2. Potential mechanisms for the effect of sirens on birth outcomes

Numerous studies have demonstrated the importance of intrauterine shocks to mothers in determining birth outcomes. Many of the shocks investigated in these studies alter the physical environment pregnant women are exposed to in terms of environmental factors such as air quality [Currie and Walker (Reference Currie and Walker2011); Currie et al. (Reference Currie, Greenstone and Moretti2011); Currie and Schwandt (Reference Currie and Schwandt2016)], nutritional quality [Lumey et al. (Reference Lumey, Ravelli, Wiessing, Koppe, Treffers and Stein1993); Almond and Mazumder (Reference Almond and Mazumder2011)], or the availability of prenatal care [Evans and Lien (Reference Evans and Lien2005)]. Other shocks relate to the mother's psychological condition while pregnant, which has been largely investigated in terms of maternal stress. Maternal stress has been shown to increase the production of cortisol, which is related to premature delivery [Lockwood (Reference Lockwood1999); Hobel (Reference Hobel2004)]. Several studies consider natural disasters [Torche (Reference Torche2011); Currie and Rossin-Slater (Reference Currie and Rossin-Slater2011)], bereavement [Persson and Rossin-Slater (Reference Persson and Rossin-Slater2018)], or increases in discrimination and racism [Lauderdale (Reference Lauderdale2006)] to show a causal detrimental relationship between stress experienced by mothers during various phases of their pregnancies and birth outcomes several months later.

Civilian stress can also increase due to armed conflict or violent incidents. In accordance with this, numerous studies have found adverse effects in response to stress induced from armed conflict during the early stages of pregnancy. Mansour and Rees (Reference Mansour and Rees2012) show that mothers in the Palestinian territories who resided in a town with higher fatalities by Israeli security forces during the Second Intifada gave birth 6–9 months later to infants with a low birth weight in higher probabilities. Camacho (Reference Camacho2008) investigates land mine explosions in Colombia as a source of stress during pregnancy and finds a detrimental effect on birth weight as a function of the number of land mine explosions during early pregnancy. Quintana-Domeque and Ródenas-Serrano (Reference Quintana-Domeque and Ródenas-Serrano2017) investigate the effect of terror attacks in Spain over a period of more than 20 years and find that terror bomb casualties during the first trimester resulted in lower birth weight. Torche and Shwed (Reference Torche and Shwed2015) exploit Israel's second Lebanon War in 2006, which similarly affected residents of northern Israel with rocket attacks for several weeks as OPE did for southern Israel, and find that exposure to the conflict early in the pregnancy had negative effects on birth outcomes. The authors do not examine localities separately and do not utilize data on sirens, as we do in the present study. Gluck et al. (Reference Gluck, Grin, Mizrachi, Leytes, Namazov, Anteby, Bar and Kovo2019) examine maternal stress and birth outcomes in response to OPE as well. However, their study focuses on comparing birth outcomes between mothers residing in close proximity to Gaza who experienced prolonged exposure to sirens prior to OPE and mothers residing in central Israel for whom the exposure to sirens during OPE was a new phenomenon. Their entire sample consists of births that were all exposed to OPE across different trimesters so it is not possible to conclude from the paper regarding the broad effects of OPE, in comparison to pregnancies that did not experience OPE. They find that OPE adversely affected birth outcomes when experienced during the first and second trimester, in comparison to third trimester exposure, and that the effect depended on whether prior to OPE the mother had already experienced sirens. Brown (Reference Brown2018) finds that exposure to greater violent conflict due to drug wars in Mexico during early pregnancy decreased birth weight. Lastly, Eskenazi et al. (Reference Eskenazi, Marks, Catalano, Bruckner and Toniolo2007) exploit the stress induced by the September 11 attacks in New York and show increased probability of low birth weight among mothers experiencing the event early in their pregnancy.

