 Measure immigration is, in a way, a little bit like trying to get the measurements of a person when all you have are reflections of that person from a distorting mirror. One of those mirrors that kids play with, at the fair, those that makes you taller or wider or curvy and so on. And all data have their own imperfections. In the past we only had a few mirrors from which we could gather information about migration, typically mirrors like administrative data or surveys. Now we have many more mirrors. Those include geolocated data from, say, Twitter data or call data records from cell phone providers. We have advertisement data that are produced by most social media companies like Facebook, LinkedIn, Google and so on. As well as many new opportunities in order to ask questions, in order to recruit participants into a service, for example via advertisements like the Facebook advertisement platform. There are many opportunities and challenges. The main opportunity is that we can really move forward in areas like understanding the impact of events like natural disasters, expanding the definition and scope of migration estimates, evaluating migrations of professionals or very highly skilled individuals, understanding relationships between internal and international migration or short-term mobility and long-term mobility, evaluating cultural patterns of integration, for example in cultural tastes and so on. So these are very exciting questions that we can address with new data. At the same time, these data are difficult to understand and require collaborations with companies often. And the goals or the objectives of companies and scientists are not always aligned. And so aligning those goals is important together with bringing in national statistical offices in order to produce something that is sustainable in the long term. So it really depends on the questions. So for certain types of questions, we actually got very good estimates. For example, in the case of estimating the flows of people from Puerto Rico to mainland US or to continental US after Eureka and Maria, we got very accurate estimates of about 180,000 people moving to continental US and then some of them coming back. But for other questions where we really need sustainable patterns over a long time, we're not there yet in terms of guaranteeing the sustainability. So in terms of one-off studies, we have many, many of those. But what we really need is to build partnerships between academia and industry in order to build sustainable patterns.