Search

This paper studies the dynamics of unemployment (u) and its natural rate (u*), with u* measured by real-time estimates for 29 countries from the OECD. We find strong evidence of hysteresis: an innovation in u causes u* to change in the same direction, and therefore has permanent effects. For our baseline specification, a one percentage point deviation of u from u* for one year has a long-run effect of 0.16 points on both variables. When we allow asymmetry, we find, perhaps surprisingly, that decreases in u have larger long-run effects than increases in u.

Significance This paper examines whether the opening of K–12 schools may lead to the spread of COVID-19. Analyzing how an increase of COVID-19 cases is related to the timing of opening K–12 schools in the United States, we find that counties that opened K–12 schools with in-person learning experienced an increase in the growth rate of cases by 5 percentage points on average, controlling for a variety of policies, past infection rates, and other factors. This association of K–12 school visits with case growth is stronger when mask wearing is not mandated for staff at school. These findings support policies that promote masking and other precautionary measures at schools and giving vaccine priority to education workers., This paper empirically examines how the opening of K–12 schools is associated with the spread of COVID-19 using county-level panel data in the United States. As preliminary evidence, our event-study analysis indicates that cases and deaths in counties with in-person or hybrid opening relative to those with remote opening substantially increased after the school opening date, especially for counties without any mask mandate for staff. Our main analysis uses a dynamic panel data model for case and death growth rates, where we control for dynamically evolving mitigation policies, past infection levels, and additive county-level and state-week “fixed” effects. This analysis shows that an increase in visits to both K–12 schools and colleges is associated with a subsequent increase in case and death growth rates. The estimates indicate that fully opening K–12 schools with in-person learning is associated with a 5 (SE = 2) percentage points increase in the growth rate of cases. We also find that the association of K–12 school visits or in-person school openings with case growth is stronger for counties that do not require staff to wear masks at schools. These findings support policies that promote masking and other precautionary measures at schools and giving vaccine priority to education workers.

How much has the US-China tariff war impacted economic outcomes in China? We address this question using high-frequency night lights data, together with measures of the trade exposure of fine grid locations constructed from Chinese firms' geo-coordinates. Exploiting within-grid variation over time and controlling extensively for grid-specific contemporaneous trends, we find that each 1 percentage point increase in exposure to the US tariffs was associated with a 0.59% reduction in night-time luminosity. We combine these with structural elasticities that relate night lights to economic outcomes, motivated by the statistical framework of Henderson et al. (2012). The negative impact of the tariff war was highly skewed across locations: While grids with negligible direct exposure to the US tariffs accounted for up to 70% of China's population, we infer that the 2.5% of the population in grids with the largest US tariff shocks saw a 2.52% (1.62%) decrease in income per capita (manufacturing employment) relative to unaffected grids. By contrast, we do not find significant effects from China's retaliatory tariffs.

In this paper, we study the role of coworker referrals for labor market outcomes. Using comprehensive Danish administrative data covering the period 1980 to 2005, we first document a strong tendency of workers to follow their former coworkers into the same establishments and provide evidence that these mobility patterns are likely driven by coworker referrals. Treating the presence of a former coworker in an establishment at the time of hiring as a proxy for a referral, we then show that referred workers initially earn 4.6 percent higher wages and are 2.3 percentage points less likely to leave their employers than workers hired through the external market. Consistent with a theoretical framework characterized by higher initial uncertainty in the external market but the possibility of subsequent learning about match-specific productivity, we show that these initial differences gradually decline as tenure increases. We structurally estimate a stylized model using indirect inference and find that the noise of the initial signal about a worker's productivity is 14.5 percent lower in the referral market than in the external market, and that firms learn about their workers' true match-specific productivity with a probability of 48.4 percent per year. Counterfactual simulations show that average wages are lower in the absence of a referral market, primarily because of lower average match productivity in the external market.

Background The COVID-19 pandemic disproportionately affected minorities in population rates of infection, hospitalization, and mortality. However, little is known about the broader racial disparities in fears and perceptions about the pandemic and getting treated. Objective To examine disparities in perceived risks of COVID-19 and getting medical care. Methods Using the nationally representative Stanford University School of Medicine Coronavirus Attitudes and Behaviors Survey fielded in May of 2020, we examine racial and ethnic disparities in eight measures on the perceived risks of COVID-19. We use regression analysis to risk adjust perceptions controlling for 10 socioeconomic, demographic, and health variables. Results Black respondents were 15 percentage points more likely than White respondents to believe the pandemic would not end by Summer 2020 (92% vs 77%, p

In the last decade, Gulf countries have imposed hiring quotas to promote the participation of natives in the private sector and address high levels of unemployment, particularly among women and the youth. This paper explores how one such policy, Nitaqat in Saudi Arabia, affected the outcomes of exporting firms, the most productive sector of the non-oil economy. We find that whereas the policy was successful in increasing the employment of Saudi nationals by these firms, it came at a high cost. In the year following the announcement of the policy, relative to firms above the quota, firms below the quota were 1.5 percentage points more likely to exit the market, 7 percentage points less likely to export, and conditional on exporting, the value of their exports fell by 14 percent. Additionally, surviving treated firms reduced their labor force by 10 percent. We find that to comply with the policy, firms hired mostly lower-wage, low-skilled Saudis. The policy doubled the share of women in treated firms. Importantly, we find that these short-term effects persisted for at least three years after the policy’s implementation.

This paper advances the use of pre-analysis plans in non-experimental research settings. In a study of recent minimum wage changes, we demonstrate how analyses of medium- and long-run impacts of policy interventions can be pre-specified as extensions to short-run analyses. Further, our pre-analysis plan includes comparisons of the effects of large vs. small minimum wage increases, which is a theoretically motivated dimension of heterogeneity. We discuss how these use cases harness the strengths of pre-analysis plans while mitigating their weaknesses. This project’s initial analyses explored CPS and ACS data from 2011 through 2015. Alongside these analyses, we pre-committed to analyses incorporating CPS and ACS data extending through 2019. Averaging across the specifications in our pre-analysis plan, we estimate that relatively large minimum wage increases reduced employment rates among low-skilled individuals by just over 2.5 percentage points. Our estimates of the effects of relatively small minimum wage increases vary across data sets and specifications but are, on average, both economically and statistically indistinguishable from zero. We estimate that medium-run effects exceed short-run effects and that the elasticity of employment with respect to the minimum wage is substantially more negative for large minimum wage increases than for small increases.

OBJECTIVE The proportion of multiple births has risen rapidly worldwide. Multiple births are likely to affect birth weight, which results in low birth weight (LBW) of less than 2500 g, possibly, because multiples are more likely to be born prematurely or less than 37 weeks into pregnancy. Using data from South Korea, this study aims to estimate the contribution of the rise in multiples to the rise in LBW incidence. METHODS Based on data from 2000 to 2017, we estimated the effect of multiples on LBW rates using linear regression analysis. Based on the regression analysis and the change in the proportion of multiples during this period, we calculated the contribution of the rise in multiples to the rise in LBW incidence using the total differential. In other words, we divided the change in LBW during the period due to the change in multiples by the total change during the period. The data are from the birth registry of the National Statistical Office, which contains information on the 8.4 million live births during the period 2000-2017. RESULTS We found that a 1 percentage point increase in multiples increases the proportion of LBWs by 0.495 percentage points. In addition, because the changes in the proportion of multiples and LBWs from 2000 to 2017 are 2.2 and 2.4 percentage points, respectively, 1.1 percentage points or 45.8% of the increase in LBWs over the period is due to the increase in multiples. CONCLUSION Since the Korean government introduced a measure to reduce the number of transferred embryos recently, one may expect that multiples in Korea would reduce in the near future, as it did in other countries. Subsequently, the incidence of LBW children is also likely to reduce, which is desirable in terms of the children's health outcomes.

