1. Traditional workforce planning focuses too much on numbers and makes predictions about the future that may not come to pass given uncertainties.
2. The document proposes using scenarios and modeling to better understand how different factors could impact future demand for skills and competencies rather than just headcounts.
3. It presents an example of modeling pharmacy workforce needs over time under different scenarios that show the range of possible outcomes is much wider than a single prediction, emphasizing the importance of stress testing policies against multiple futures.
2. Main points
Consider the unpredictability of the future
Present uncertainty to decision makers
Model skills and competences not numbers
Understand what drives future demand
Use scenarios to capture uncertainty
Stress test before you implement
3. How many doctors do we need?
Do we trust this?
Trainedhospitaldoctors
(thousands)
60
50
40
30
Year
2014 2040
Supply
Demand
?
4. Policy
What if the future is not what we
expect?
Failure
Megatrends
ProblemSystem
Events
11. Scenario generation workshop
Vary across different
futures
Formal elicitation
protocol
Monte Carlo simulation
Probability
Value
Quantify critical parameters
17. 1. Focusing on numbers is not enough.
2. We don’t know what skills and
competences are needed in future.
3. We don’t know who is best to
provide them.
4. We don’t know what drives demand.
Some embarrassing admissions...
18. What about the rest of the system?
Health
Public health
Social care
2%
98%
Workforces not yet
modelled
Workforces modelled to date
19. What about the rest of the system?
Health
Public health
Social care
10% Other health and
support
21% Paid adult care
and support
24%
Volunteer adult
care and support
43%
Unpaid adult care
and support
2% Workforces modelled to date
20. Skills and competences framework
Competences
Skills
FacilitationLeadershipWellbeing
Knowledge Personal
Types of skill
Prevent, Enable, Assess, Plan, Treat, Rehabilitate, Relieve, Link
Level of skill
1 to 6 representing intensity, experience and accountability
Quantitative skills
Qualitative skills
21. A new challenge: Horizon 2035
What skills and
competences
do we have?
What might we
need in future?
22. Step-by-step process
1. What skills does your workforce have
today?
2. What drives the demand for skills?
3. How does the workforce meet this
demand?
4. How does skills demand change by
2035?
23. 1. What skills does your workforce
have today?
Prevent
Enable
Assess
Plan
Treat
Rehabilitate
Relieve
Link
Unpaid adult social care workforce
Nurses
Dentists
Medical generalists
Medical specialists
Volunteer care and support workforce
Other workforce groups
Workforce groups
25. Population
Learning disabilities
Oral health
Singular demand for service
Maternal and perinatal
Infectious disease
Mental long-term conditions
Physical long-term conditions
Skill level: 1 2 3 4 5
3. How does the workforce meet this
demand?
Prevent
Enable
Assess
Plan
Treat
Rehabilitate
Relieve
Link
Unpaid adult social care workforce
Nurses
Dentists
Medical generalists
Medical specialists
Volunteer care and support workforce
Other workforce groups
Workforce groups Increasing concentration and experience
26. Skill level
6
5
4
3
2
1
Example: Level 5 link
skills associated with
infectious disease. This
could include public
health consultants
liaising with community
services and
stakeholders to deliver
public health
programmes.
Multidimensional problem
27. How demand is met today
Physical
LTC
Mental
LTC
Learning
disabilities
Oral
health
Infectious
diseases
Maternity
Singular
demand
28. How demand is met today
Percentage of hours by demand source
Physical LTC
Mental LTC
Learning disabilities
Oral health
Infectious diseases
Maternity
Singular demand
29. Sources of uncertainty
Quantifying critical parameters
Reference future:
Level 5
-
2
4
6
8
10
12
14
16
ALL Life
expectancy
Fertility Migration Wellness Productivity
Key:
median
Hoursin2035(billions)
40%
40%
10%
10%
36. Main points
Consider the unpredictability of the future
Present uncertainty to decision makers
Model skills and competences not numbers
Understand what drives future demand
Use scenarios to capture uncertainty
Stress test before you implement
37. Next steps
What is the optimal mix of skills?
To minimise cost? To maximise flexibility?
Can we standardise scenarios?
Socio-technological-economic and wellbeing
futures?
New robust planning framework
39. More information
Horizon 2035
Future demand for skills:
Initial results
http://www.cfwi.org.uk/publications
/horizon-2035-future-demand-for-
skills-initial-results
40. t
Rethinking workforce
planning for the future
Dr. Graham Willis
Head of Research and Development, CfWI
E: graham.willis@cfwi.org.uk
T: +44(0)78 1234 0405
Hinweis der Redaktion
Many things can push a policy off-track…
Our environment, our system can change – we can have new governments. We have an election in the UK next year.
Our problem might change – we might be trying to solve the wrong thing.
Events can happen – we might get a medical breakthrough or a pandemic.
And we have megatrends – trend we cannot easily change, like an increasing and ageing population. We have to live with them.