While armed conflict or violence can affect birth outcomes through the anxiety and stress they may generate among pregnant women, behavioral responses among pregnant women during armed conflict can also affect birth outcomes. OPE was a conflict lasting less than 2 months, so migration or fertility responses are not relevant for this setting. Nevertheless, we do test whether our results may be driven by migration patterns shortly after OPE and find no evidence for this (see Table B1 in online Appendix).Footnote ³ However, changes in prenatal care can be a likely behavioral response to OPE. On the one hand, OPE increased the risk of performing outdoor activities, including routine activities such as medical check-ups and prenatal care. This could have led to a decrease in prenatal care during OPE. On the other hand, it has been shown that pregnant women can increase pro-health behavior—particularly prenatal care—in response to violent conflict, so as to better shield their fetus from the negative shock. Torche and Villarreal (Reference Torche and Villarreal2014) find improved birth weight outcomes in response to greater exposure to homicides due to drug wars in Mexico. Whether prenatal care decreases or increases in response to OPE may depend on pregnant women's ability to limit or minimize the risk of obtaining prenatal care.

A few studies on armed conflict during pregnancy exploit regional and temporal variation in the intensity measures of armed conflict, such as casualties [Mansour and Rees (Reference Mansour and Rees2012); Quintana-Domeque and Ródenas-Serrano (Reference Quintana-Domeque and Ródenas-Serrano2017)], explosions [Camacho (Reference Camacho2008)], or homicide rates [Brown (Reference Brown2018)]. Our study also exploits such variation by assigning to each mother the number of sirens she experienced in her locality of residence during each trimester of her pregnancy.

Environmental factors were not affected in any way by OPE. The combat fighting was confined to the Gaza strip, where Israelis do not reside, and missiles shot at Israel do not produce any environmental pollutants or chemical hazards. While OPE did affect the economy in Israel, this effect was marginal for the vast majority of households with an estimated loss of 0.4% of GDP.

Absences from work increased during OPE, as child care services were not available in many localities and some employees may have had concerns on traveling to reach their work destinations.Footnote ⁴ For pregnant women, this could lead to a substantial decrease in work attendance, as paid parental leave in Israel does not include the prenatal period.Footnote ⁵ However, results on the effects of work attendance on birth outcomes are not conclusive, with evidence that both adverse [Wüst (Reference Wüst2015)], beneficial [Vrijkotte et al. (Reference Vrijkotte, Van Der Wal, Van Eijsden and Bonsel2009); Dave and Yang (Reference Dave and Yang2020)], or null [Ahammer et al. (Reference Ahammer, Halla and Schneeweis2020)] birth outcome effects are associated with decreased work attendance. Thus, although we indeed see increased absence from work using data from the Israel Labor Survey (see Figure C1 in online Appendix), we do not reach any conclusions related to this potential mechanism in our main analysis, given the inconclusiveness regarding the effect this may have on birth outcomes.Footnote ⁶

3. Data

Data for our analysis are primarily from two sources. The first source is restricted confidential data for all births delivered in Soroka University Medical Center (Soroka) between January 2007 and early August 2015. The second source documents all sirens in Israel by locality and date beginning in June 2014 and ending in September 2014.

Soroka University Medical Center (Soroka) is the only hospital serving the Negev region in southern Israel. The hospital is located in Beer Sheva, the largest city in southern Israel, and provides medical services to approximately one million residents of the South, from Kiryat Gat and Ashkelon to Eilat. Soroka is the third largest hospital in Israel, with 1,151 hospital beds. The nearest hospital to Soroka is roughly an hour drive from Soroka. Thus, for the vast majority of mothers giving birth in the Negev region, Soroka was the only option for delivery.Footnote ⁷ The number of annual births in Soroka ranged from over 12,000 in 2007 to roughly 15,000 in 2014, steadily increasing over the years. Of these, roughly 55% of births are to Arab mothers and 45% are to Jewish mothers. We focus our analysis on Jewish births, due to concern that the locality for some of the Arab mothers is imprecise, as a large fraction of Arabs in southern Israel reside in unrecognized communities that are not documented in the Soroka birth data. During OPE, Soroka was fully functional. The hospital complex has built-in protective shelters in nearly every hallway in case of a missile attack. All labor and delivery rooms (as well as operation rooms) in Soroka are fully protective in case of a missile attack.