The reproductive calendar is a data collection tool that collects month-by-month retrospective histories of contraceptive use. This survey instrument is implemented in large-scale demographic surveys, but its reliability is not well-understood. Our analysis helps to address this research gap, using longitudinal panel data with overlapping calendars from urban Kenya. Our findings indicate calendar data collected in 2014 underestimated 2012 reports of current use by 5 percentage points. And while the overall percentage of women reporting at least one episode of contraceptive use was similar across the two calendars (67 percent vs. 70 percent), there was notable disagreement in contraceptive behavior when comparing the histories of individual women; less than 20 percent of women with any contraceptive use reported the exact same pattern of use in both calendars. Low calendar reliability was especially apparent for younger women and those with complicated contraceptive histories. Individual-level discordance resulted in a small difference in 12-month discontinuation rates for the period of calendar overlap; when surveyed in 2014, women reported a 12-month discontinuation rate of 39 percent, compared to a rate of 34 percent reported in 2012. When using retrospective calendar data, attention must be paid to the potential for individual reporting errors.

BACKGROUND: Randomized trials are generally performed from a frequentist perspective, which can conflate absence of evidence with evidence of absence. The RECOVERY trial evaluated convalescent plasma for patients hospitalized with coronavirus disease 2019 (COVID-19) and concluded that there was no evidence of an effect. Re-analysis from a Bayesian perspective is warranted. METHODS: Outcome data were extracted from the RECOVERY trial by serostatus and time of presentation. A Bayesian re-analysis with a wide variety of priors (vague, optimistic, sceptical, and pessimistic) was performed, calculating the posterior probability for: any benefit, an absolute risk difference of 0.5% (small benefit, number needed to treat 200), and an absolute risk difference of one percentage point (modest benefit, number needed to treat 100). RESULTS: Across all patients, when analysed with a vague prior, the likelihood of any benefit or a modest benefit with convalescent plasma was estimated to be 64% and 18%, respectively. The estimated chance of any benefit was 95% if presenting within 7 days of symptoms, or 17% if presenting after this. In patients without a detectable antibody response at presentation, the chance of any benefit was 85%. However, it was only 20% in patients with a detectable antibody response at presentation. CONCLUSIONS: Bayesian re-analysis suggests that convalescent plasma reduces mortality by at least one percentage point among the 39% of patients who present within 7 days of symptoms, and that there is a 67% chance of the same mortality reduction in the 38% who are seronegative at the time of presentation. This is in contrast to the results in people who already have antibodies when they present. This biologically plausible finding bears witness to the advantage of Bayesian analyses over misuse of hypothesis tests to inform decisions.

Background The world has made considerable progress in the reduction of adolescent maternity and early marriage. However, this progress has been uneven, with many countries finding themselves far from achieving the Sustainable Development Goals in this dimension. We assessed levels and trends over time in adolescent marriage and maternity prevalence within the West and Central African region as well as their correlation with select macro-level indicators for income and social institutions. Methods We estimated country-specific prevalence rates using survey data (pooled cross-sectional) conducted between 1986 and 2017. The pooled sample provides information on 262 721 adolescent girls between the ages of 15 and 19. We assessed the relative country-level trends by comparing prevalence rates from the first and latest available survey in each country. We further analyzed regional trends by country income group (low- and middle-income) and examined the association of prevalence rates with measurements of gender discrimination and social institutions at the country-level. Estimations were conducted using survey weights and country-specific weights for population shares in the pooled sample. Results Prevalence of adolescent maternity declined from 30.1 percent (95% confidence interval (CI) = 29.6%-32.2%) in the 1990s, to 28.7 percent (95% CI = 27.9%-29.6%) in the 2000s and 26.2 percent (95% CI = 25.4%-27.1%) in the 2010s. Adolescent marriage rates decreased from 37.3 percent (95% CI = 35.5%-39.1%) in the 1990s to 27.5 percent (95% CI = 26.5%-28.6%) in the 2000s, and to 24.9 percent (95% CI = 24.1%-25.7%) in the 2010s. Between 1986 and 2017, adolescent marriage decreased in all countries except for the Central African Republic (with a rise from 39% to 55%) and Niger (56% to 61%). The prevalence of adolescent maternity decreased in all but three countries: Congo, Dem. Rep. (25% to 37%), Niger (36% to 40%), and the Central African Republic (36% to 49%). When grouped by income level, the prevalence was 8 percentage points higher in low-income countries than in middle-income countries in both outcomes. We did not establish any statisticly significant association between adolescent marriage and maternity with country-level measures of discrimination against women. However, we found evidence of an association between specific legal measures of protection against early marriage and lower prevalence rates for both early marriage and maternity. Conclusions Despite considerable progress in the reduction of adolescent maternity and marriage over the last 30 years, current levels of both indicators remain overall high in the WCA region, with high heterogeneity across individual countries. Countries with higher income level and higher standard in legal protection of young girls perform consistently better on both indicators. The prevalence rates of adolescent marriage and maternity reversed over the course of three decades, so that nowadays adolescent maternity rates exceed adolescent marriage rates in most countries. Further research is needed to understand the weak or non-existent association between adolescent marriage and maternity with gender discrimination and social institutions.

IntroductionCancer-related employment disruption contributes to financial toxicity and associated clinical outcomes through income loss and changes in health insurance and may not be uniformly experienced. We examined racial/ethnic differences in the financial consequences of employment disruption.MethodsWe surveyed a national sample of cancer patients employed at diagnosis who had received assistance from a national nonprofit about the impact of cancer diagnosis and treatment on employment. We used logistic regression models to examine racial/ethnic differences in income loss and changes in health insurance coverage.ResultsOf 619 cancer patients included, 63% identified as Non-Hispanic/Latinx (NH) White, 18% as NH Black, 9% as Hispanic/Latinx, 5% as other racial/ethnic identities, and 5% unreported. Over 83% reported taking a significant amount of time off from work during cancer diagnosis and treatment, leading to substantial income loss for 64% and changes in insurance coverage for 31%. NH Black respondents had a 10.2 percentage point (95% CI: 4.8 – 19.9) higher probability of experiencing substantial income loss compared to NH White respondents, and Hispanic or Latinx respondents had a 12.4 percentage point (95% CI: 0.3 – 24.5) higher probability compared to NH White respondents, controlling for clinical characteristics (i.e., cancer type, stage and age at diagnosis, and time since diagnosis). Similarly, NH Black respondents had a 9.3 percentage point (95% CI: -0.7 – 19.3) higher probability of experiencing changes in health insurance compared to NH White respondents, and Hispanic or Latinx respondents had a 10.0 percentage point (95% CI: -3.0 – 23.0) higher probability compared to NH White respondents.DiscussionCompared with NH White respondents, NH Black and Hispanic/Latinx respondents more commonly reported employment-related income loss and health insurance changes. Given documented racial/ethnic differences in job types, benefit generosity, and employment protections as a result of historic marginalization, policies to reduce employment disruption and its associated financial impact must be developed with a racial equity lens.

