You have ${pages_left} free articles remaining.
Subscribe now to continue getting the industry’s most current and thoughtful review of hospital performance.
January 2019
Spotlight: The “Goldilocks Zone” of Budgeting – Part II
ic_arrow_up.svg
Hide Page Sections
ic_arrow_drop_down_black_24px.svg
Show Page Sections
In the December 2018 Flash Report, the idea of the “Goldilocks Zone,” the percentage of performance over- and/or under-budget that could be defined as “accurate,” was introduced.
Organizations typically use budgets as a “path” to measure the financial success of their operations. If they perform favorable to budget, that means there are “extra” resources available to the organization that were not anticipated. If they perform unfavorable to budget, this suggests that hospital operators did not manage the business as planned. The overarching assumption of this “path” is that organizations are accurate budgeters. But, just how accurate are organizations when it comes to budgeting?
This month’s Spotlight will continue to challenge the long-held belief that performance to budget is an adequate measure of success with the point of view that budget accuracy should be prioritized instead. Accuracy has to be defined; what percentage of over- and/or under-budgeting is accurate? What are the appropriate thresholds [over- and under-budget] for health systems?
Overview
Data Background
Final Considerations
Overview
Getting Started
Outside the Goldilocks Zone
So what does this all mean?
To summarize, only a small percentage (20%) of organizations perform within +/- 3% of their budgeted Operating Margin. Based on these findings, this validates that comparing performance to budget for the majority of organizations is not the best approach for qualifying satisfactory or unsatisfactory performance. Performing above budget is not always unfavorable as falling within a few [percentage] points is actually favorable. The focus on attaining an accurate budget should be prioritized over comparing actual performance to budget. Once an organization can create an accurate budget, the exercise of comparing actual performance to budget becomes a more adequate measure.
Based on the data, those organizations that most often find themselves within the Goldilocks Zone share several characteristics:
  • The ability to accurately project the next year’s volumes
  • Operating margin targets that fall within a few points of the prior year’s operating margin
  • A more conservative approach to setting targets – not setting them too high or below the prior year’s actual performance
What we also found, and as we have already mentioned, change is slow. Whether it’s the growth in patient population or seeing financial improvement from newly implemented change initiatives, the financial realization occurs much slower than oftentimes budgeted for. If rapid change is desired and is to be enacted from the top-down, budget approaches such as rolling forecasts and zero-based budgeting may be greater drivers of significant change.

In an exercise to validate the predictive ability of the prior year’s performance, a predictive model was created where historic data was used to the determine the most correlated variables to apply in calculating the current year’s operating margins – the outcomes this model can be seen in Figure 8.
Figure 8: Predicting Operating Margin
Data Source: Kaufman Hall Axiom Data Date Range: January 2016 to November 2018 YTD
For our model, operating margin was calculated at the monthly level for each hospital. The variables of the prior year’s actual performance and total expenses were our highest weighted coefficients. Similar to Figure 7, a random sample was used to visualize the outputs for Figure 8. In the results, the majority of our predictions fell within the Goldilocks Zone of +/- 3% of actual performance. In researching a sample of outlying circumstances, it was identified that these were hospitals in the middle of financial turnarounds – turnarounds need a more precise model of prediction than the general population. It’s worth noting one predictive model was applied for all hospitals in this exercise. As each hospital is different, each hospital would need their own predictive model – and variables – to better forecast the next year’s performance. A key takeaway, however, is that by applying one model to all hospitals and receiving very accurate results across the population, it proves hospitals may not be as volatile as they are often perceived.
Analysis assumptions and data background
In December 2018’s Spotlight, the data were extracted from Kaufman Hall’s Axiom Database for the date range of January 2017 to November 2018. This month, to expand the dataset to increase the population of the analysis, data were extracted from January 2016 to November 2018.
If you are not familiar with Kaufman Hall’s Axiom Database, this is the repository that stores all of the budget and [actual] performance data of hospitals that subscribe to Kaufman Hall’s budgeting platform – this includes about 715 organizations across the United States. This is a sufficient population size to use for this analysis, and includes health systems of all types and sizes.
