At June 30, 2017, the University reported $1,515,810 as deferred outflows of resources related to pensions resulting from University contributions subsequent to the measurement date that will be recognized as a reduction of the net pension liability in the following fiscal year. Deferred outflows and deferred inflows of resources at June 30, 2017, related to pensions will be recognized in pension expense as follows:

Non-hazardous

2018 $ (396,213)

2019 (243,524)

2020 (44,071)

2021 (24,728)

$ (708,536)

Actuarial assumptions - The total pension liability at the measurement date of June 30, 2016 was determined by actuarial valuations as of June 30, 2016, using the following actuarial assumptions, applied to all periods included in the measurement.

2016

Inflation 3.25%

Salary increases, average including inflation 4.00

Investment rate of return, net of expense, non-hazardous, including inflation 6.75

For the June 30, 2016 and 2015 actuarial valuation, the rates of mortality for non-hazardous, for the period before retirement are according to the RP-2000 Combined Mortality Table projected with Scale BB to 2013 (multiplied by 50% for males and 30% for females). The rates of mortality for the period after service retirement are according to the RP-2000 Combined Mortality Table projected with Scale BB to 2013 (set back 1 year for females) for all healthy retired members and beneficiaries. The RP-2000 Combined Disabled Mortality Table projected with Scale BB to 2013 (set back 4 years for males) is used for the period after disability retirement.

The actuarial assumptions used in the June 30, 2016 valuation were based on the results of an actuarial experience study for the period July 1, 2008 - June 30, 2013.

The long-term expected return on plan assets is reviewed as part of the regular experience studies prepared every five years for KERS. The most recent analysis is outlined in a report dated December 5, 2015. Several factors are considered in evaluating the long-term rate of return assumption. Including long-term historical data, estimates inherent in current market data, and a log-normal distribution analysis in which best-estimate ranges of expected future real rates of return (expected return, net of investment expense and inflation) were developed by the investment consultant for each major asset class. These ranges were combined to produce the long-term expected rate of return by weighting the expected future real rates of return by the target asset allocation percentage and then adding expected Inflation. The assumption is intended to be a long-term assumption and is not expected to change absent a significant change in the asset allocation, a change in the inflation assumption, or a fundamental change in the market that alters expected returns in future years.

32.