This study presents a recent estimate (for year 2017) of ice thickness and volume of a small glacier called Patsio Glacier located in Western Himalayas. Also, the present study aims to examine whether the improvement in GlabTop2 based estimates obtained for Chhota Shigri Glacier through the use of relatively accurate and high-resolution Digital Elevation Model (DEM) and glacier specific model parameterisation is consistent and can be extended to other glaciers. The version of GlabTop2 model used in this study is GlabTop2_IITB (Glacier Bed Topography version 2_Indian Institute of Technology Bombay). For validation purpose, ground penetrating radar (GPR) survey-based ice thickness estimates available at four cross-section profiles collected during July 2017 was used. Root mean square error (RMSE) of the modelled ice thickness estimates is about 14 m. Average and the maximum thickness of the modelled glacier ice is about 45 and 129 m, respectively. The obtained results when compared with an earlier study carried out using velocity-based approach indicate a significant improvement in the modelling accuracy. Moreover, a 16 percentage points reduction in uncertainty of the estimated ice thickness is seen when compared with a previous study involving different version of GlabTop2 model for Himalaya–Karakoram region. This is perhaps due to the use of a relatively accurate and high-resolution DEM and model parameterisation scheme adopted in this study, thus confirming the consistency of the modelling approach for different glaciers. Based on the obtained ice thickness estimates, the modelled volume of the glacier stored ice was calculated to be 0.11198 ± 0.0162 km3 in 2017.

Medicare provides nearly universal health insurance to individuals at age 65 years. How eligibility for Medicare affects racial and ethnic disparities in access to care and health is poorly understood.To assess the association of Medicare with racial and ethnic disparities in access to care and health.This cross-sectional study uses regression discontinuity to compare racial and ethnic disparities before and after age 65 years, the age at which US adults are eligible for Medicare. There are a total of 2 434 320 respondents in the Behavioral Risk Factor Surveillance System and 44 587 state-age-year observations in the US Centers for Disease Control and Prevention Wide-Ranging Online Data for Epidemiologic Research Data (eg, the mortality rate for individuals age 63 years in New York in 2017) from January 2008 to December 2018. The data were analyzed between February and May 2021.Eligibility for Medicare at age 65 years.Proportions of respondents with health insurance, as well as self-reported health and mortality. To examine access, whether respondents had a usual source of care, encountered cost-related barriers to care, or received influenza vaccines was assessed.Of 2 434 320 participants, 192 346 were Black individuals, 104 294 were Hispanic individuals, and 892 177 were men. Immediately after age 65 years, insurance coverage increased more for Black respondents (from 86.3% to 95.8% or 9.5 percentage points; 95% CI, 7.6-11.4) and Hispanic respondents (from 77.4% to 91.3% or 13.9 percentage points; 95% CI, 12.0-15.8) than White respondents (from 92.0% to 98.5% or 6.5 percentage points; 95% CI, 6.1-7.0). This was associated with a 53% reduction compared with the size of the disparity between White and Black individuals before age 65 years (5.7% to 2.7% or 3.0 percentage points; 95% CI, 0.9-5.1; P = .003) and a 51% reduction compared with the size of the disparity between White and Hispanic individuals before age 65 years (14.6% to 7.2% or 7.4 percentage points; 95% CI, 5.3-9.5; P .001). Medicare eligibility was associated with narrowed disparities between White and Hispanic individuals in access to care, lowering disparities in access to a usual source of care from 10.5% to 7.5% (P = .05), cost-related barriers to care from 11.4% to 6.9% (P .001), and influenza vaccination rates from 8.1% to 3.3% (P = .01). For disparities between White and Black individuals, access to a usual source of care before and after age 65 years was not significantly different: 1.2% to 0.0% (P = .24), cost-related barriers to care from 5.8% to 4.3% (P = .22), and influenza vaccinations from 11.0% to 10.3% (P = .60). The share of people in poor self-reported health decreased by 3.8 percentage points for Hispanic respondents, 2.6 percentage points for Black respondents, and 0.2 percentage points for White respondents. Mortality-related disparities at age 65 years were unchanged. Medicare's association with reduced disparities largely persisted after the US Affordable Care Act took effect in 2014.In this cross-sectional study that uses a regression discontinuity design, eligibility for Medicare at age 65 years was associated with marked reductions in racial and ethnic disparities in insurance coverage, access to care, and self-reported health.

Recent empirical work suggests that there are large agglomeration gains from working and living in developing country cities. These estimates find that doubling city size is associated with an increase in productivity by 19 percent in China, 12 percent in India, and 17 percent in Africa. These agglomeration benefits are considerably higher relative to developed country cities, which are in the range of 4 to 6 percent. However, many developing country cities are costly, crowded, and disconnected, and face slow structural transformation. To understand the true productivity advantages of cities in developing countries, this paper systematically evaluates more than 1,200 elasticity estimates from 70 studies in 33 countries. Using a frontier methodology for conducting meta-analysis, it finds that the elasticity estimates in developing countries are at most 1 percentage point higher than in advanced economies, but not significantly so. The paper provides novel estimates of the elasticity of pollution, homicide, and congestion, using a large sample of developing and developed country cities. No evidence is found for productivity gains in light of the high and increasing costs of working in developing country cities.

In this paper, we propose a data augmentation framework for Optical Character Recognition (OCR). The proposed framework is able to synthesize new viewing angles and illumination scenarios, effectively enriching any available OCR dataset. Its modular structure allows to be modified to match individual user requirements. The framework enables to comfortably scale the enlargement factor of the available dataset. Furthermore, the proposed method is not restricted to single frame OCR but can also be applied to video OCR. We demonstrate the performance of our framework by augmenting a 15% subset of the common Brno Mobile OCR dataset. Our proposed framework is capable of leveraging the performance of OCR applications especially for small datasets. Applying the proposed method, improvements of up to 2.79 percentage points in terms of Character Error Rate (CER), and up to 7.88 percentage points in terms of Word Error Rate (WER) are achieved on the subset. Especially the recognition of challenging text lines can be improved. The CER may be decreased by up to 14.92 percentage points and the WER by up to 18.19 percentage points for this class. Moreover, we are able to achieve smaller error rates when training on the 15% subset augmented with the proposed method than on the original nonaugmented full dataset.

The accurate characterisation of the instrumental spectral performance of field spectroradiometers acquires particular importance in view of the instruments' widespread use for calibrating and validating airborne- and satellite-based optical sensor systems. Typically, spectral calibration relies on the basic assumption of a Gaussian instrumental spectral response function (ISRF). We test alternative ISRF parameterisations and examine the resulting changes in spectral performance of four ASD field spectroradiometers. We then use a MODTRAN-simulated spectrum to compare an ideal Bottom Of Atmosphere (BOA) reflectance with the reflectances derived from ASD-convolved radiances using nominal (per ASD specifications), Gaussian and Symmetric Super Gaussian parameterisations of the spectroradiometers. We highlight the impact of the small but significant differences by retrieving the Photochemical Reflectance Index (PRI) values, which change by a few percentage points compared to the reference PRI.

In response to the Covid-19 outbreak, among other previous ``non-pharmaceutical interventions99, on March 22, 2020 the Italian Government imposed an economic lockdown and ordered the closing of all non-essential economic activities. This paper estimates the causal effect of this measure on mortality by Covid-19 and on mobility patterns. The identification of the causal effect exploits the variation in the number of active workers across municipalities induced by the economic lockdown. The difference-in-difference empirical design compares outcomes in municipalities above and below the median variation in the share of active population before and after the lockdown within a province, also controlling for municipality-specific dynamics, daily-shocks at the provincial level and municipal unobserved characteristics. Our results show that the intensity of the economic lockdown is associated to a statistically significant reduction in mortality by Covid-19 and, in particular, for age groups between 40-64 and older. Back of the envelope calculations indicate that 4,793 deaths were avoided, in the 26 days between April 5 to April 30, in the 3,518 municipalities which experienced a more intense lockdown. Assuming linearity, a 1 percentage point reduction in the share of active population caused a 1.32 percentage points reduction in mortality by Covid-19. We also find that the economic lockdown, as expected, led to a reduction in human mobility. Several robustness checks corroborate our empirical findings.

