"Statistician Senior (underfill Intermediate)"

A physician-health policy researcher in the Department of Internal Medicine, Division of General Medicine, seeks a full-time Statistician Senior (underfill to Intermediate) to join a dynamic, interdisciplinary research team. Our team performs health economics research to better understand and develop policies to lower health care spending and improve healthcare delivery and patient outcomes. Core topic areas include the effects of corporate ownership and market structure on system outcomes and how payment reform is reshaping U.S. primary care. We use large national datasets and advanced statistical (econometric) methods to generate evidence that informs health policy and clinical practice.

The successful candidate will manage, clean, and analyze administrative datasets (e.g., Medicare claims, prescribing, and corporate ownership data); write reproducible code in Stata, SAS, R, or Python; and contribute to study design, statistical modeling, and publication. We are seeking a statistician who is curious, detail-oriented, and motivated by policy-relevant questions, able to work both independently and collaboratively across institutions.

The position is based at the University of Michigan (U-M) and will involve close collaboration with Brown University Center for Advancing Health Policy through Research (CAHPR), led by a health economist faculty PI. The candidate will be financially supported by both U-M and Brown University (approximately 50/50) but will hold an employment relationship only at U-M. The candidate will have the opportunity for deep collaborations across Michigan Medicine, U-Ms Institute for Healthcare Policy and Innovation, and Brown University CAHPR, offering a rich environment for interdisciplinary research with leading economists and health policy scholars. The position offers outstanding professional growth, including opportunities for skill development, mentorship, and authorship on peer-reviewed papers.

1. Data management

Supporting Actions: Clean, organize, and link large administrative datasets (e.g., Medicare claims, prescribing data, and corporate ownership data) from multiple sources. Develop and maintain reproducible data pipelines and documentation to ensure transparency and data integrity. Conduct quality assurance procedures, including audits, validation checks, and version control. Assess data quality and completeness, address data limitations, and prepare analytic files for use by research teams. Ensure all datasets meet standards for reproducibility and documentation required for publication and data sharing.

2. Statistical analysis and visualization

Supporting Actions: Conduct descriptive and inferential analyses to answer policy-relevant research questions. Apply econometric and causal-inference methods (e.g., difference-in-differences, instrumental variables, propensity-score methods) and selected machine-learning algorithms. Write reproducible code in Stata, R, SAS, or Python; summarize and interpret results for manuscripts, presentations, and grant proposals; and contribute to the development of statistical models and study design. Generate clear, publication-quality visualizations and tables that communicate analytic findings to technical and policy audiences.

3. Assist with writing papers and grant proposals

Supporting Actions: Draft and refine statistical methods and results sections for manuscripts and grant proposals. Interpret and summarize analytic findings for publication and presentation. Prepare reports for internal and external review committees. Participate in manuscript development for peer-reviewed journals and proposal writing, particularly those sections related to study design, analytic approach, and data sources. Provide pre-award analytic support to projects, such as conducting preliminary analyses and synthesizing results. Engage in departmental and cross-institution research seminars and workshops. Co-authorship is available for substantive contributions.

4. Supervise and/or mentor trainees

Supporting Actions: Supervise and/or mentor students and trainees, consult on methodological and statistical issues, and provide analytic and technical support as needed. Provide guidance on study design, analytic planning (including power calculations, model selection, and sample size estimation), and preparing presentations and publication-quality manuscripts. Collaborate closely with faculty and analysts to ensure analytic accuracy and reproducibility across projects.

Senior Level

• Masters degree in relevant field, per below.
• Three or more years of experience in data management and statistical analysis

Intermediate Level

• Bachelors degree in relevant field, per below.
• One or more years of experience in data management and statistical analysis.

Both

• Degree in statistics, biostatistics, economics, health services research, public health, public policy, data science, computer science, or a related field focused on data analysis and interpretation.
• Proficiency with one or more statistical software packages (Stata, R, Python, or SAS).
• Demonstrated ability to apply regression and descriptive analytic techniques to quantitative data.
• Strong organizational, analytical, and written communication skills, with close attention to detail.
• Ability to work independently and collaboratively across multiple projects and deadlines

Preferred Qualifications

• Experience analyzing and managing large, complex datasets, including health care claims (Medicare and/or commercial) or electronic health record data
• Experience applying causal-inference and econometric methods (e.g., difference-in-differences, instrumental variables, propensity score methods).
• Experience with machine learning algorithms or predictive modeling.
• Experience preparing analytic code, documentation, and supporting materials for publication.
• Ability to produce clear analytical documentation and visual presentations of statistical results (charts, tables, and other visual aids).
• Familiarity with longitudinal data analysis, methods for handling missing data, and resampling techniques such as bootstrapping.
• Knowledge of the U.S. health care system and health policy context.
• Training or experience mentoring students or junior analysts.

This position is based at the North Campus Research Complex (NCRC) in Ann Arbor and offers a hybrid work schedule with regular on-site presence expected. Fully remote arrangements are not available.

This position may be underfilled at a lower classification depending on the qualifications of the selected candidate.

Two Current Projects:

1. Impact of corporate consolidation on primary care

Insurers are targeting primary care practices serving older adults in Medicare Advantage (MA), acquiring or contracting with them to generate profit by cutting costs, intensifying coding, and capturing quality bonuses. Integration may improve quality or expand geriatric care but could harm patients if insurers ration care or avoid complex patients. Despite this transformation, little is known about how integration shapes care for older adults due to obscured ownership and contracting arrangements. We will (1) identify insurer-integrated practices by applying machine learning to Medicare claims, prescribing data, and corporate ownership sources; (2) isolate the causal impact of insurer-practice integration on patient access and quality.

2. Upcoding and risk adjustment in Medicare Accountable Care Organizations (ACOs)

Accountable Care Organizations (ACOs) are groups of clinicians and organizations that share responsibility for the cost and quality of care for Medicare beneficiaries. To discourage cherry-picking of healthier patients, Medicare adjusts payments based on each patients expected health risk a process known as risk adjustment. Yet this system can be gamed: organizations can intensify diagnostic coding to make patients appear sicker than they are, increasing bonuses without making care more efficient. Concerns about such practices are especially pronounced in the new ACO REACH program, which allows direct participation by corporate and insurer-owned entities with sophisticated data and coding infrastructures. This project evaluates coding intensity in ACO REACH by applying econometric methods to Medicare claims and enrollment data, testing whether organizational integration and ownership structure are linked to differential growth in risk scores. The findings will inform policy efforts to ensure that value-based payment promotes efficiency and equity rather than artificial risk inflation.

Additional Information. Finalists will complete a two-stage interview process, including an initial communication interview and a take-home coding exercise.