Now that is an evil of the government (and some larger corporations) not related
to bidding or even contracting, and I've been party to that many times.
The budget cycles in the US Federal government - and from your post, Paul, it
sounds like the UK government as well, is fraught with problems that are
antithetical to the type of reforms that Scott Ambler is suggesting.
An Agency is given a budgetary amount for the next fiscal year. In the US
Federal government this is done two years in advance, by law. In other words,
right now the Federal government agencies are preparing their budget requests
for the fiscal year 2011 that starts October 1, 2010. These requests still have
to be approved by Congress which may not occur until the summer of 2010, and
sometimes, as in the past, they don't occur until after the new fiscal year
causing the government to stop for lack of budget. It has happened.
When considering the budget the budget making group - different groups in
different agencies - consider first what was spent in this year's budget. The
rule is simple. "If you did not spend it, you did not need it, therefore your
budget will be reduced accordingly". Agencies and Departments down to the
lowest level of government entity for which there is a budget allocated react
with the dictum: "Spend it or lose it." Even when a department legitimately
saves money and feels as though they might be able to get along on less, the
managers know how difficult it is to increase a budget so they will tend to try
to make sure their budget amount is not reduced. There is also the personal ego
and power aspect for management which follows the formula: money (budget) =
power. There is also an understood corollary: FTE (full time equivalent or the
number of people working for you) = power.

What all this means is that an agency that under-spends its budget will start
spending at the end of each fiscal year to use up whatever remaining budget is
left. Of course the opposite is true as well, when there is a shortage toward
the end of the fiscal year, contracts will be suspended until the start of the
next fiscal year when more money will be released. This is very good for
training because in many cases, agencies figure out ways of using up money
allocated for other categories in training resulting in a increased demand for
training at the end of each fiscal year. The justification is that there is no
such thing as unneeded training. This is music to the ears of those of you in
this forum who do training. There are also a great number of small contracts -
usually $25K or less - that can be awarded without competition.

This all has nothing to do with the RFP process, Scott (Ambler). This has to do
with the budgeting and control process of the government or corporation. And
the scenario that Paul describes is an exact manifestation of that process.

Because of the legal budget constraints and the common way those under the
constraints work around them, it is difficult to award an agile contract in the
way Scott suggests. There must be at least a fixed ceiling on the effort to keep
to the budget allocated. Then, regardless of the success of an effort, when the
budget is used up, the contract must, by law, terminate because funding is no
longer available to pay the contractor until the next fiscal year provided there
are funds allocated in that year's budget. Yes, more funds can be requested,
but the time and effort to secure the funds, even reallocating from another
budget line, usually takes more time than simply waiting for the start of the
next fiscal year.

So here's the quandary to resolve using agile techniques. Budgets are by law
determined two years in advance and may be cut or changed by Congress which
means the end result is also law of the land. Once determined, that budget line
item must be spent for that line (e.g. a specific project to achieve a specific
goal, such as build a bridge where one is not needed, or fund a $350 toilet seat
cover) and that line only and must be completely used up and not exceeded.
Management is then evaluated on its ability to achieve the stated goals using
that allocated budget amount.

Put yourself now in the position of a manager of an agency. You have $240K to
get a software project completed. Would you award it to a contractor who
proposes 30K sprints with no guaranty of achieving the goal, but will guaranty a
quality result and better software product for the government? Here are the
three possible results you are facing. The contractor completes the job in
eight sprints spending the $240K and everyone is happy. The contractor
generates exceptional software over the eight months, but because of additional
requests on the part of the government and so forth, does not complete the job
requiring the government to stop work completely (assuming the job was to be
completed within the original 240K no additional money was budgeted for the
following fiscal year. Had such money been requested it would probably be cut by
the time of approval). It is possible to reallocate funds from another budget
line item, but this is illegal...if caught. The third option sounds great: the
possibility that the contractor finishes early, say after the fifth sprint. The
goal is achieved completely - there is no additional work to be done on this
project. Reallocating the remaining funds to another project is also
illegal...if caught. I think that most if not all agency and department heads
would opt for a fixed price contract even when they know it may not work because
at the end of the contract you can hold the contractor's feet to the fire to
complete the job even when the contractor has to spend more than was bid. It's
safer and less work. Remember that the people who are responsible to make sure
the contractor fulfills his obligation are not the agency management or the
agency staff. The contracting offices of each agency is responsible. This is
another Federal law.

There is one other thing to remember. All of these laws were in reaction to
some contractor and agency head at some time ripping off the taxpayers' money to
line their own pockets. It is still done, and there will be more laws to
prevent it.

So I'm wondering if there is an agile or any other approach that would allow for
iterative, high quality software development that still adheres to the laws,
protects against abuse, and achieves the goals of the agencies to serve the
general public.

An appropriate topic on the US Independence Day. Happy Fourth to all.



From: Paul Oldfield
Sent: Saturday, July 04, 2009 6:30 AM
To: agilemodeling@yahoogroups.com
Subject: Re: [AM] RFPs the Agile Way





(responding to Steve, Scott)

I came across an interesting example - gov't work -
where the customer really did want fixed price, but
on investigation were prepared to consider variable
scope, especially if we could offer to deliver in their
order of priority. Seems they had a fixed budget
and must in effect spend all that and no more, or
bad things happen.

Paul Oldfield
Capgemini






--------------------------------------------------------------------------------



No virus found in this incoming message.
Checked by AVG - www.avg.com
Version: 8.5.375 / Virus Database: 270.13.3/2217 - Release Date: 07/03/09
18:11:00


[Non-text portions of this message have been removed]