Purpose Proton therapy allows for more conformal dose distributions and lower organ at risk and healthy tissue doses than conventional photon-based radiotherapy, but uncertainties in the proton range currently prevent proton therapy from making full use of these advantages. Numerous developments therefore aim to reduce such range uncertainties. In this work, we quantify the benefits of reductions in range uncertainty for treatments of skull base tumors. Methods The study encompassed 10 skull base patients with clival tumors. For every patient, six treatment plans robust to setup errors of 2 mm and range errors from 0% to 5% were created. The determined metrics included the brainstem and optic chiasm normal tissue complication probability (NTCP) with the endpoints of necrosis and blindness, respectively, as well as the healthy tissue volume receiving at least 70% of the prescription dose. Results A range uncertainty reduction from the current level of 4% to a potentially achievable level of 1% reduced the probability of brainstem necrosis by up to 1.3 percentage points in the nominal scenario in which neither setup nor range errors occur and by up to 2.9 percentage points in the worst-case scenario. Such a range uncertainty reduction also reduced the optic chiasm NTCP with the endpoint of blindness by up to 0.9 percentage points in the nominal scenario and by up to 2.2 percentage points in the worst-case scenario. The decrease in the healthy tissue volume receiving at least 70% of the prescription dose ranged from -7.8 to 24.1 cc in the nominal scenario and from -3.4 to 38.4 cc in the worst-case scenario. Conclusion The benefits quantified as part of this study serve as a guideline of the OAR and healthy tissue dose benefits that range monitoring techniques may be able to achieve. Benefits were observed between all levels of range uncertainty. Even smaller range uncertainty reductions may therefore be beneficial.

We study the results of a massive nationwide correspondence experiment sending more than 83,000 fictitious applications with randomized characteristics to geographically dispersed jobs posted by 108 of the largest U.S. employers. Distinctively Black names reduce the probability of employer contact by 2.1 percentage points relative to distinctively white names. The magnitude of this racial gap in contact rates differs substantially across firms, exhibiting a between-company standard deviation of 1.9 percentage points. Despite an insignificant average gap in contact rates between male and female applicants, we find a between-company standard deviation in gender contact gaps of 2.7 percentage points, revealing that some firms favor male applicants while others favor women. Company-specific racial contact gaps are temporally and spatially persistent, and negatively correlated with firm profitability, federal contractor status, and a measure of recruiting centralization. Discrimination exhibits little geographical dispersion, but two digit industry explains roughly half of the cross-firm variation in both racial and gender contact gaps. Contact gaps are highly concentrated in particular companies, with firms in the top quintile of racial discrimination responsible for nearly half of lost contacts to Black applicants in the experiment. Controlling false discovery rates to the 5% level, 23 individual companies are found to discriminate against Black applicants. Our findings establish that systemic illegal discrimination is concentrated among a select set of large employers, many of which can be identified with high confidence using large scale inference methods.

Background\ud \ud Feedback-informed treatment (FIT) involves using computerized routine outcome monitoring technology to alert therapists to cases that are not responding well to psychotherapy, prompting them to identify and resolve obstacles to improvement. In this study, we present the first health economic evaluation of FIT, compared to usual care, to enable decision makers to judge whether this approach represents a good investment for health systems.\ud \ud \ud \ud Methods\ud \ud This randomised controlled trial included 2233 patients clustered within 77 therapists who were randomly assigned to a FIT group (n = 1176) or a usual care control group (n = 1057). Treatment response was monitored using patient-reported depression (PHQ-9) and anxiety (GAD-7) measures. Therapists in the FIT group had access to a computerized algorithm that alerted them to cases that were “not on track”, compared to normative clinical data. Health service costs included the cost of training therapists to use FIT and the cost of therapy sessions in each arm. The incremental cost-effectiveness of FIT was assessed relative to usual care, using multilevel modelling.\ud \ud \ud \ud Results\ud \ud FIT was associated with an increased probability of reliable symptomatic improvement by 8.09 percentage points (95% CI: 4.16%–12.03%) which was statistically significant. The incremental cost of FIT was £15.17 (95% CI: £6.95 to £37.29) per patient and was not statistically significant. The incremental cost-effectiveness ratio (ICER) per additional case of reliable improvement was £187.4 (95% CI: £126.7 to £501.5); this confidence interval shows that the relative cost-effectiveness is between FIT being a dominant strategy (i.e. more effective and also cost-saving) to FIT being more effective at a modest incremental cost to the health system.\ud \ud \ud \ud Conclusions\ud \ud The FIT strategy increases the probability of reliable improvement in routine clinical practice and may be associated with a small (but uncertain) incremental cost. FIT is likely to be a cost-effective strategy for mental health services.

Background Although effects on labour is one of the most tangible and attributable climate impact, our quantification of these effects is insufficient and based on weak methodologies. Partly, this gap is due to the inability to resolve different impact channels, such as changes in time allocation (labour supply) and slowdown of work (labour productivity). Explicitly resolving those in a multi-model inter-comparison framework can help to improve estimates of the effects of climate change on labour effectiveness. Methods In this empirical, multi-model study, we used a large collection of micro-survey data aggregated to subnational regions across the world to estimate new, robust global and regional temperature and wet-bulb globe temperature exposure-response functions (ERFs) for labour supply. We then assessed the uncertainty in existing labour productivity response functions and derived an augmented mean function. Finally, we combined these two dimensions of labour into a single compound metric (effective labour effects). This combined measure allowed us to estimate the effect of future climate change on both the number of hours worked and on the productivity of workers during their working hours under 1·5°C, 2·0°C, and 3·0°C of global warming. We separately analysed low-exposure (indoors or outdoors in the shade) and high-exposure (outdoor in the sun) sectors. Findings We found differentiated empirical regional and sectoral ERF's for labour supply. Current climate conditions already negatively affect labour effectiveness, particularly in tropical countries. Future climate change will reduce global total labour in the low-exposure sectors by 18 percentage points (range −48·8 to 5·3) under a scenario of 3·0°C warming (24·8 percentage points in the high-exposure sectors). The reductions will be 25·9 percentage points (–48·8 to 2·7) in Africa, 18·6 percentage points (–33·6 to 5·3) in Asia, and 10·4 percentage points (–35·0 to 2·6) in the Americas in the low-exposure sectors. These regional effects are projected to be substantially higher for labour outdoors in full sunlight compared with indoors (or outdoors in the shade) with the average reductions in total labour projected to be 32·8 percentage points (–66·3 to 1·6) in Africa, 25·0 percentage points (–66·3 to 7·0) in Asia, and 16·7 percentage points (–45·5 to 4·4) in the Americas. Interpretation Both labour supply and productivity are projected to decrease under future climate change in most parts of the world, and particularly in tropical regions. Parts of sub-Saharan Africa, south Asia, and southeast Asia are at highest risk under future warming scenarios. The heterogeneous regional response functions suggest that it is necessary to move away from one-size-fits-all response functions to investigate the climate effect on labour. Our findings imply income and distributional consequences in terms of increased inequality and poverty, especially in low-income countries, where the labour effects are projected to be high.