All of these can lead to policy failure, especially if we only thing about one expected future.
The solution is to think about many futures…
We generate a set of possible futures. One of these is our expected future, but we model all of them.
[If time: telescope story, actually they look backwards in time!]
They are plausible, you can’t say they couldn’t happen.
They are challenging, you need to make sure your plans could cope.
They are consistent, they have a logical flow of cause and effect, they make sense.
We want an approach that will allow us to create policies that are robust, that take account of the uncertainty of the future, and will lead to good outcomes.
This is the framework we have developed. [explain]
It is transparent and participatory. We wanted a method that open and easy to understand – unlike our previous Excel model which was a black box and impossible to understand. We also wanted to involve stakeholders in all stages. This is really important in situations that are highly political – and there is nothing more politically charged that the healthcare system.
So, for example, they sense-check the model design and outputs.
We do horizon scanning to understand the system.
We know the future is complex.
We have future issues – like workforce cost.
We have factors that are driving change – like our ageing population
We have events that might surprise us – like a technology breakthrough
How do we make sense of this? We ask people for their ideas about the future. We do this online in our ideas bank. Ideas have a narrative or story about how the future might evolve, which they also quantify:
Probability of the idea happening
Impact if it does
Workforces that might be impacted
Their stakeholder group
This allows us to understand people’s worries and concerns, and to map the system, the driving forces and feedback.
We do scenario generation to explore the future.
We hold stakeholder workshops.
We analyse the factors from horizon scanning.
We find the key factors to build scenarios.
We check they are consistent.
We produce narrative scenarios.
We quantify them for modelling.
Factors are analysed by their uncertainty of outcome and their impact on the focal question.
And it’s going to vary.
Some are more important than others.
Some are less important.
Some are predetermined, we have to live with them, like our ageing population.
But those that are highest impact and uncertainty of outcome are the key ones we build scenarios around.
How do we know if the key factors we select all make sense together?
We use a method called cross impact analysis.
Why not health and care? To avoid disagreements!
Here is a simple example. We have 4 factors. Each can evolve in two ways – high or low (we could have more).
Explain A
B is also consistent.
C has low population and GDP, but high energy usage and low emissions?
A and B could be our scenarios. We usually generate 5 to 8 scenarios this way.
To model scenarios we need to quantify the values of parameters that will vary across different futures.
Examples are:
Health needs of the population
Working hours of the workforce
Productivity of the healthcare system
We ask an expert panel to produce a probability distribution to express their uncertainty. This is really important.
We use a formal elicitation protocol – SHELF.
We do this for the expected future, then the scenarios.
We take all these future values and their uncertainty, and do a Monte Carol simulation.
We do workforce modelling to simulate these possible futures.
The model takes:
Data we have and assumptions that we make
Policy levers that we can control – like intake or length of training
Critical parameters, their value in each future scenario – like the health needs of the population.
An example from pharmacy. Here are four scenarios.
Red is supply, blue is demand. The fans show the spread of uncertainty from our Monte Carlo simulation. The outer light band is 95% probability.
The good news is that we are not short of pharmacists
The bad news is that we are training far more pharmacists than we need. Some pharmacists training today won’t get jobs in the future. We are spending money on training that might be better spent elsewhere.
In every future there is over-supply – the desirable futures and the less desirable.
A, B, C are a one-off reduction in training intake.
D, E, F are a phased reductions over 5 years, then a 3% increase over 10 years
Policy option B – one-off 35% reduction in training intake
Policy option E – phased 10% reduction in training intake over 5 years, then 3% reduction for ten years
Scenario 4 is the most challenging for pharmacists – internet ordering, remote diagnosis, 3D printing of drugs
Policy option B is not politically acceptable
Policy option E gives flexibility and can fine-tune if things change – which they will
Although we have made good progress in health professions – the rest of the workforce is unexplored territory!
Another 10% for the rest of the health workforce
Then huge numbers moving into social care, volunteers and unpaid carers.
Number alone will not support effective workforce planning.
We need to think about the skills and competences that people have, and their capacity to gain new skills.
This requires us to think about the kinds of skills and their level of intensity.
This is our new challenge from our Department of Health – to understand what skills and competences we have today, and how they may need to change meet the future, looking out to 2035.
This is an example output from Horizon 2025.
This is the reference future for skills demand for the whole system. Note the spread of uncertainty.!
This is an example output from Horizon 2025.
This is the reference future for skills demand for the whole system. Note the spread of uncertainty!
This is an example output from Horizon 2025.
We have developed a set of six scenarios for the future.
This is an example output from Horizon 2025.
Top three are:
LTC physical
LTC mental
Learning disabilities
Level 1 has highest growth (L1 = no training; L2 = some training &/or experience; L3 = 3 to 4 yrs formal edu; … L6 - +6 yrs eg consultant
Levels 5 and 6 have highest uncertainty