State Health Expenditure Dataset (SHED), 2000-2013 (ICPSR 36741)

Amount (original)

This variable comprises the amount of money expended on a particular record. All figures are nominal (not inflation adjusted). Note: Negative values may be observed. They likely reflect reconciliation efforts and/or intra-agency transfers. Researchers should be aware of negative values, but also keep in mind that these negative values are included in the Census' dataset (and thus also included in the final estimates put out by Census).

Amount inflation (BLS) adjusted

This variable comprises the amount of money expended on a particular record. All figures are real (inflation adjusted). The Amount variable was inflation adjusted by the project team using BLS estimators.

Categories of spending

This variable enumerates spending for *32 and *27 into categories of non Public Health spending, with the exception of a broad area of Public Health (value 7). Categories were created by the project team, in consultation with an extended Foundational Public Health Services framework.

Category

This variable indicates which major category of spending the expenditure falls in. It uses the Foundational Public Health Services model, as well as a category for "Not Public Health." This variable is expanded upon in the OtherCat variable.

Census state ID

This data element is the identification number assigned by the US Census Division of State and Local Finance to the state. Note: These data elements are different that Census's FIPS assignments.

Code (Not/Maybe/Yes PH)

This variable is the result of re-coding done by the Johns Hopkins and de Beaumont team, using the Foundational Public Health Services model of what constitutes Public Health spending, as separate from e.g., community health care or disability-related health care spending. Note: For purposes of aggregation (and creation of the PHS variable), "Maybe" codes are included in the Yes-Public Health total.

Department

Variable indicates the type of department in which a record's expenditure occurs. These include the designated State Health Agency, Behavioral Health (and Mental Health) departments, Environmental protections, and all other agencies.

Expenditure area

Data in this variable are listed as either functional code 32 (All Health - Other) or 27 (Environmental Health; available 2007 and later). These are created by the Census as part of their broader activities around coding expenditure data from state governments. Functionally, 27 is a subcode of 32 (i.e., both are aggregated together in national estimates of Health-Other spending), so this variable may be of limited utility for data users.

Function

This variable was created by the Census to indicate the type or function of spending associated with a particular code (i.e., in this dataset *32 or *27). These numbers are aggregated into direct, intergovernmental transfers, and equipment expenditures in the Type - Type of expenditure variable.

LVL1 Description

These variables represent where in the hierarchy the expenditure occurs. Each state is organized differently, and often states report these differently across different years in the dataset. These variables are as originally provided by the states to the Census, and have not been cleaned. These variables, however, were used by the Johns Hopkins project team to create the variables Code - Code (Not/Maybe/Yes PH) PHS - Public Health Spending Cat - Category. Coding was based on the Foundational Public Health Services model. The variables are categorized in a hierarchal manner with Level 1 representing the highest level in a state government (e.g., Department of Health) to Level 6 representing the most granular level (e.g., a particular communicable disease control program). - Level 1

LVL2 Description

These variables represent where in the hierarchy the expenditure occurs. Each state is organized differently, and often states report these differently across different years in the dataset. These variables are as originally provided by the states to the Census, and have not been cleaned. These variables, however, were used by the Johns Hopkins project team to create the variables Code - Code (Not/Maybe/Yes PH) PHS - Public Health Spending Cat - Category. Coding was based on the Foundational Public Health Services model. The variables are categorized in a hierarchal manner with Level 1 representing the highest level in a state government (e.g., Department of Health) to Level 6 representing the most granular level (e.g., a particular communicable disease control program). - Level 2

LVL3 Description

These variables represent where in the hierarchy the expenditure occurs. Each state is organized differently, and often states report these differently across different years in the dataset. These variables are as originally provided by the states to the Census, and have not been cleaned. These variables, however, were used by the Johns Hopkins project team to create the variables Code - Code (Not/Maybe/Yes PH) PHS - Public Health Spending Cat - Category. Coding was based on the Foundational Public Health Services model. The variables are categorized in a hierarchal manner with Level 1 representing the highest level in a state government (e.g., Department of Health) to Level 6 representing the most granular level (e.g., a particular communicable disease control program). - Level 3

LVL4 Description

These variables represent where in the hierarchy the expenditure occurs. Each state is organized differently, and often states report these differently across different years in the dataset. These variables are as originally provided by the states to the Census, and have not been cleaned. These variables, however, were used by the Johns Hopkins project team to create the variables Code - Code (Not/Maybe/Yes PH) PHS - Public Health Spending Cat - Category. Coding was based on the Foundational Public Health Services model. The variables are categorized in a hierarchal manner with Level 1 representing the highest level in a state government (e.g., Department of Health) to Level 6 representing the most granular level (e.g., a particular communicable disease control program). - Level 4

LVL5 Description

These variables represent where in the hierarchy the expenditure occurs. Each state is organized differently, and often states report these differently across different years in the dataset. These variables are as originally provided by the states to the Census, and have not been cleaned. These variables, however, were used by the Johns Hopkins project team to create the variables Code - Code (Not/Maybe/Yes PH) PHS - Public Health Spending Cat - Category. Coding was based on the Foundational Public Health Services model. The variables are categorized in a hierarchal manner with Level 1 representing the highest level in a state government (e.g., Department of Health) to Level 6 representing the most granular level (e.g., a particular communicable disease control program). - Level 5

LVL6 Description

These variables represent where in the hierarchy the expenditure occurs. Each state is organized differently, and often states report these differently across different years in the dataset. These variables are as originally provided by the states to the Census, and have not been cleaned. These variables, however, were used by the Johns Hopkins project team to create the variables Code - Code (Not/Maybe/Yes PH) PHS - Public Health Spending Cat - Category. Coding was based on the Foundational Public Health Services model. The variables are categorized in a hierarchal manner with Level 1 representing the highest level in a state government (e.g., Department of Health) to Level 6 representing the most granular level (e.g., a particular communicable disease control program). - Level 6

Object Description

Data in this variable constitute object codes (sometimes known as class codes) as originally provided to the Census by the states. These are the most granular descriptor of the type of spending. Whereas the Record hierarchy descriptors (LVL1-LVL6), describe where an expenditure occurs, the object description variable contains information on what the monies are expended upon.

Public Health Spending

This variable is an aggregation of "Yes Public Health" (Code=3) and "Maybe Public Health" (Code=2). In short, it relates to the Johns Hopkins project team's desire to create a conservative revision of official Census estimates - i.e., assuming the Census coded correctly unless re-coding could directly show otherwise (where Code=3).

State (Abbreviation)

This variable indicates which state the expenditure occurs in, using a 2-letter postal abbreviation.

Subcategory

This variable is a code of subcategories within the Cat - Category variable, using the Foundational Public Health Services model, as well as categories identified by the Centers for Medicare and Medicaid Services' National Health Expenditure Accounts methods manual. This variable includes subcategories for all records, i.e., both for public health and nonpublic health expenditures.

Type of expenditure

This variable aggregates the function of the expenditure into three types: direct expenditures (function=E, F, G), intergovernmental transfers (M, N, R), and equipment (K). This variable was generated by the Johns Hopkins project team using original Census data (variable Function - Function).

Umbrella Agency status

This variable indicates whether the designated State Health Agency is part of an umbrella agency / super-agency - e.g., an agency that performs encompasses Public Health and Medicaid versus a standalone Public Health agency. Variable comes from the Profile Study conducted by the Association of State and Territorial Health Officials.

Year

This variable represents the fiscal year of the associated expenditure.