Icelandic Cattle is a local dairy cattle breed in Iceland. With about 26,000 breeding females, it is by far the largest among the indigenous Nordic cattle breeds. The objective of this study was to investigate the feasibility of genomic selection in Icelandic Cattle. Pedigree-based best linear unbiased prediction (PBLUP) and single-step genomic best linear unbiased prediction (ssGBLUP) were compared. Accuracy, bias, and dispersion of estimated breeding values (EBV) for milk yield (MY), fat yield (FY), protein yield (PY), and somatic cell score (SCS) were estimated in a cross validation-based design. Accuracy (r^) was estimated by the correlation between EBV and corrected phenotype in a validation set. The accuracy (r^) of predictions using ssGBLUP increased by 13, 23, 19, and 20 percentage points for MY, FY, PY, and SCS for genotyped animals, compared with PBLUP. The accuracy of nongenotyped animals was not improved for MY and PY, but increased by 0.9 and 3.5 percentage points for FY and SCS. We used the linear regression (LR) method to quantify relative improvements in accuracy, bias (Δ^), and dispersion (b^) of EBV. Using the LR method, the relative improvements in accuracy of validation from PBLUP to ssGBLUP were 43%, 60%, 50%, and 48% for genotyped animals for MY, FY, PY, and SCS. Single-step GBLUP EBV were less underestimated (Δ^), and less overdispersed (b^) than PBLUP EBV for FY and PY. Pedigree-based BLUP EBV were close to unbiased for MY and SCS. Single-step GBLUP underestimated MY EBV but overestimated SCS EBV. Based on the average accuracy of 0.45 for ssGBLUP EBV obtained in this study, selection intensities according to the breeding scheme of Icelandic Cattle, and assuming a generation interval of 2.0 yr for sires of bulls, sires of dams and dams of bulls, genetic gain in Icelandic Cattle could be increased by about 50% relative to the current breeding scheme.

We evaluate the impacts that health supporting environments have on residents’ health and well-being. Using a stratified multi-stage sampling method, we select a sample of 12,360 permanent adult residents aged 15–69, and collect information on their health literacy level, as well as their demographic background and health. This individual level data is then merged with the administrative health supporting environment data. More than two thirds of residents self-reported having good/excellent health, and the percent of adults living in communities with healthy parks, healthy trails, and healthy huts in their community is 23 percent, 43 percent, and 25 percent, respectively. Controlling for a series of confounding factors at the community and individual levels, we find that healthy parks and healthy trails are positively correlated with self-reported health, which increases the probability of self-reporting good health by 2.0 percentage points (p &lt, 0.10) and 6.0 percentage points (p &lt, 0.01), respectively. Access to healthy huts is negatively associated with self-reported health, decreasing the probability of self-reporting good health by 5.0 percentage points (p &lt, 0.01). Health literacy plays a role in moderating the effect of health parks, and a positive effect is more likely to be observed among adults with lower health literacy. Health supporting environments may play a role in reducing the likelihood of undiagnosed diseases and changing residents’ lifestyles, which promotes the health and well-being of residents, especially among those with inadequate health literacy.

Container management platform needs to allocate resources reasonably for containers running different services. At present, the container mainly allocates resources through the responsive elastic scaling scheme, and the speed of resource adjustment is difficult to guarantee, while the predictive elastic scaling scheme has not reached the actual production requirements in terms of speed and accuracy. In this paper, after studying the existing responsive elastic expansion scheme and predictive elastic expansion scheme, a container elastic expansion scheme based on combined prediction model is proposed. Combining the advantages of ARIMA model and SVM model in short-term time series prediction, a combined prediction algorithm model based on ARIMA and SVM is designed. Compared with the existing prediction algorithm model, the accuracy of the combined prediction algorithm is improved by about 9 percentage points, and the accuracy can reach 92.03%, which provides a strong support for the realization of the elastic expansion of the predictive container.

How should schools assign students to more rigorous math courses so as best to help their academic outcomes? We identify several hundred California middle schools that used 7th-grade test scores to place students into 8th-grade algebra courses and use a regression discontinuity design to estimate average impacts and heterogeneity across schools. Enrolling in 8th-grade algebra boosts students’ enrollment in advanced math in ninth grade by 30 percentage points and eleventh grade by 16 percentage points. Math scores in tenth grade rise by 0.05 standard deviations. Women, students of color, and English-language learners benefit disproportionately from placement into early algebra. Importantly, the benefits of 8th-grade algebra are substantially larger in schools that set their eligibility threshold higher in the baseline achievement distribution. This suggests a potential tradeoff between increased access and rates of subsequent math success.

Deep neural network approaches to speaker verification have proven successful, but typical computational requirements of State-Of-The-Art (SOTA) systems make them unsuited for embedded applications. In this work, we present a two-stage model architecture orders of magnitude smaller than common solutions (237.5K learning parameters, 11.5MFLOPS) reaching a competitive result of 3.31% Equal Error Rate (EER) on the well established VoxCeleb1 verification test set. We demonstrate the possibility of running our solution on small devices typical of IoT systems such as the Raspberry Pi 3B with a latency smaller than 200ms on a 5s long utterance. Additionally, we evaluate our model on the acoustically challenging VOiCES corpus. We report a limited increase in EER of 2.6 percentage points with respect to the best scoring model of the 2019 VOiCES from a Distance Challenge, against a reduction of 25.6 times in the number of learning parameters.

BackgroundThe COVID-19 pandemic has created global shortages of personal protective equipment (PPE) such as medical exam gloves, forcing healthcare workers to either forgo or reuse PPE to keep themselves and patients safe from infection. In severely resource-constrained situations, limited cycles of disinfection and extended use of gloves is recommended by the U.S. Centers for Disease Control and Prevention (CDC) to conserve supplies. However, these guidelines are based on limited evidence.MethodsSerial cycles of hand hygiene were performed on gloved hands using alcohol-based hand rub (ABHR) (six and ten cycles), 0.1% sodium hypochlorite (bleach) solution (ten cycles), or soap and water (ten cycles) on three types of latex and three types of nitrile medical exam gloves, purchased in the United States and India. A modified FDA-approved water-leak test was performed to evaluate glove integrity after repeated applications of these disinfecting agents. 80 gloves per disinfectant-glove type combination were tested. Within each glove type the proportion of gloves that failed the water-leak test for each disinfectant was compared to that of the control using a non-inferiority design with a non-inferiority margin of five percentage points. Results were also aggregated by glove material, and combined for overall results.FindingsWhen aggregated by glove material, the dilute bleach exposure demonstrated the lowest difference in proportion failed between treatment and control arms: −2.5 percentage points (95% CI: −5.3 to 0.3) for nitrile, 0.6 percentage points (95% CI: −2.6 to 3.8) for non-powdered latex. For US-purchased gloves tested with six and ten applications of ABHR, the mean difference in failure risk between treatment and control gloves was within the prespecified non-inferiority margin of five percentage points or less, though some findings were inconclusive because confidence intervals extended beyond the non-inferiority margin. The aggregated difference in failure risk between treatment and control gloves was 3.5 percentage points (0.6 to 6.4) for soap and water, and 2.3 percentage points (−0.5 to 5.0) and 5.0 percentage points (1.8 to 8.2) for 10 and 6 applications of ABHR, respectively. The majority of leaks occurred in the interdigital webs (35%) and on the fingers (34%).ConclusionCurrent guidelines do not recommend extended use of a single-use PPE under normal supply conditions. However, our findings indicate that some combinations of glove types and disinfection methods may allow for extended use under crisis conditions. We found that ten applications of dilute bleach solution have the least impact on glove integrity, compared to repeated applications of ABHR and soap and water. However, the majority of glove and exposure combinations were inconclusive with respect to non-inferiority with a 5 percentage point non-inferiority margin. Testing specific glove and disinfectant combinations may be worthwhile for settings facing glove shortages during which extended use is necessary. The modified water-leak testing method used here is a low-resource method that could easily be reproduced in different contexts.

