diff --git a/pandas/core/arrays/sparse/array.py b/pandas/core/arrays/sparse/array.py
index b8375af797b3a..76c9d013575f0 100644
--- a/pandas/core/arrays/sparse/array.py
+++ b/pandas/core/arrays/sparse/array.py
@@ -21,7 +21,6 @@
     astype_nansafe,
     construct_1d_arraylike_from_scalar,
     find_common_type,
-    infer_dtype_from_scalar,
     maybe_box_datetimelike,
 )
 from pandas.core.dtypes.common import (
@@ -328,8 +327,8 @@ def __init__(
             else:
                 npoints = sparse_index.length
 
-            dtype = infer_dtype_from_scalar(data)[0]
-            data = construct_1d_arraylike_from_scalar(data, npoints, dtype)
+            data = construct_1d_arraylike_from_scalar(data, npoints, dtype=None)
+            dtype = data.dtype
 
         if dtype is not None:
             dtype = pandas_dtype(dtype)
diff --git a/pandas/core/construction.py b/pandas/core/construction.py
index 44224f9709699..ad18cb8dd5bc6 100644
--- a/pandas/core/construction.py
+++ b/pandas/core/construction.py
@@ -21,7 +21,6 @@
     construct_1d_arraylike_from_scalar,
     construct_1d_ndarray_preserving_na,
     construct_1d_object_array_from_listlike,
-    infer_dtype_from_scalar,
     maybe_cast_to_datetime,
     maybe_cast_to_integer_array,
     maybe_castable,
@@ -480,17 +479,13 @@ def sanitize_array(
             subarr = _try_cast(data, dtype, copy, raise_cast_failure)
         else:
             subarr = maybe_convert_platform(data)
-
-        subarr = maybe_cast_to_datetime(subarr, dtype)
+            subarr = maybe_cast_to_datetime(subarr, dtype)
 
     elif isinstance(data, range):
         # GH#16804
         arr = np.arange(data.start, data.stop, data.step, dtype="int64")
         subarr = _try_cast(arr, dtype, copy, raise_cast_failure)
     elif lib.is_scalar(data) and index is not None and dtype is not None:
-        data = maybe_cast_to_datetime(data, dtype)
-        if not lib.is_scalar(data):
-            data = data[0]
         subarr = construct_1d_arraylike_from_scalar(data, len(index), dtype)
     else:
         subarr = _try_cast(data, dtype, copy, raise_cast_failure)
@@ -500,16 +495,7 @@ def sanitize_array(
         if isinstance(data, list):  # pragma: no cover
             subarr = np.array(data, dtype=object)
         elif index is not None:
-            value = data
-
-            # figure out the dtype from the value (upcast if necessary)
-            if dtype is None:
-                dtype, value = infer_dtype_from_scalar(value, pandas_dtype=True)
-            else:
-                # need to possibly convert the value here
-                value = maybe_cast_to_datetime(value, dtype)
-
-            subarr = construct_1d_arraylike_from_scalar(value, len(index), dtype)
+            subarr = construct_1d_arraylike_from_scalar(data, len(index), dtype)
 
         else:
             return subarr.item()
diff --git a/pandas/core/dtypes/cast.py b/pandas/core/dtypes/cast.py
index 165e63e23d60e..fafa1f51c823e 100644
--- a/pandas/core/dtypes/cast.py
+++ b/pandas/core/dtypes/cast.py
@@ -1566,7 +1566,7 @@ def find_common_type(types: List[DtypeObj]) -> DtypeObj:
 
 
 def construct_1d_arraylike_from_scalar(
-    value: Scalar, length: int, dtype: DtypeObj
+    value: Scalar, length: int, dtype: Optional[DtypeObj]
 ) -> ArrayLike:
     """
     create a np.ndarray / pandas type of specified shape and dtype
@@ -1583,6 +1583,10 @@ def construct_1d_arraylike_from_scalar(
     np.ndarray / pandas type of length, filled with value
 
     """
+
+    if dtype is None:
+        dtype, value = infer_dtype_from_scalar(value, pandas_dtype=True)
+
     if is_extension_array_dtype(dtype):
         cls = dtype.construct_array_type()
         subarr = cls._from_sequence([value] * length, dtype=dtype)
diff --git a/pandas/core/frame.py b/pandas/core/frame.py
index de60cda382fba..9cad404f2ce82 100644
--- a/pandas/core/frame.py
+++ b/pandas/core/frame.py
@@ -83,7 +83,6 @@
     infer_dtype_from_scalar,
     invalidate_string_dtypes,
     maybe_box_datetimelike,
-    maybe_cast_to_datetime,
     maybe_casted_values,
     maybe_convert_platform,
     maybe_downcast_to_dtype,
@@ -3942,14 +3941,7 @@ def reindexer(value):
                 value = maybe_infer_to_datetimelike(value)
 
         else:
-            # cast ignores pandas dtypes. so save the dtype first
-            infer_dtype, fill_value = infer_dtype_from_scalar(value, pandas_dtype=True)
-
-            value = construct_1d_arraylike_from_scalar(
-                fill_value, len(self), infer_dtype
-            )
-
-            value = maybe_cast_to_datetime(value, infer_dtype)
+            value = construct_1d_arraylike_from_scalar(value, len(self), dtype=None)
 
         # return internal types directly
         if is_extension_array_dtype(value):