Our birth data detail for each birth the month of birth, the month of release from the hospital, mother's age, birth parity, locality, whether she is Jewish or Arab, and a vast range of pre- and post-birth health conditions and outcomes documented by ICD codes. Mother identification numbers are provided, so that it is possible to identify multiple births for the same mother over time.

Due to data confidentiality, only the month of birth and release from the hospital are provided, rather than exact dates. Our treatment variables—the number of sirens experienced by the mother during her pregnancy—are relative to the conception date. In order to derive an approximate date of conception, we first assume that all birthdates are on the 15th of the birth month provided, except in cases where the release month from the hospital proceeds the birth month, which for the most part should reflect births at the end of the month with a release date in the following month. Thus, for those births we assume a birthdate at the end of the birth month provided. This should generate errors of at most two weeks in the exact date of birth.Footnote ⁸ For the conception date, we take the birthdate we derive, subtract from it the gestation length provided in days, and add to that 14 days.

Our main sample is limited to Jewish mothers ages 20–45 with a birth parity of 10 or less and gestation length of at least 180 days. We exclude mothers from localities that are within 7 kilometers of the Gaza strip. These towns and communities—officially termed the Gaza Envelope in Israel—experienced rocket and mortar firings for several years prior to OPE and their residents were also more likely to leave during OPE.

Most of our analysis focuses on the sample of births conceived during 2011, 2013, and the first half of 2014. We avoid potential selection of births that were conceived during and immediately after OPE, by excluding from our sample all births with a conception date beginning July 1, 2014. We also exclude births conceived during 2012 due to an eight-day military operation during November 2012, which also involved rocket attacks. We thus have two conception years—2013 and 2014 with varying exposure to OPE and sirens during the pregnancy trimesters, along with 2011 conceptions, which complement 2013 control group conceptions, such that the sample includes conceptions that were not exposed to OPE during pregnancy from all calendar months.Footnote ⁹ When using samples from longer time frames for our analysis, we exclude pregnancies exposed to the other military operation in Gaza preceding OPE—Operation Cast Iron in December–January 2008.

Our outcome variables consist of the following post-natal maternal and newborn health conditions: birth weight in grams along with an indicator for low birth weight (LBW) (≤ 2500 grams), and gestation length in days along with an indicator for pre-term delivery (<37 weeks gestation length).Footnote ¹⁰ We utilize the health conditions available in the Soroka birth data to control for the following pre-birth conditions in our regression analysis: maternal age (categorical: 20–26, 26–30, 30–35, 35–40, 40–45), child sex, maternal abortion history, birth parity (categorical), high fertility, past cesarean delivery, late pregnancy loss in the past, and infertility treatment.Footnote ¹¹

Our birth data are merged with data on all sirens by date and locality during OPE and the month preceding OPE. Israel's Home Front Command—an IDF branch responsible for instructing the civilian population on coping with military threats facing Israel—has on its web site detailed documentation of all sirens by date and locality beginning in mid-July 2014. Because we needed data on sirens from as early as June 2014, we complemented the Home Front Command data with documentation of all sirens by date and locality from a provider of a cell phone application that alerts of sirens in Israel.Footnote ¹²

Israel's Home Front Command divides large cities in Israel into areas such that a single rocket targeted toward the city does not automatically initiate sirens across the entire city but rather just for the areas that are at risk. The data on sirens and maternal residence are just at the locality level. As such, it is likely that the number of sirens in large cities does not reflect the true number of sirens mothers experienced in the locality. For this reason, for the three largest cities in our data—Beer Sheva, Ashkelon, and Ashdod—we divide the total number of sirens by two.Footnote ¹³

Based on our derived conception date, we assign to each mother the number of sirens in her locality during each trimester of her pregnancy. If exposure to OPE or to sirens affected the birthdate, then the number of sirens a mother experiences is endogenous to our outcomes of interest. To address this potential violation of our exclusion restriction, we follow [Persson and Rossin-Slater (Reference Persson and Rossin-Slater2018)], and define our treatment variables—the number of sirens experienced during pregnancy trimesters—relative to the expected due date at full term rather than the actual birthdate.