Importance: The association between age-related hearing loss (ARHL) and physical or cognitive frailty has been poorly explored. These associations could define new perspectives for delaying frailty-related processes in older age. Objective: To examine whether peripheral ARHL and age-related central auditory processing disorder (CAPD) are independently associated with physical or cognitive frailty. Design, Setting, and Participants: This cross-sectional study analyzed registry data from December 31, 2014, on 1929 older (≥65 years) participants of the Salus in Apulia Study (Southern Italy) who underwent audiologic, physical, and neuropsychological assessment. Data analysis was performed from December 12, 2019, to January 4, 2020. Main Outcomes and Measures: Prevalence of peripheral ARHL in older individuals with physical and/or cognitive frailty and those without frailty assessed using the Fried criteria (physical) and the Mini-Mental State Examination (cognitive). Multivariable logistic regression models were used to assess associations of audiologic variables with frailty phenotype. Results: Data from 1929 participants (mean [SD] age, 73.6 [6.3] years; 974 male [50.5%]) were eligible for the analyses. The prevalence of peripheral ARHL was higher in the physical frailty group (96 [26.6%]) than in the nonfrail group (329 [21.0%]) (difference, 5.61 percentage points; 95% CI, 0.63-10.59 percentage points) and in the cognitive frailty group (40 [38.8%]) than in the nonfrail group (385 [21.1%]) (difference, 17.75 percentage points; 95% CI, 8.2-27.3 percentage points). Age-related CAPD was more prevalent in the physical frailty group (62 [17.2%]) than in the nonfrail group (219 [14.0%]) (difference, 3.21 percentage points; 95% CI, -1.04 to 7.46 percentage points) and in the cognitive frailty group (28 [27.2%]) than in the nonfrail group (253 [13.9%]) (difference, 13.33 percentage points; 95% CI, 4.10-22.21 percentage points). In the multivariable models, age-related CAPD was associated with cognitive frailty in the fully adjusted model (odds ratio [OR], 1.889; 95% CI, 1.094-3.311). There was also an inverse association between the unitary increase in Synthetic Sentence Identification With the Ipsilateral Competitive Message scores, indicating a lower likelihood of this disorder, and cognitive frailty (OR, 0.989; 95% CI, 0.988-0.999). Peripheral ARHL was associated with cognitive frailty only in the partially adjusted model (OR, 1.725; 95% CI, 1.008-2.937). Conclusions and Relevance: In this cross-sectional study of 1929 participants, age-related CAPD was independently associated with cognitive frailty. Whether the management of ARHL may help prevent the development of different frailty phenotypes or improve their clinical consequences should be addressed in longitudinal studies and, eventually, well-designed randomized clinical trials..

Purpose To examine the association between the introduction of a school-based health center (SBHC) and high school graduation rates. Methods We use school-level longitudinal data from Colorado that combines data on the opening of SBHCs in high schools with 4-year high school graduation rates overall and by gender between 2000 and 2018. The analytic sample consists of high schools without an SBHC in 2000 (n = 132). We compare high schools that opened SBHCs over the period to those that did not and run school-level panel fixed effects models to assess the relationship between opening an SBHC and change in high school graduation rates. Results Schools that subsequently opened SBHCs had larger minority populations and lower average graduation rates in 2000. Opening an SBHC was associated with a 4.1 percentage point increase in the overall graduation rate (p = .077). The gender-stratified analyses indicate young men’s graduation rates were most sensitive to the presence of an SBHC, increasing 4.8 percentage points (p = .051), compared to young women’s graduation rates increasing 3.0 percentage points (p = .163). Conclusions Our findings suggest that the benefits of SBHC access may extend beyond health-specific outcomes to graduation rates.

Purpose The present study evaluated whether Pennsylvania's 2016 law requiring a retail license for the sale of e-cigarettes was associated with adolescent e-cigarette use. Methods Data were drawn from the 2015–2017 Youth Risk Behavior Surveillance System. We examined the prepolicy and postpolicy change in e-cigarette use for the state with the retail licensing requirement (Pennsylvania) compared with control states (New York and Virginia). Results Results showed that e-cigarette licensing policy was significantly associated with e-cigarette use. E-cigarette use among Pennsylvania adolescents reduced by 5.2 percentage points in 2017 when compared with New York adolescents, and a corresponding 21.6% decrease from its baseline prevalence level in 2015. Similarly, there was a 7.4 percentage point decrease in e-cigarette use in Pennsylvania when compared with Virginia (30.7% relative decrease from the baseline prevalence). Conclusions An e-cigarette retail licensing requirement may be a useful policy tool in reducing e-cigarette use among adolescents.

Since 1988, 27 states have introduced No Pass, No Drive laws, which tie a teenager’s ability to receive and maintain a driver’s license to various school-related outcomes -- most commonly, enrollment and attendance. Truancy-Based No Pass, No Drive policies target only attendance – teens that fail to meet a minimum attendance requirement lose their driver’s license. However, these policies allow students to drop out of school without facing this penalty. These policies increase the annual dropout rate by between 32 and 45 percent (1.4 to 2 percentage points). Enrollment-Based No Pass, No Drive policies, the largest group of policies, which target both enrollment and attendance, have negligible effects on dropout rates, but decrease the Averaged Freshman Graduation Rate (AFGR) by more than one percentage point. However, this lower graduation rate stems from students delaying their dropout decision by up to two years. As a result, these students are retained in the ninth and tenth grades, increasing ninth grade enrollment by 2.8 percent relative to eighth grade enrollment the year prior; this causes an artificial reduction in the graduation rate, rather than a reduction in the true likelihood that a student will graduate.

This paper investigates how the ability to innovate affects firms' decisions to operate informally and the aggregate consequences of their sectoral choice. I embed a sectoral choice model, where firms choose to operate in the formal or informal economy, into a richer general equilibrium environment to analyze the aggregate effects of firm-level decisions in response to government taxation. I calibrate the model and conduct simulations to quantify the impacts on the aggregate economy. I find that a change in tax rates from 50% to 60% leads to a 20.9% reduction in the size of the formal sector. This change is accompanied by a 0.07 percentage point reduction in TFP growth per year. Given that countries like Mali, Mexico, and Sri Lanka impose total tax rates near 50%, these findings have significant and applicable policy implications across a broad range of lesser developed countries. Even at lower tax rates, for instance 10%, a 10% increase, decreases the size of the formal sector by more than 7.7%.

Explaining cross-country differences in current accounts is difficult. While pay-as-you-go pensions reduce the need to save for retirement, contributions to capital-funded pensions are saved for future consumption. An overlapping-generations analysis shows that capital-funded pensions increase net foreign assets holdings. With a multi-pillar system whose capital-funded part accounts for 18% of pensions, the Austrian current account balance would be 1 percentage point of gross domestic product (GDP) higher than with pure pay-as-you-go pensions in 20 years. By comparison, the Austrian current account surplus averages 1.8% of GDP. Empirically, I find that the current account of high-income countries increases with the coverage and replacement rates of capital-funded pensions.

The paper tests the idea that major demographic shifts can affect housing prices. We first build an overlapping generation model and analytically solve for the equilibrium price of the asset. The model predicts that economies with a higher fraction of old people in the overall population have lower house prices. We empirically test this hypothesis using data on house prices and demographic variables from the Organization for Economic Co-operation and Development (OECD). We find that if population growth increases by one percentage point, house price growth increases by 1.4 percentage points.

