RFC 7230 does not give any instructions on how to parse the "reason text" element of the status line, so no encoding is implied or specified.
The ABNF grammar indicates that your string would be valid if presented in any "ASCII-compatible" encoding, such as UTF-8:
reason-phrase = *( HTAB / SP / VCHAR / obs-text )
VCHAR
means "visible ASCII character", and obs-text
is any byte in the range 0x80 to 0xFF.
However, there is no definition of how those bytes should be interpreted. Instead, the specification recommends that clients do not parse it at all:
The reason-phrase element exists for the sole purpose of providing a
textual description associated with the numeric status code, mostly
out of deference to earlier Internet application protocols that were
more frequently used with interactive text clients. A client SHOULD
ignore the reason-phrase content.
So go ahead and put whatever text you like there if you're using it for low-level debugging or with a custom client. But don't be surprised if standard clients either ignore it or misinterpret it as a different character encoding.