We note that we only observe maternal locality when giving birth, rather than during pregnancy. This does not pose a problem to our analysis as long as migration is minimal (so that the noise in assigning sirens based on locality at birth remains minimal) and migration patterns post-OPE are not systematically correlated with the number of sirens in a locality. We test for this using data from the Israel Central Bureau of Statistics (CBS) on the migration balance for all towns and cities in Israel in 2015. Migration balances are relatively small—2.5–4.8 individuals per 1,000 residents. Furthermore, we do not find a correlation between migration balances per 1,000 residents in Jewish towns and the total number of sirens the town experienced during OPE (coefficient estimates for various regressions were positive with p-values ranging between 0.21 and 0.38). Actual results are presented in Table B1 in online Appendix.

Our birth and siren data are complemented by locality-level data from the CBS. This particularly includes the locality socioeconomic ranking, an official measure in integers ranging between 1 (the lowest) and 10 (the highest), which is based on population characteristics such as the mean age, dependency ratio, the share of families with 4 or more children, educational attainment, employment and retirement, and living standards (mean income, vehicle ownership, and travel abroad). We separately examine effects for mothers residing in localities that are ranked low and high socioeconomically among Jewish localities.

3.1. Descriptive statistics

Table 1 presents summary statistics for pre-birth characteristics and siren exposure for the main sample of births used in our regression analysis—births conceived during 2011, 2013, and the first half of 2014. We test for potential selection of mothers based on exposure to OPE at pregnancy trimesters with t-tests for differences in the mean characteristics between mothers who did not experience OPE and mothers who did experience OPE in any of the trimesters. P-values for these t-tests are presented in columns 3, 5, and 7. Most pre-birth characteristics are balanced between the sub-samples, with the exception of differences in the abortion history and c-section history indicators between mothers experiencing OPE during the third trimester and mothers who did not experience OPE. These are two differences out of 24 differences that are tested for in Table 1, so statistically significant differences for two tests are possible by chance. We further note that there are several p-values for the t-test results presented in Table 1 that are less than 0.2. In total, there are 6 such p-values, which can also result by chance in a series of t-tests.Footnote ¹⁴

Table 1. Summary statistics—pre-birth characteristics

Notes: The sample is limited to all Jewish births in Soroka Hospital conceived during 2011, 2013, and the first half of 2014 with gestation length of at least 180 days and residents of localities in Southern Israel. The number of births is smaller for first trimester exposure to OPE due to the exclusion of all births conceived during OPE to avoid selection concern. Columns 1, 2, 4, and 6 present means with standard deviations in parenthesis. Columns 3, 5, and 7 present p-values for a t-test of the difference between the means of the relevant trimester during OPE and those that did not experience OPE during pregnancy. The samples for experiencing OPE during the first, second, or third pregnancy trimesters are not mutually exclusive.

The last row in Table 1 shows the mean number of sirens experienced by mothers exposed to OPE during each of the pregnancy trimesters. Figure 1 displays how the number of sirens is distributed across births that are exposed to OPE during each trimester. Bunching at some of the siren values in Figure 1 is due to more births being attributed to larger towns that experienced that amount of sirens. We note though that even within these towns, there is variation across births exposed to OPE in the number of sirens due to different dates for each of the trimesters across births. In Figure D1 in online Appendix, we present the same histograms as in Figure 1, only separately based on the socioeconomic ranking of mothers’ locality.

