# Fertility and its Meaning: Evidence from Search Behavior

## TL;DR

This study uses Google search data and survey information to analyze socio-economic differences in fertility-related behaviors and to predict regional fertility variations and trends in the U.S.

## Contribution

It introduces a novel approach combining search behavior data with survey data to understand fertility differences and predict regional fertility patterns.

## Key findings

- Google search queries reveal socio-economic fertility differences.
- Predictive models explain 75% of regional fertility variation.
- Web search data can track fertility trend changes.

## Abstract

Fertility choices are linked to the different preferences and constraints of individuals and couples, and vary importantly by socio-economic status, as well by cultural and institutional context. The meaning of childbearing and child-rearing, therefore, differs between individuals and across groups. In this paper, we combine data from Google Correlate and Google Trends for the U.S. with ground truth data from the American Community Survey to derive new insights into fertility and its meaning. First, we show that Google Correlate can be used to illustrate socio-economic differences on the circumstances around pregnancy and birth: e.g., searches for "flying while pregnant" are linked to high income fertility, and "paternity test" are linked to non-marital fertility. Second, we combine several search queries to build predictive models of regional variation in fertility, explaining about 75% of the variance. Third, we explore if aggregated web search data can also be used to model fertility trends.