before and after the policy, as this would confuse the effects of the policy if only the delivery of the healthiest children were brought forward. In addition, October and November are used as control months to rule out the effect of other possible environmental factors.

With regard to birth weight, there is a certain positive tendency over the course of time, i.e. the average weight of the babies is slightly higher each year. However, the year of the policy change is clearly an extreme case, with an average birth weight 20 grams lower than the previous year. Perhaps 20 grams seems like a very small amount, but it is important to remember that this graph shows the average weight of all the babies born on those dates, both those born pre-term and the rest. 

Taking into account that the deliveries brought forward represented around 6% of the total births, additional analyses suggest that the babies affected by the cancellation of the subsidy were born weighing between 130 and 300 grams below the average, i.e. with a birth weight up to 9% lower. However, the proportion of babies with low or extremely low birth weights (below 2,500 grams or below 1,500 grams, respectively) did not increase, which suggests that the planned early deliveries fundamentally occurred with very advanced pregnancies. This is further confirmed by analyses of the weeks of gestation at delivery. To put it another way, it does not appear that such foolhardy acts were committed as inducing delivery several weeks before the due date, but rather that the early deliveries were fundamentally related to births with an estimated due date close to the turn of the year.

To analyse the health indicators after birth, hospital admissions of those born during the first 33 months of life are compared, i.e., up to the age of nearly three years. In this case, it does not seem that the children affected by the change in the baby cheque rules had more hospitalisations during the first week of life than the children born in other years. The rates are 0.138 for babies born around the turn of the year of 2010-11 and 0.143 for those born in other years. However, a considerable increase is observed in the hospitalisations of this group of babies when they were aged between 7 and 30 days old with respect to children born in other years. This increase is numbered at over 400 hospital admissions in comparison with other years, representing an increase of approximately 20%. For older children the differences are almost negligible.
 

In general terms, the most frequent diagnoses among the babies hospitalised are perinatal conditions, which follow immediately on from birth (representing nearly 34% of hospitalisations) and respiratory conditions (representing nearly 24%), followed by digestive problems or infectious diseases. When comparing the incidence of these conditions in babies born in different years, it is confirmed that during the first months of life the babies affected by the change in the baby cheque suffered 25% more hospitalisations due to respiratory problems (fundamentally bronchitis) than the children born in other years.

In summary, the data reflect that the babies affected by the cancellation of the baby cheque weighed less at birth (between 130 and 300 grams) and suffered a considerable increase (approximately 20%) in hospitalisation rates during the first two months of life, fundamentally due to respiratory illnesses. All the effects observed on health arising from a planned early delivery date cannot be attributed to other aspects that might confuse the relationship. In this sense, checks have been made to ensure that the results are not affected by a series of factors that could be intervening such as, for example, differing air quality or different hospital congestion levels. 
 

Therefore, it can be safely concluded that the worse health indicators of these babies (lower birth weight and higher number of hospitalisations) is the result of a lower gestational age at birth and, therefore, a lower degree of foetal maturity, as a consequence of a planned early delivery date. Furthermore, the results observed by taking advantage of this natural experiment of the baby cheque cancellation also have external validity because the profiles of those who had their delivery date brought forward in response to the baby cheque policy change (higher educational level, better socioeconomic position, with previous births, multiple births and/or private health insurance) are also those of people who under normal conditions would have scheduled the birth date. 

The main conclusion of this research is that planned early delivery for non-medical reasons has negative effects on the health of the babies, who are born with a lower birth weight and suffer a higher number of hospitalisations in their early months of life, especially due to respiratory conditions, although the negative effects seem to disappear from the second month onwards. It would be advisable, therefore, for families and healthcare agents alike to be aware of these effects when making decisions and that sufficient, reliable and quality information is available regarding the subject.

However, in addition, this study has at least two other policy implications. The first is that, given that individuals and families do respond to economic incentives, it would have been more appropriate for the cancellation of the baby cheque to have been less abrupt, instead taking the form, for example, of a progressive reduction in the subsidy over the course of several months. The second implication highlights the importance of administrative data for research purposes. Evaluating the effects of public policies helps to design better measures and programmes based on evidence and on their effectiveness.

BORRA MARCOS, C., L. GONZÁLEZ and A. SEVILLA (2017): "The impact of scheduling birth early on infant health", Journal of the European Economic Association. 

FOMINAYA, C. (2010): "Cheque bebé de alto riesgo", ABC, 30/12/2010 

MINISTERIO DE SANIDAD (2012): Informe sobre la atención al parto y nacimiento en el Sistema Nacional de Salud, Madrid: Ministerio de Sanidad, Servicios Sociales e Igualdad.

OECD (2011): "Caesarean sections", en Health at a glance 2011: OECD indicators, París: OECD Publishing.

OECD (2017): "Caesarean sections", en Health at a glance 2017: OECD indicators, París: OECD Publishin