Figure 1. Number of sirens experienced during OPE by pregnancy trimester exposure to OPE.

Notes: The sample is limited to all Jewish births in Soroka Hospital conceived during 2013 and through the first half of 2014 with gestation length of at least 180 days and residents of localities in Southern Israel. Each histogram is limited to births that were exposed to OPE during the relevant trimester. The number of births is smaller for first trimester exposure to OPE due to the exclusion of all births conceived during OPE to avoid selection concern. Total number of localities is 85.

4. Empirical strategy

Our main regression specification examines the linear relationship between sirens experienced during each of the pregnancy trimesters and birth outcomes. This results in the following regression specification:

(1)$$\eqalign{Outcome_{ilym} & = \alpha_{0} + \sum_{t = 1}^{3}\alpha_{1}^{t}Sirens_{ilym}^{t}\cr & \quad + \eta X'_{iym} + \rho_{m} + \theta_{ly} + \varepsilon_{ilym}.}$$

Our dependent variables are birth outcomes for mother i in locality l with a child conceived in year and month (y, m). The main coefficients of interest in equation (1) are $\alpha _{1}^{t}$ for tε {1, 2, 3} that represent trimesters 1,2, and 3, respectively. These three coefficients of interest indicate how birth outcomes change for each additional siren the mother experiences during the specific pregnancy trimester in comparison to mothers who experienced no sirens during their pregnancy.

Equation (1) controls for pre-birth maternal and fetal characteristics [X′_iym—mother's age (categorical), child sex, maternal abortion history, birth parity (categorical), high fertility, past cesarean delivery, late pregnancy loss in the past, and infertility treatment].Footnote ¹⁵ ρ _m are month of conception fixed effects.

The intensity of sirens may be correlated with locality characteristics that may also be determinants of birth outcomes—for example, sirens were relatively more frequent in larger localities and the larger localities in southern Israel are relatively more disadvantaged. Locality fixed effects can control for that. However, if locality characteristics are changing over time and this is correlated with OPE exposure at the end of our sample period, then locality fixed effects cannot control for that. As such, our regression specification is more conservative and controls for locality-year-of-conception fixed effects, θ _ly, rather than just locality fixed effects.

Our results will demonstrate the importance of accounting for differential effects of sirens on birth outcomes based on socioeconomic status. Our only source for characterizing maternal socioeconomic status is the socioeconomic ranking of the mother's town, categorized on a scale of 1–10 by the Israel CBS. As such, we utilize a variation of equation (1) by interacting our siren count variables for each trimester with indicator variables for the mother residing in a low socioeconomically ranked town (4 or less in the CBS ranking) or high socioeconomically ranked town (5 or more). This results in the following regression specification:

(2)$$\eqalign{Outcome_{ilym} & = \beta_{0} + \sum_{t = 1}^{3}\beta_{1}^{t}Sirens_{ilym}^{t}\ast LowSocio_{ly}\cr & \quad + \sum_{t = 1}^{3}\beta_{2}^{t}Sirens_{ilym}^{t}\ast HighSocio_{ly} + \eta X'_{iym} + \rho_{m} + \theta_{ly} + \varepsilon_{ilym}.}$$

In equation (2), all variables are as defined in equation (1), only there are now six coefficients of interest—the three siren counts for each trimester interacted with the high and low socioeconomic categorizations.

We argue that our coefficients of interest are capturing a causal relationship between birth outcomes and sirens during OPE. This is under the assumption that absent OPE, birth outcomes would have continued on the same path they were on prior to OPE. Because OPE—especially its exact timing and its length—can be viewed as an exogenous shock in Israel, this identification assumption seems highly reasonable. We stress further that all treated births in our sample were conceived prior to OPE, and that it was not possible to foresee OPE even a month before its onset. Given the locality-year-of-conception fixed effects, we can broadly generalize our identification strategy to comparing between mothers conceiving during the same year and residing in the same locality, only their conception dates vary by several months. This difference in the conception date generates exogenous variation in the number of sirens each of these mothers experienced. This is while accounting for seasonality of conception by including in our regression specification month of conception fixed effects.Footnote ¹⁶