This study examines the association of unemployment variation with intimate partner violence using representative data from thirty-one developing countries, from 2005 to 2016. It finds that a 1 percent increase in the male unemployment rate is associated with an increase in the incidence of physical violence against women by 0.50 percentage points, or 2.75 percent. This is consistent with financial and psychological stress generated by unemployment. Female unemployment rates have the opposite effect, a 1 percent decrease being associated with an increase in the probability of victimization of 0.52 percentage points, or 2.87 percent. That an improvement in women's employment opportunities is associated with increased violence is consistent with male backlash. The study finds that this pattern of behaviors emerges entirely from countries in which women have more limited access to divorce than men.

Objective We analyze the effect of neighborhood limited English proficiency on third graders' reading achievement in public elementary schools in the U.S. state of Georgia. Methods Neighborhood limited English proficiency and school‐related outcomes are more likely to be endogenously determined at the neighborhood level. We correct for this endogeneity concern using foreign‐born population as the instrument for neighborhood limited English proficiency. Results Our results show that a 1 percentage point increase in limited English proficient households in the census tract reduced proficient learners and above on third‐grade reading by 0.20 percentage point in the U.S. state of Georgia. Conclusion Improving neighborhood characteristics may help improve students' achievement.

OBJECTIVES: Medicaid and Children’s Health Insurance Program plans publicly report quality measures, including follow-up care after psychiatric hospitalization. We aimed to understand failure to meet this measure, including measurement definitions and enrollee characteristics, while investigating how follow-up affects subsequent psychiatric hospitalizations and emergency department (ED) visits. METHODS: Administrative data representing Alabama’s Children’s Health Insurance Program from 2013 to 2016 were used to identify qualifying psychiatric hospitalizations and follow-up care with a mental health provider within 7 to 30 days of discharge. Using relaxed measure definitions, follow-up care was extended to include visits at 45 to 60 days and visits to a primary care provider. Logit regressions estimated enrollee characteristics associated with follow-up care and, separately, the likelihood of subsequent psychiatric hospitalizations and/or ED visits within 30, 60, and 120 days. RESULTS: We observed 1072 psychiatric hospitalizations during the study period. Of these, 356 (33.2%) received follow-up within 7 days and 566 (52.8%) received it within 30 days. Relaxed measure definitions captured minimal additional follow-up visits. The likelihood of follow-up was lower for both 7 days (−18 percentage points; 95% confidence interval [CI] −26 to −10 percentage points) and 30 days (−26 percentage points; 95% CI −35 to −17 percentage points) regarding hospitalization stays of ≥8 days. Meeting the measure reduced the likelihood of subsequent psychiatric hospitalizations within 60 days by 3 percentage points (95% CI −6 to −1 percentage point). CONCLUSIONS: Among children, receipt of timely follow-up care after a psychiatric hospitalization is low and not sensitive to measurement definitions. Follow-up care may reduce the need for future psychiatric hospitalizations and/or ED visits.

This paper analyses the unemployment hysteresis in the Czech Republic on data from 1999 to 2016. The hysteresis is modelled by allowing for the impact of cyclical unemployment on the non-accelerating inflation rate of unemployment. Models are estimated using the Bayesian approach and provide robust evidence in favour of the hysteresis. The estimates imply that in response to an increase in the cyclical unemployment of 1 percentage point, the non-accelerating inflation rate of unemployment increases by 0.18 percentage points.

P-wave first-motion polarity is the most useful information in determining the focal mechanisms of earthquakes, particularly for smaller earthquakes. Algorithms have been developed to automatically determine P-wave first-motion polarity, but the performance level of the conventional algorithms remains lower than that of human experts. In this study, we develop a model of the convolutional neural networks (CNNs) to determine the P-wave first-motion polarity of observed seismic waveforms under the condition that P-wave arrival times determined by human experts are known in advance. In training and testing the CNN model, we use about 130 thousand 250 Hz and about 40 thousand 100 Hz waveform data observed in the San-in and the northern Kinki regions, western Japan, where three to four times larger number of waveform data were obtained in the former region than in the latter. First, we train the CNN models using 250 Hz and 100 Hz waveform data, respectively, from both regions. The accuracies of the CNN models are 97.9% for the 250 Hz data and 95.4% for the 100 Hz data. Next, to examine the regional dependence, we divide the waveform data sets according to the observation region, and then we train new CNN models with the data from one region and test them using the data from the other region. We find that the accuracy is generally high ($${ \gtrsim }$$≳ 95%) and the regional dependence is within about 2%. This suggests that there is almost no need to retrain the CNN model by regions. We also find that the accuracy is significantly lower when the number of training data is less than 10 thousand, and that the performance of the CNN models is a few percentage points higher when using 250 Hz data compared to 100 Hz data. Distribution maps, on which polarities determined by human experts and the CNN models are plotted, suggest that the performance of the CNN models is better than that of human experts.

Remote sensing is considered a valuable tool for monitoring the impacts of global change on vegetation species composition, condition and distribution. Multi-season imagery has been shown to improve the classification of vegetation communities though the contribution of the winter season in multi-seasonal classifications remains to be assessed. The capability of multi-season images of RapidEye, a new generation high spatial resolution (5 m) space-borne sensor containing an additional red-edge band, was evaluated for the classification of several wetland and dryland communities. RapidEye images were obtained for four seasons (winter, spring, summer and autumn) between 2011 and 2012 for a subtropical coastal region of South Africa. The separability of nine wetland and dryland communities was assessed for each season using the Partial Least Square Random Forest (PLS-RF) algorithm. The four-seasons approach yielded a higher overall classification accuracy (OA = 86 ± 2.8%) when compared to using any single-season classification. The highest single-season accuracies were obtained in spring (80 ± 2.9%), summer (80 ± 3.1%), and autumn (79 ± 3.4%) compared to the winter (66 ± 3.1%). A three-season combination of autumn, winter and spring yielded the highest average OA (86 ± 3.1%), maximised the user’s accuracies and minimised the number of comparable pairs confused. The inclusion of indices in the classification scenarios showed a minor (±1 percentage points) difference in the average overall and user’s accuracies compared to the classification results where only bands were used. The red-edge band of RapidEye increased the overall and average user’s accuracy for most of the scenarios by 2–6 percentage points and thus contributed to the separability of communities which are dominated by evergreen tree species.

OBJECTIVE: To measure discordance between aggregate estimates of means‐tested coverage from the American Community Survey (ACS) and administrative counts and examine the association of discordance with ACA Medicaid expansion. DATA SOURCES: 2010‐2016 ACS and counts of Medicaid and Children's Health Insurance Program enrollment from the Centers for Medicare & Medicaid Services. STUDY DESIGN: State‐by‐year counts of means‐tested coverage from the ACS were compared to administrative counts using percentage differences. Discordance was compared for states that did and did not adopt expansion using difference‐in‐differences. We then contrasted the effect of expansion on means‐tested coverage estimated from the ACS with results from administrative data. DATA COLLECTION/EXTRACTION: Survey and administrative data. PRINCIPAL FINDINGS: One year before expansion there was a 0.8 and 4 percent overcount in expansion and nonexpansion states, respectively. By 2016, there was a 10.64 percent undercount in expansion states vs a 0.02 percent undercount in nonexpansion states. The ACS suggests that expansion increased means‐tested coverage in the full population by three percentage points, relative to five percentage points suggested by administrative records. CONCLUSIONS: Discordance between the ACS and administrative records has increased over time. The ACS underestimates the impact of Medicaid expansion, relative to administrative counts.

