Groen, J. J. J. and M. B. Nattinger (2020). Alternative Indicators for Chinese Economic Activity Using Sparse PLS Regression. Economic Policy Review (forthcoming).
In this paper, we investigate underlying economic growth in China. There is substantial evidence that China's official GDP number is artificially smooth. To better model cyclical trends, we apply sparse partial least squares (PLS) regressions. In these regressions, we target Chinese imports, as reported by their major trade partners; retail sales; and an industry-level diffusion index. We use on our RHS a large set of variables spanning finance, trade, production and prices. We find that our factors identify realistic Chinese macroeconomic business cycles. Then, we decompose movements in our underlying factors into global demand, Chinese credit, and Chinese monetary policy components. Our analysis suggests that global demand was the main driver of the Chinese economy from the recession until mid-2017, when strong contractions in credit, magnified by an underwhelming monetary policy response, caused an economic downturn.
Neuner, J., A. B. Nattinger, T. W. Yen, E. L. McGinley, M. B. Nattinger, L. E. Pezzin. Temporal trends and regional variation in the utilization of low-value breast cancer care: has the Choosing Wisely campaign made a difference? Breast Cancer Res Treat 2019 Jul;176(1):205-215
In this paper, we attempt to answer the following question: Has the Choosing Wisely campaign been effective at preventing the overuse of ineffective breast cancer treatments? We identified annual cohorts of women with incident, early-stage breast cancer and estimated the prevalence of four initial treatment and six surveillance metrics deemed as low-value breast cancer care using MarketScan Commercial Claims and Medicare Supplemental and Coordination of Benefits databases. We implemented multivariate logistic regressions to estimate temporal trends and regional variation in the use of the low-value methods, and focused on the year of the Choosing Wisely campaign's inception. Unfortunately, we found that the Choosing Wisely campaign was ineffective at preventing the overuse of low-value treatments; fortunately, use of these methods is falling over time for other reasons.
Kong, A. L., A. B. Nattinger, E. L. McGinley, M. B. Nattinger, L. E. Pezzin. The influence of socioeconomic status, tumor characteristics and patterns of breast cancer care on breast cancer specific survival among elderly women. Journal of Clinical Oncology 2018 36:15 suppl, 6557-6557
In this published abstract of forthcoming work, we investigate the relationships between patient demographic and socieoeconomic status, tumor characteristics, initial and follow-up breast cancer care and 3-year breast cancer mortality among a population-based cohort of elderly women with incident breast cancer. We identified women with newly diagnosed breast cancer in 2006-2009 from the Surveillance and Epidemiology End Result study linked with Medicare claims, and applied a Classification and Regression Tree (CART) model to 15 individual indicators of neo-adjuvant and adjuvant breast cancer treatment, tumor characteristics, and patient demographic and SES variables to identify patterns (i.e. combinations of variables) with the greatest discriminant value in predicting 3-year mortality by cause of death. We found nineteen unique patterns that were identified as best discriminating 3-year mortality by cause of death. The single best discriminator between high and low breast cancer mortality was found by CART to be the number of positive nodes, followed by the use of radiation therapy and tumor stage. The patient's socioeconomic status was a discriminant factor in four out of the ten patterns associated with high breast cancer mortality.