Bill Status Data Model : Terminology Post #1
Sean McGrath
We have made a good start in the "Towards a Bill Status Data Model"
thread. Thanks to Tom Bruce, Karen Suhaka, and Shay Wilson for
contributions so far! I intend to wait a couple more days for further
contributions on that thread before replying to the excellent
points/concerns/options being raised there.
While that thread is cooking however, I think we can usefully start the
search for shared terminology. Getting agreement on basic terms (nouns,
verbs, adjectives) is fundamental to any data modeling exercise. It is
absolutely critical in legislative informatics because the terminology
differs so greatly from state to state - and in some cases - chamber to
chamber.
I am going to number each piece of terminology discussion so that we can
refer to them easily in the conversations that will hopefully ensue. One
e-mail per term to keep things as simple as possible.
Bill Status Data Model : Term #1 : "Bill"
The word "Bill" as in "Bill Status", is, I think a "given" in the sense
that it has very common currency in the world at large. The media speak
of bills, as do Members, lobbyests etc. Many existing computer systems
use the term. I.e. "bill drafting", "billtrack50", "manybills" etc.
However, when you look a little closer, it is not obvious what makes a
bill a bill. We could say that a "bill" is any piece of business
formally named and worked through the publicly visible legislative process.
Possible objections:
- Some legislatures have a formal distinction between "bills" and other
forms of formally named business items. In some legislatures, the term
"bill" is reserved for items that go through the visible legislative
process with the intent to change law. This distinguishes them - in some
states - from items that are also formally worked through the
legislative process but that do not attempt to change law. A common
example is a "Resolution".
- Some legislatures have a single "bill" numbering convention but
reserve numeric ranges for things that that legislature does not
consider to be bills. e.g. Resolutions.
- Some legislatures have items beyond "Bills" and "Resolutions". E.g.
Executive Redirection Orders
- Some legislatures use the term "measure" as a "catch all" term to
cover Bills+Resolutions etc. - even if the term "measure" is not used on
the legislative documents themselves.
Some states use the word "Act" starting with the first public drafts.
Some use "Bill" and the word "Bill" changes to "Act" upon passage. Some
use "Bill for an Act" on the way through. Others go with "Act" from the
get-go.
One final example of the potential complexities if we try to get too
granular in the data model here. Proposed changes to state Constitions
are pretty much guaranteed to be exotic animals! They may or may not be
numbered differently. They may or may not be considered "Resolutions" -
even though they change law (presuming you accept that state
Constitutions constitute "law" in your legisprucence taxonomy:-)
I propose that we use the term "bill" in conversation colloquially (like
everybody else) but to avoid confusion in the data model, we use the
term "measure" as a catch-all term. Any distinction between measures in
terms of chamber of origin or potential legislative impact or flavor of
Resolution or constitutional-based legal nuance we leave out of scope of
the data model. If useful semantics can be derived by looking at the
measure name (or some other attribute we may add to our data model),
then well and good.
The other alternative would be to adopt the term "bill" in its loose
colloquial sense directly in the data model. My concern with that is
that existing terminological practice suggest the word "measure" has
currency as a catch-all term and we run the risk of adding to the
confusion unless we go with existing common practice.
Proposal:
Rather than get embroiled in the worlds of legisprudence or political
science or Plato's Theory of Forms (interesting though they are!), I
propose we "go with the flow" and use the term "bill" loosely in
language but in our data model use the term "measure". In other words,
the bill status data model is a data model for reporting the status of
"measures". In some legislatures there will be a 1-to-1 between measures
and what that legislature thinks of as bills. In others, some measures
will be "bills" (local terminology) and others will be "resolutions"
(local terminology) or other more exotic animals like "executive
redirection orders" (local terminology).
I propose that our data model does not impose a taxonomy on measures but
rather leave it to the state-specific conventions for the measures to
add whatever legislative semantic is required. Examples:
- In Kansas "HR6027" is a House Resolution Measure. I can tell that by
the presence of "HR" at the start of the name and the fact that the
"bill" is in the numeric range 6000-7000.
- In Pennsylvania, the numbering scheme for bills subsumes Resolutions
e.g. Senate Bill 55 of 2003-4 is a "Joint Resolution". I.e. there is no
separate numeric range for Resos. I can tell it is a Resolution by the
title given to the Measure.
- In Wyoming, joint resolutions are bills named "HJ<N>" and "SJ<N>".
There is a separate numeric range for Resos.
The only absolute requirement for a measure in our model is that it have
a name. That name may be completely opaque, it may have semantics -but
we are not going to concern ourselves with that in the data model
because of the complexities above. How does that sound?
This is probably the most critical noun for us to get down before we
move on to any adjectives or verbs in the model. If the proposed use of
"measure" is acceptable, I propose to move on to some possible attributes.
Regards,
Sean