BACKGROUND Prices for total joint arthroplasty vary widely. Insurers have experimented with reference-based benefit designs (reference pricing) to control costs by setting a contribution limit that covers lower-priced facilities but necessitates higher out-of-pocket payments at higher-priced facilities. The purpose of this study was to evaluate the impact of reference pricing on the cost and quality of care for total joint arthroplasty. METHODS The California Public Employees' Retirement System (CalPERS) implemented reference pricing for total joint arthroplasty in January 2011. We obtained data on 2,023 CalPERS patients who underwent total joint arthroplasty from January 2009 to December 2013 and comparison group data on 8,024 non-CalPERS patients from the same time period. Trends in 9 cost and quality-related metrics were compared between the CalPERS group and the comparison group: patient choice of a lower-priced hospital, insurer payment, consumer payment, 90-day complication rate, 90-day readmission rate, annual surgical volume of the chosen hospital, length of stay, travel distance, and rate of discharge to home. The impact of reference pricing was estimated with difference-in-differences multivariable regressions, adjusting for covariates. RESULTS An increase of 19 percentage points (95% confidence interval [CI], 13.0 to 25.6 percentage points; p < 0.01) in the selection of lower-priced hospitals was attributable to reference pricing, with a concurrent mean savings for the insurer of $5,067 (95% CI, $2,315 to $7,819; p < 0.01) and an increase in the mean patient out-of-pocket payment of $1,991 (95% CI, $1,053 to $2,929; p < 0.01). No significant change in any quality indicator was attributable to reference pricing, with the exception of an 8% reduction (95% CI, 3.3% to 12.7% reduction; p < 0.01) in the length of stay for hip replacement. CONCLUSIONS Reference pricing motivates patients to choose lower-priced hospitals for total joint arthroplasty, with no measurable adverse impact on quality. Reference pricing represents a viable strategy in the shift toward value-based care.

This study uses a discontinuity in the Swedish tertiary eligibility requirement to estimate the probability of enrolling in tertiary education, and the payoff thereof. Regression discontinuity results, show that achieving tertiary eligibility in upper-secondary education, increases the probability of enrolling in tertiary education by around 10–15 and 7 percentage points for students who enrolled on an academic and vocational track, respectively. For academic students, this implies 5% higher earnings for men, while for women it increases the probability of having positive incomes by 2%. Thus, academic students at the margin of tertiary education receive a substantial tertiary education payoff. (JEL I21, I26, I28).

This paper addresses measurement error in food security in the USA. In particular, it uses a Bayesian 4-parameter IRT model to look at the likelihood of over- or under-reporting of the conditions that comprise the food security module (FSM), the data collection administered in many US surveys to assess and monitor food insecurity. While this model’s parameters are only partially identified, we learn about the likely values of these parameters by using a Bayesian framework. My results suggest significant under-reporting of more severe food security items, particularly those in the child module. I find no evidence of over-reporting of food hardships. I show that, under conservative assumptions, this model predicts food insecurity prevalence between 1 and 3 percentage points higher than current estimates, or roughly 4 to 15 percent of prevalence, for the years 2007–2015. Results suggest much larger increases—on the order of 50 percent of prevalence—for very low food security among households that were screened into the food security module.

High nominal interest rates are associated with high deposit spreads, which is consistent with a model where banks have monopoly power and currency and deposits are substitutes. Therefore, higher interest rates raise the implicit price of banking services, increase bank profits and attract entry into the banking sector. Taking these effects into account, a one percentage point increase in inflation has a welfare cost of 0.083% of GDP, 6.7 times higher than traditional estimates.

Background. There is a lack of data on alcohol consumption over time. This study characterizes the long-term drinking patterns of people with lifetime alcohol use disorders who have engaged in treatment or informal care. Methods. We developed multinomial logit models using the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) to estimate short-term transition probabilities (TPs) among the 4 World Health Organization drinking risk levels (low, medium, high, and very high risk) and abstinence by age, sex, and race/ethnicity. We applied an optimization algorithm to convert 3-year TPs from NESARC to 1-year TPs, then used simulated annealing to calibrate TPs to a propensity-scored matched set of participants derived from a separate 16-year study of alcohol consumption. We validated the resulting long-term TPs using NESARC-III, a cross-sectional study conducted on a different cohort. Results. Across 24 demographic groups, the 1-year probability of remaining in the same state averaged 0.93, 0.81, 0.49, 0.51, and 0.63 for abstinent, low, medium, high, and very high-risk states, respectively. After calibration to the 16-year study data ( N = 420), resulting TPs produced state distributions that hit the calibration target. We find that the abstinent or low-risk states are very stable, and the annual probability of leaving the very high-risk state increases by about 20 percentage points beyond 8 years. Limitations. TPs for some demographic groups had small cell sizes. The data used to calibrate long-term TPs are based on a geographically narrow study. Conclusions. This study is the first to characterize long-term drinking patterns by combining short-term representative data with long-term data on drinking behaviors. Current research is using these patterns to estimate the long-term cost effectiveness of alcohol treatment.

Employee representatives in firms are a potentially key but not yet studied source of the impact of unions and works councils. Their actions can shape multiple drivers of firm performance, including collective bargaining, strikes, and training. This article examines the impact of union representative mandates by exploiting legal membership thresholds present in many countries. In the case of Portugal, which we examine here, while firms employing up to forty‐nine union members are required to have one union representative; this increases to two (three) union reps for firms with fifty to ninety‐nine (100–199) union members. Drawing on matched employer–employee data on the unionized sector and regression discontinuity methods, we find that a one percentage point increase in the legal union representative/members ratio leads to an increase in firm performance of at least 7 percent. This result generally holds across multiple dimensions of firm performance and appears to be driven by increased training. However, we find no effects of union representatives on firm‐level wages, given the predominance of sectoral collective bargaining.

While education levels of women have increased dramatically relative to men, women are still greatly underrepresented in Science, Technology, Engineering, and Mathematics (STEM) college programmes. We use unique data on preference rankings for all secondary school students who apply for college in Ireland and detailed information on school subjects and grades to decompose the sources of the gender gap in STEM. We find that, of the 22 percentage points raw gap, about 13 percentage points is explained by differential subject choices and grades in secondary school. Subject choices are more important than grades – we estimate male comparative advantage in STEM (as measured by subject grades) explains about 3 percentage points of the gender gap. Additionally, differences in overall achievement between girls and boys have a negligible effect. Strikingly, there remains a gender gap of 9 percentage points even for persons who have identical preparation at the end of secondary schooling (in terms of both subjects studied and grades achieved); however, this gap is only 4 percentage points for STEM-ready students. We find that gender gaps are smaller among high-achieving students and for students who go to school in more affluent areas. There is no gender gap in science (the large gaps are in engineering and technology), and we also find a smaller gender gap when we include nursing degrees in STEM, showing that the definition of STEM used is an important determinant of the conclusions reached.

Statistical institutes are considering web scraping of online prices of consumer goods as a feasible alternative to scanner data. The lack of transaction data generates the question whether web scraped data are suited for price index calculation. This article investigates this question by comparing price indices based on web scraped and scanner data for clothing and footwear in the same webshop. Scanner data and web scraped prices are often equal, with the latter being slightly higher on average. Numbers of web scraped product prices and products sold show remarkably high correlations. Given the high churn rates of clothing products, a multilateral method (Geary-Khamis) was used to calculate price indices. For 16 product categories, the indices show small overall differences between the two data sources, with year on year indices differing only by 0.3 percentage point at COICOP level (men’s and women's clothing). It remains to be investigated whether such promising results for web scraped data will also be found for other retailers., Chessa Antonio G., Griffioen Robert. Comparing Price Indices of Clothing and Footwear for Scanner Data and Web Scraped Data. In: Economie et Statistique / Economics and Statistics, n°509, 2019. Big Data and Statistics. Part 2. Big Data in the Consumer Price Index. pp. 49-68.