6. Robustness checks

6.1. Placebo analysis

We examine whether the results in section 5 are due to pre-existing trends. For this, we run the specification in equation (1) but on data that assigns the OPE exposure and siren variables as if OPE occurred in summer 2010 rather than summer 2014. The sample for this is births conceived during the first half of 2010, all of 2011, and 2009 beginning in April.Footnote ²¹ The results presented in Table 5 are reassuring. While the coefficient estimate for third trimester sirens is statistically significant for pre-term birth, all other 11 coefficient estimates are not statistically significant. With the exception of the third trimester birth weight coefficient estimate, with a p-value of 0.18, all other p-values are at least 0.24. Thus, this result may be by chance.

Table 5. Placebo analysis—assigning OPE to summer 2010 births

Notes: Number of localities is 78. The coefficient estimates presented are α ₁ − α ₃ from equation (1). Birth-specific control variables are: mother's age (categorical), child sex, maternal abortion history, birth parity (categorical), high fertility, past cesarean delivery, late pregnancy loss in the past, and infertility treatment. Mean of dependent variable is the mean for non-OPE births in the sample, in the above case births not assigned OPE treatment based on the placebo OPE. Standard errors are clustered at the locality level.

***p < 0.01, **p < 0.05, *p < 0.1.

6.2. Testing for selection of mothers giving birth during operation protective edge

One can argue that it is problematic to include births that took place during OPE in Soroka Hospital, as there may be selection of a different population giving birth during this period. As we will show below, this does not appear to be the case. Furthermore, these births are the main sample of births that experienced OPE during their third trimester. Given this, excluding these births would limit substantially our ability to estimate the effect of third trimester sirens on birth outcomes. Lastly, we note that the effects reported in response to third trimester births are mixed in Table 3. While most coefficient estimates point at worse birth outcomes in response to third trimester sirens, there is also a statistically significant decline in the probability of a pre-term birth among mothers ranked low socioeconomically in response to third trimester sirens. Thus, it is not clear which direction the selection would be, if there were any.

If different mothers came to give birth in Soroka during OPE, then the distribution of birth months over the course of the year of OPE—2014—should be different in comparison to other years. We would expect less births during July–August 2014 when OPE was taking place. We test for this in Table 6 by running regressions with the dependent variable being an indicator for a birth occurring during July–August. These regressions are run for the sample of births occurring in 2007–2009, 2011–2012, and 2014 during January through October, in order to exclude births exposed to other military operations (see section 3 on sample restrictions due to this). Two alternative specifications explore two different explanatory variables of interest. In the first, the main explanatory variable is an indicator for a birth taking place during 2014, thus comparing the probability of birth during July–August 2014 to that of all other years in the regression sample. In the second specification, the main explanatory variable is the total amount of sirens in the mother's locality during OPE divided by 20.Footnote ²² The latter regression specification controls also for year-of-birth fixed effects. All regression specifications for Table 6 control for locality fixed effects, linear time trends, and birth-specific controls. Standard errors are clustered in the year of birth level in the former regression and at the locality level in the latter regression.

Table 6. Testing for different probability of births during July–August 2014

Notes: The sample includes birth occurring during January–October in 2007–2009, 2011–2012, and 2014. Birth-specific control variables are: mother's age (categorical), child sex, maternal abortion history, birth parity (categorical), high fertility, past cesarean delivery, late pregnancy loss in the past, and infertility treatment. Columns (3)–(4) limit the sample to maternal localities that are ranked socioeconomically at level 4 or less. Columns (5)–(6) limit the sample to maternal localities that are ranked socioeconomically at level 5 or more. Standard errors in columns (1), (3), and (5) are clustered at the year of birth level, while in columns (2), (4), and (6) at the locality level.

