Geographic microtargeting of social assistance with high-resolution poverty maps.

Hundreds of millions of poor families receive some form of targeted social assistance. Many of these antipoverty programs involve some degree of geographic targeting, where aid is prioritized to the poorest regions of the country. However, policy makers in many low-resource settings lack the disaggregated poverty data required to make effective geographic targeting decisions. Using several independent datasets from Nigeria, this paper shows that high-resolution poverty maps, constructed by applying machine learning algorithms to satellite imagery and other nontraditional geospatial data, can improve the targeting of government cash transfers to poor families. Specifically, we find that geographic targeting relying on machine learning-based poverty maps can reduce errors of exclusion and inclusion relative to geographic targeting based on recent nationally representative survey data. This result holds for antipoverty programs that target both the poor and the extreme poor and for initiatives of varying sizes. We also find no evidence that machine learning-based maps increase targeting disparities by demographic groups, such as gender or religion. Based in part on these findings, the Government of Nigeria used this approach to geographically target emergency cash transfers in response to the COVID-19 pandemic.

Mobile phone data reveal the effects of violence on internal displacement in Afghanistan.

Nearly 50 million people globally have been internally displaced due to conflict, persecution and human rights violations. However, the study of internally displaced persons-and the design of policies to assist them-is complicated by the fact that these people are often underrepresented in surveys and official statistics. We develop an approach to measure the impact of violence on internal displacement using anonymized high-frequency mobile phone data. We use this approach to quantify the short- and long-term impacts of violence on internal displacement in Afghanistan, a country that has experienced decades of conflict. Our results highlight how displacement depends on the nature of violence. High-casualty events, and violence involving the Islamic State, cause the most displacement. Provincial capitals act as magnets for people fleeing violence in outlying areas. Our work illustrates the potential for non-traditional data sources to facilitate research and policymaking in conflict settings.

Machine learning and phone data can improve targeting of humanitarian aid.

The COVID-19 pandemic has devastated many low- and middle-income countries, causing widespread food insecurity and a sharp decline in living standards1. In response to this crisis, governments and humanitarian organizations worldwide have distributed social assistance to more than 1.5 billion people2. Targeting is a central challenge in administering these programmes: it remains a difficult task to rapidly identify those with the greatest need given available data3,4. Here we show that data from mobile phone networks can improve the targeting of humanitarian assistance. Our approach uses traditional survey data to train machine-learning algorithms to recognize patterns of poverty in mobile phone data; the trained algorithms can then prioritize aid to the poorest mobile subscribers. We evaluate this approach by studying a flagship emergency cash transfer program in Togo, which used these algorithms to disburse millions of US dollars worth of COVID-19 relief aid. Our analysis compares outcomes-including exclusion errors, total social welfare and measures of fairness-under different targeting regimes. Relative to the geographic targeting options considered by the Government of Togo, the machine-learning approach reduces errors of exclusion by 4-21%. Relative to methods requiring a comprehensive social registry (a hypothetical exercise; no such registry exists in Togo), the machine-learning approach increases exclusion errors by 9-35%. These results highlight the potential for new data sources to complement traditional methods for targeting humanitarian assistance, particularly in crisis settings in which traditional data are missing or out of date.

Microestimates of wealth for all low- and middle-income countries.

Many critical policy decisions, from strategic investments to the allocation of humanitarian aid, rely on data about the geographic distribution of wealth and poverty. Yet many poverty maps are out of date or exist only at very coarse levels of granularity. Here we develop microestimates of the relative wealth and poverty of the populated surface of all 135 low- and middle-income countries (LMICs) at 2.4 km resolution. The estimates are built by applying machine-learning algorithms to vast and heterogeneous data from satellites, mobile phone networks, and topographic maps, as well as aggregated and deidentified connectivity data from Facebook. We train and calibrate the estimates using nationally representative household survey data from 56 LMICs and then validate their accuracy using four independent sources of household survey data from 18 countries. We also provide confidence intervals for each microestimate to facilitate responsible downstream use. These estimates are provided free for public use in the hope that they enable targeted policy response to the COVID-19 pandemic, provide the foundation for insights into the causes and consequences of economic development and growth, and promote responsible policymaking in support of sustainable development.

Public mobility data enables COVID-19 forecasting and management at local and global scales.