Getting started
To first assess the population, a review of how accurately organizations budgeted Net Operating Revenue was performed. Figure 1 displays the distribution of Net Operating Revenue Performance to Budget. In the visualization, the x-axis displays what percentage under- or over-budget organizations performed at, while the distribution displays what percentage of the organizations performed at which over- or under-thresholds. When composing this distribution, it was discovered almost one-third of the organizations within the Kaufman Hall database broke the threshold of 20% under- or over-budget. For the analysis, the population was limited to only those organizations who fell within 20% of budget [as to not skew the data and results].
Figure 2: Operating Margin Performance to Budget
Data Source: Kaufman Hall Axiom Data Date Range: January 2016 to November 2018 YTD
In Figure 1, a normal distribution can be seen – the distribution is bell-shaped, and demonstrates an appropriate mean and standard deviation amongst the population in the dataset. Looking at Figure 1 in further detail, it is seen only 5.7% of organizations performed at budget; roughly 36% of organizations performed within +/- 3% of budget; and 52% of organizations performed within +/- 5% of budget. Overall, it appears the organizations within the dataset are average budgeters relative to predicting and/or budgeting for Net Operating Revenue. Using a similar approach to the accuracy assessment of Net Operating Revenue, Figure 2 explores how well organizations predict profitability by assessing the distribution of Operating Margin.
Figure 1: Net Operating Revenue Performance to Budget
Data Source: Kaufman Hall Axiom Data Date Range: January 2016 to November 2018 YTD
When comparing Operating Margin’s distribution to Net Operating Revenue’s distribution, a stark difference is displayed – the distribution is not bell-shaped, it’s rather flat. In December 2018’s Spotlight, it was calculated that only about one-third of hospitals performed within +/-3% of budget when it came to profitability, meaning there was about an equal amount of hospitals who performed above and below budget. As we expanded our dataset from 2017-2018 to 2016-2018 for the January 2019 Spotlight, the even spread of these data expanded even more; only about 20% of hospitals performed within +/-3% for this date range, with 45% performing below budget, and 51% performing above budget.
All of this begs the question “why are more hospitals not within the Goldilocks Zone?” Why are so many hospitals falling outside of +/-3% and +/-5%?
Exploration of why more hospitals are not within the Goldilocks Zone
One of the goals of this analysis was to see which variables are the best predictors of actual performance – prior year’s performance, volume, expense, or a mix of all of them? Figure 3 examines how strong a predictor the current year’s budget was to actual performance.
Figure 4: Net Operating Revenue Current Year (Y) vs. Net Operating Revenue Prior Year (X)
Data Source: Kaufman Hall Axiom Data Date Range: January 2016 to November 2018 YTD
For a quick refresher of statistics and R-Squared: R-Squared falls between 0% and 100%, and it is the statistical measure that represents the proportion of the variance for a dependent variable (Y) explained by an independent variable (X). In simpler terms, R-Squared is the percentage of how well the independent variable (x) predicts the movement of the dependent variable (Y) – in this case, how well the current year’s budget predicts the current year’s actuals. Typically, the higher the R-Squared, the stronger the correlation between the variables. Obviously, there are statistical measures for accuracy beyond R-Squared, but for this analysis, the approach was kept simple. Viewing Figure 3, the R-Squared came out to be 86.4%. It is unsure whether 86.4% is a respectable result, so the exploration of other predictor variables was needed as a mode of comparison. In Figure 4, the current year’s actual performance is compared against the prior year’s actual performance to get an idea.
Figure 3: Net Operating Revenue Current Year (Y) vs. Net Operating Revenue Budget Current Year (X)
Data Source: Kaufman Hall Axiom Data Date Range: January 2016 to November 2018 YTD
When comparing the current year’s actuals to the prior year’s actuals, an R-Squared of 99.2% was produced. Comparing the results in Figure 4 to those in Figure 3, one can surmise the prior year’s actual performance is a stronger indicator – or stronger predictor – of current performance than the current year’s budget. To explore this notion further, Figure 5 conducts this same high-level analysis, but instead uses Operating Margin.