***p < 0.01, **p < 0.05, *p < 0.1.

The results in Table 6 do not show statistically significant changes in the probability of a birth during July–August 2014, in comparison to other years—both in general (first column) and as a function of the number of sirens in the mother's locality (second column). Furthermore, these results hold when examining only the sample of births from mothers either from disadvantaged or more advantaged localities. Thus, the evidence does not suggest more or less births in Soroka during OPE.Footnote ²³

6.3. Different specifications

Equations (1) and (2) assume a linear relationship between sirens and birth outcomes. We verify whether this is appropriate for capturing the potential effect of sirens on birth outcomes using two alternative specifications: a non-parametric specification and a non-linear quadratic specification.

For the non-parametric specification, we split the siren count variables for each trimester into indicator variables for the following: 1–20 sirens, 21–40 sirens, and 41 + sirens, producing a total of 9 coefficients of interest. The excluded group consists of mothers who did not experience sirens during their pregnancy. This regression specification takes the following form:

(3)$$\eqalign{Outcome_{ilym} & = \gamma_{0} + \sum_{t = 1}^{3}\sum_{k = 1}^{3}\gamma_{k}^{t}SirensIndic_{ilym}^{t, k}\cr & \quad + \eta X'_{iym} + \rho_{m} + \theta_{ly} + \varepsilon_{ilym}.}$$

All dependent and control variables are as in equation (1). The main coefficients of interest in equation (3) are $\gamma _{1}^{t}-\gamma _{3}^{t}$ for tε {1, 2, 3} that represent trimesters 1,2, and 3, respectively. These 9 coefficients of interest indicate how birth outcomes change when the mother experienced the respective range of sirens during the specific pregnancy trimester in comparison to mothers who experienced no sirens during their respective trimester.

The results of this specification are outlined in Figure 2 for our four main dependent variables. The coefficient estimates for each siren range and pregnancy trimester are the dots and the vertical lines represent their 95% confidence intervals. The results suggest both adverse and beneficial effects of sirens on birth outcomes. For example, a large amount of sirens during the first trimester appears to positively affect both birth weight and gestation length. On the other hand, a large range of sirens during the third trimester decreases birth weight and gestation length and increases the probability of low birth weight. With the exception of the pre-term dependent variable, which exhibits statistically significant effects in response to the lowest siren range, all other patterns exhibited in Figure 2 are consistent with a linear relationship between sirens and the birth outcome.

Figure 2. Non-parametric effect of sirens on birth outcomes.

Notes: The sample is limited to all Jewish births in Soroka Hospital conceived during 2011, 2013, and 2014 through June with gestation length of at least 180 days and residents of 85 localities in southern Israel. See section 3 for information on the sample. Each figure plots the coefficient estimates γ ^t₁−γ ^t₃ for tϵ {1, 2, 3} from a equation (3). Each dot represents the coefficient estimate for the indicator variable on the horizontal axis and the vertical lines branching from the dot are the 95% confidence intervals. The various trimester dots for each of the siren count variables are supposed to be vertically aligned but are not for presentation purposes. Birth-specific control variables are: mother's age (categorical), child sex, maternal abortion history, birth parity (categorical), high fertility, past cesarean delivery, late pregnancy loss in the past, and infertility treatment. Standard errors are clustered at the locality level.

Due to some patterns in Figure 2 that may suggest a non-linear relationship between sirens and birth outcomes, we present in Figure 3 results from a specification that allows for a quadratic relationship between sirens and birth outcomes, differentiated by low and high maternal locality socioeconomic ranking (as in equation (2)). The figure presents predicted values from these specifications for various siren counts. First, the results exhibit for birth weight and low birth weight a qualitatively different effect of sirens based on maternal locality socioeconomic ranking, in particular in response to first and third trimester sirens. For the first trimester response, this is highly consistent with the results in Table 3. Second, when the predicted values for positive siren counts are statistically different from those for zero siren counts, the patterns plotted do not overwhelmingly appear non-linear. This lends further support to our linear specifications in equations (1) and (2).