Policymakers everywhere are working to determine the set of restrictions that will effectively contain the spread of COVID-19 without excessively stifling economic activity. We show that publicly available data on human mobility-collected by Google, Facebook, and other providers-can be used to evaluate the effectiveness of non-pharmaceutical interventions (NPIs) and forecast the spread of COVID-19. This approach uses simple and transparent statistical models to estimate the effect of NPIs on mobility, and basic machine learning methods to generate 10-day forecasts of COVID-19 cases. An advantage of the approach is that it involves minimal assumptions about disease dynamics, and requires only publicly-available data. We evaluate this approach using local and regional data from China, France, Italy, South Korea, and the United States, as well as national data from 80 countries around the world. We find that NPIs are associated with significant reductions in human mobility, and that changes in mobility can be used to forecast COVID-19 infections.

The strength of long-range ties in population-scale social networks.

Long-range connections that span large social networks are widely assumed to be weak, composed of sporadic and emotionally distant relationships. However, researchers historically have lacked the population-scale network data needed to verify the predicted weakness. Using data from 11 culturally diverse population-scale networks on four continents-encompassing 56 million Twitter users and 58 million mobile phone subscribers-we find that long-range ties are nearly as strong as social ties embedded within a small circle of friends. These high-bandwidth connections have important implications for diffusion and social integration.

Using gene expression ratios to predict outcome among patients with mesothelioma.

BACKGROUND: We have recently demonstrated that simple ratios of the expression levels of selected genes in tumor samples can be used to distinguish among types of thoracic malignancies. We examined whether this technique could predict treatment-related outcome for patients with mesothelioma. METHODS: We used gene expression profiling data previously collected from 17 mesothelioma patients with different overall survival times to define two outcome-related groups of patients and to train an expression ratio-based outcome predictor model. A Student's t test was used to identify genes among the two outcome groups that had statistically significant, inversely correlated expression levels; those genes were used to form prognostic expression ratios. We used a combination of several highly accurate expression ratios and cross-validation techniques to assess the internal consistency of this predictor model, quantitative reverse transcription-polymerase chain reaction of tumor RNA to confirm the microarray data, and Kaplan-Meier survival analysis to validate the model among an independent set of 29 mesothelioma tumors. All statistical tests were two-sided. RESULTS: We developed an expression ratio-based test capable of identifying 100% (17/17) of the samples used to train the model. This test remained highly accurate (88%, 15/17) after cross-validation. A four-gene expression ratio test statistically significantly (P =.0035) predicted treatment-related patient outcome in mesothelioma independent of the histologic subtype of the tumor. CONCLUSIONS: Gene expression ratio-based analysis accurately predicts treatment-related outcome in mesothelioma samples. This technique could impact the clinical treatment of mesothelioma by allowing the preoperative identification of patients with widely divergent prognoses.

Translation of microarray data into clinically relevant cancer diagnostic tests using gene expression ratios in lung cancer and mesothelioma.

The pathological distinction between malignant pleural mesothelioma (MPM)and adenocarcinoma (ADCA) of the lung can be cumbersome using established methods. We propose that a simple technique, based on the expression levels of a small number of genes, can be useful in the early and accurate diagnosis of MPM and lung cancer. This method is designed to accurately distinguish between genetically disparate tissues using gene expression ratios and rationally chosen thresholds. Here we have tested the fidelity of ratio-based diagnosis in differentiating between MPM and lung cancer in 181 tissue samples (31 MPM and 150 ADCA). A training set of 32 samples (16 MPM and 16 ADCA) was used to identify pairs of genes with highly significant, inversely correlated expression levels to form a total of 15 diagnostic ratios using expression profiling data. Any single ratio of the 15 examined was at least 90% accurate in predicting diagnosis for the remaining 149 samples (e.g., test set). We then examined (in the test set) the accuracy of multiple ratios combined to form a simple diagnostic tool. Using two and three expression ratios, we found that the differential diagnoses of MPM and lung ADCA were 95% and 99% accurate, respectively. We propose that using gene expression ratios is an accurate and inexpensive technique with direct clinical applicability for distinguishing between MPM and lung cancer. Furthermore, we provide evidence suggesting that this technique can be equally accurate in other clinical scenarios.