Figure 5: Operating Margin Current Year (Y) vs. Current Year Budget and Prior Year Actual
Data Source: Kaufman Hall Axiom Data Date Range: January 2016 to November 2018 YTD
In Figure 5, the visualization on the left displays the comparison of Operating Margin Current Year vs. Current Year Budget, and the visualization on the right displays the comparison of Operating Margin Current Year vs. Prior Year’s Actuals. The Current Year Budget produced an R-Squared of 76.2%, while the Prior Year’s Actuals produced an R-Squared of 89.3%. Again, the performance of the prior year is more accurate predictor of the current year’s performance than the current year’s budget. Many organizations begin the budgeting process by first projecting the next year’s volumes. To do this, organizations typically rely on historic data to make these projections, and configure these projections based on phenomena in the current year. Since most of the line items within the budgeting process are dependent on volume, if volume is inaccurately projected, budget inaccuracy will only be compounded further down the budgeting process. This makes predicting volume the most important factor in creating an accurate budget.
Figure 6: The Three Big Steps of the Budgeting Process
Data Source: Kaufman Hall Axiom Data Date Range: January 2016 to November 2018 YTD
At a high-level, the analyses introduced thus far have mirrored the thought process of the first two steps in the budgeting process. The analysis started with an assessment Net Operating Revenue (Step 1), followed by assessment of Operating Margin (Step 2). Based on what has been examined, in building the budget for the next year, it would be more accurate to use the current year’s actual performance numbers – less any outlying circumstances (i.e., unit closures, new construction) – and not making any significant adjustments. Based on the data, where hospitals seem to develop inaccuracies during budget planning are in two key areas: 1) Making significant adjustments to the current year’s actual performance when planning for next year’s budget (e.g., large decreases in LOS from performance improvement initiatives, large increases in volumes based on population health assessments) 2) Step 3: Calculating expenses and reconciliation, and balancing the gap(s) between expectation and reality During Steps 1 and 2, popular “adjustments” to the current year’s actual performance involve the expectations of growth in volume, increased acuity, decreased length of stay based on initiative of the hospital to improve performance, and specific inpatient service lines bringing in more patients than the previous year. The reality is these areas of expected growth and improvement are not as significant as health systems anticipate. In short: change is slow. The financial realization of increased volumes and improved processes takes longer than what is usually expected. The “adjustments” [made in Steps 1 and 2] are often compounded in the Expense and Reconciliation process (Step 3) during budget planning. As new target volumes are introduced, so are target expenses, and having to bridge the gap to align these two areas to meet target margins often comes with even more “adjustments.” By applying adjustments on top of other adjustments, the current year’s actual performance numbers become much diluted, thus making them a less accurate predictor of the next year’s performance than the current year’s actual numbers.
In Figure 7, we see a sample of hospitals’ Current Year Actual Operating Margins, Current Year Budgeted Operating Margins, and Prior Year Actual Operating Margins.
Due to the large volume of hospitals used in this analysis, a random sample of the population was taken to visualize Figure 7. Based on what is seen in Figure 7, more often than not, hospitals tend to over-budget in comparison to their actual operating margins, and the prior year’s actual operating margins tend to be closer to the current year’s actual performance. Intuitively, all of this makes sense. In the environment of executives’ performance being measured by actuals to budget, budget numbers will be more inflated to create enough wiggle room to perform below budget, thus appearing to perform well.
Figure 7: Operating Margin Performance to Actuals and Budget
Data Source: Kaufman Hall Axiom Data Date Range: January 2016 to November 2018 YTD
Data Background
Final Considerations
Overview
Getting Started
Outside the Goldilocks Zone
©2018 Kaufman, Hall & Associates, LLC
scroll_up.svg
kha_logo.svg
National Hospital Flash Report
menu_icon.svg
mail.svg
Share article
Sign up now for access to the latest news and reports
SUBMIT
Become a subscriber and receive our monthly reports. Choose the option that best fits your needs.
SUBMIT
Monthly subscription: $100/month
Annual subscription: $600/year
Quarterly subscription: $200/quarter