Figure 3. Sirens and birth outcomes—non-linear specification differentiated by socioeconomic status. (a) Birth weight and gestation length. (b) Low birth weight and preterm.

Notes: The figures plot the predictive values and the 95% confidence intervals from results for regression specifications that are similar to equation (2) only the siren variables are quadratic. Siren values are listed on the horizontal axis during the first, second, and third trimester of pregnancy in the left, middle, and right columns, respectively

In the online Appendix, we further challenge our regression specification by examining the relationship between sirens per 1,000 locality population and birth outcomes, under the assumption that the effect of a siren may be smaller in larger localities that are more spread. The results are not as precise as those presented in Table 3.

6.4. Different gestation length thresholds

In section 3, we explain that the number of sirens assigned to each mother in constructing the sirens treatment variable is based on the mother's due date rather than her child's birth date. This addresses the potential endogeneity of the birthdate, given that our variables of interest affect gestational length. Our sample is limited to births with a gestational length of at least 180 days. This results in a small amount of mothers—those who gave birth substantially earlier than their due date—having a large amount of sirens assigned to them despite not experiencing these sirens during their pregnancy. This distortion would primarily be relevant to the number of sirens assigned during the third trimester.

To test whether assignment of sirens based on the mother's due date affected our results in any way, we present regression results in Table 7 for the same specification as in Table 3, only the minimal gestation length threshold varies. The advantage of this more strict sample criterion is that the number of sirens more precisely captures the actual number of sirens mothers experienced in their locality during their third trimester. The disadvantage of this sample restriction is reduced variation in our outcomes of interest. Despite this disadvantage, the results in Table 7 still show statistically significant benefits in response to sirens among mothers ranked high socioeconomically and adverse effects for mothers ranked low socioeconomically in response to sirens during the first trimester.

Table 7. Operation protective edge and birth outcomes—different gestational length thresholds for sample

Notes: The number of localities is 83–85. See section 3 for information on the sample. The coefficient estimates presented are β ₁ − β ₆ in equation (2). The first row indicates the minimal gestational length threshold for the sample in the regression. Birth-specific control variables are: mother's age (categorical), child sex, maternal abortion history, birth parity (categorical), high fertility, past cesarean delivery, late pregnancy loss in the past, and infertility treatment. Mean of dependent variable is the mean for non-OPE births in the sample. Standard errors are clustered at the locality level.

***p < 0.01, **p < 0.05, *p < 0.1.

7. Concluding remarks

As the economics literature has progressed from demonstrating the importance of birth weight for long-term adult outcomes [Behrman and Rosenzweig (Reference Behrman and Rosenzweig2004); Black et al. (Reference Black, Devereux and Salvanes2007)] to establishing the central role intrauterine shocks play in determining birth weight and other birth outcomes [Almond and Currie (Reference Almond and Currie2011)], an abundant research has emerged that evaluates various potential intrauterine shocks and their effect on birth outcomes. When possible, these studies also evaluate the effect of these shocks on long-term adult outcomes or even offspring outcomes. This study evaluates one such shock—Operation Protective Edge in Israel—which resulted in the civilian population experiencing numerous daily siren warnings of rocket attacks over a two-month period.

Our results show that the effect of a conflict on birth outcomes can be both positive and negative, depending on behavioral responses to this shock, in particularly in terms of prenatal care. Furthermore, prenatal care responses—at least in our setting—vary qualitatively based on maternal socioeconomic status. Mothers ranked high socioeconomically likely have the resources to increase pro-health behavior during pregnancy in response to OPE, while mothers ranked low socioeconomically do not appear to have this option and even respond behaviorally in ways that are detrimental to their pregnancies. These differential responses produce qualitatively different birth outcome effects, thus demonstrating the potential for conflict—and possibly other negative shocks that can involve intrauterine exposure—to aggravate further socioeconomically-based inequality.