The speed of light has been slowing down since it's been measured over the last 200 years. Some will say that it's due to improved and more accurate measuring techniques, but then they have to explain why the exponential curve just happens to be "by chance" a perfect match for the YEC model.
In 1982, Dr. Barry Setterfield, an Australian astronomer, made an amazing discovery that he first presented at the 1983 National Creation Conference. Dr. Setterfield tabulated all the measurements of the speed of light since 1675. The computer-generated curve showed that the speed of light was slowing at an exponential rate. Extrapolating back to 4082 BC, the approximate time of creation of the earth, the speed of light was 107 – 1011 times its present velocity. Immediately, the scientific community set out to discredit Dr. Setterfield's report, only to verify that he was indeed correct to an extremely high statistical probability. No one could explain this exponential decay.
Ten years later in 1992, I realized that the Hubble space telescope and others show space filling with gas and dust. Many enormous clouds of dust and gas in space are so dense that they completely obscure the light from any stars behind them. Hubble's Bubble is a good example. This enormous cloud of gas six light-years across is in the constellation Cassiopeia. Dust and gas decrease the speed of light. This would explain Setterfield's discovery that the speed of light was much greater at the time of creation of the earth. Light waves are the result of vibrations of electrons in matter. In my opinion, light waves are propagated instantaneously in an absolute vacuum. When a light wave strikes some matter, the infinitesimal inertia of the electrons requires an infinitesimal amount of time for the electrons in the matter to vibrate at the frequency of the light wave. This results in a delay in light passing through any medium. If this is true, light from the stars would be seen on the earth the instant the stars were created because space had no gas or dust at the time of creation.
I suggest that a simple experiment would be to measure the velocity of light in a container holding several atmospheres of air, and then measure it as a vacuum is gradually created in the container. I predict that its velocity will increase exponentially as the vacuum is created. In view of Dr, Setterfield's discovery, the earth is not 12 – 20 billion years old, but only about 6,000 years old. Confirmation of a 6000-year age of the earth was discovered by correcting errors found by Creation Science nuclear decay of radioactive isotopes.

The exponential decay of the velocity of light in space with time is given in Setterfield, Barry, 1983, The Velocity of Light and the Age of the Universe, National Creation Conference, Minneapolis – St. Paul. This graph shows the speed of light slowing down as space is filled with dust and gas.
Http://www.intelligentdesigntheory.info/speed-light.html
MODIFICATIONS TO LIGHT-SPEED MODEL
During the period from 1980 until 1987, my light-speed research was developed around the 350 years of data on light-speed, c, and the best part of 90 years of accumulated data from the associated atomic constants. The theoretical approach adopted was one that was in agreement with the observational evidence that was amassed during that period. The results were outlined in the Report of August 1987, which was initially issued under the auspices of SRI International and Flinders University. Since then, the data base has been extended, mainly through astronomical evidence, and ultimate causes have been sought for the observed phenomena.
The astronomical data have required the energy conservation approach adopted earlier to be extended to cover an additional effect that is not apparent over the short term in terrestrial data. When this is done, some minor changes occur in the basic approach adopted during the 1980's. Importantly, the astronomical and geological evidence extend the data base back to the frontiers of the universe and with it our understanding of the probable processes acting to bring about the observed effects.
As a result of this extension, some of the conclusions reached in the 1987 Report require modification. The new model is presented in Behavior of the Zero Point Energy and Atomic Constants. A listing of the most important of these changes now follows.
OUTLINE OF NEW MODEL:
It is generally accepted that the fabric of space was stretched out in response to processes operating at the inception of the cosmos. It is proposed that this stretching invested the fabric of space with an energy that eventually manifests as the zero-point energy, which is an intrinsic property of the vacuum. Evidence is deduced that the zero-point energy is increasing with time. The reason for the progressive change in the strength of the zero-point energy (ZPE) may be traced to the behaviour of the vacuum at the Planck length level. A smooth increase in the ZPE induces a smooth decline in the speed of light, c, and the rate of ticking of atomic clocks, while simultaneously smooth changes in the values of some atomic constants also occur.
Evidence in the scientific literature indicates that the ZPE sustains atomic structures universally. Therefore it is proposed that, as more energy became progressively available to them from the vacuum, atomic particle and orbit energies underwent a series of discrete isotropic increases or quantum jumps when the ZPE increased to a quantum threshold. Within the quantum interval, energy was conserved in all atomic processes, as atoms could not access fractions of a full quantum of energy from the ZPE. Thus, with increasing time, atoms emitted light that shifted in jumps towards the blue end of the spectrum. With increasing astronomical distance (looking back in time), theory indicates that the resulting redshift should increase in jumps of 2.633 km/s, in accord with Tifft's statistically treated observations that yield a quantisation of 2.667 km/s. This means that the treatment of the redshift and the Doppler formula in the 1987 Report has been superseded.
MODIFICATIONS:
From Table 12 on page 28 of the 1987 Report , Option III is now favoured over Option I by the new work. This means that e, the electric permittivity of the vacuum, and m, the vacuum's magnetic permeability, are both proportional to 1/c. If 'e' is the electronic charge, this results in the quantity e2/ε being independent of c within a quantum interval, as is its magnetic equivalent which is m2/μ. This ensures that the electric and magnetic properties of the vacuum always maintain the same proportions as is demanded with any change in the strength of the ZPE. If this were not the case, there would also be dispersion effects on light propagating through space that are not in accord with observation.
The section on Gravitation in the Report on pages 43 and 44 is now superseded by a treatment in accord with SED physics and the effect of the ZPE on point particle charges. The result is that there would be no change in gravitational phenomena, including gravitational acceleration and orbit times for planets and celestial bodies, as the ZPE increased and c declined. This arises because the term Gm is a constant, where G is the Newtonian gravitational constant and m is mass. The dynamical (orbital) clock has thus ticked at a constant rate. Although this conclusion is the same as in the Report, the means whereby it is derived is different.
The other major change has been the determination of the basic curve for light-speed. The paper undergoing review establishes a linear relationship between the redshift and the velocity of light via the behaviour of the ZPE. The redshift can then be used as a measure of the value of c. Furthermore, looking out into astronomical distances is equivalent to looking back in time. Consequently, the standard graph of redshift plotted against distance is equivalent to a graph of the speed of light plotted against time. It is the same graph with re-scaled axes. Observational evidence indicates that an oscillation is also involved. This oscillation means that the minimum value attained for c around 1970 may be followed by a slight increase or, alternatively, may be followed by further decay after this stationary period has ended. The actual behaviour at this point has yet to be determined. However, evidence is accumulating that a slight rise in c has occurred. This matter is the subject of ongoing investigation.
In secular literature, changing c models are now being considered under the heading of VSL (variable speed of light) theories. The 'cDK' reference is retained to distinguish it from other models. However the term 'DK' which came about from the word 'decay' really has nothing to do with the concept of decay as it is known in biology. This has given rise to some misconceptions about the behavior of light, which might have other implications. In fact, photons in transit do not lose energy by the mechanisms in operation in this model, so light itself is not undergoing any 'decay' when its velocity drops. The drop is due entirely to an increase in the ZPE which itself is responsible for sustaining all atomic structures throughout the cosmos.
Barry Setterfield
2nd October 2001; updated August 31, 2007
Modifications To Light Speed Model
Experiments and results
The speed of light has been observed to have exceeded the speed of 300,000 km/sec, albeit over short range, in quantum tunneling experiments (Landauer 1993, Brown 1995).
Enjoy.